jlpoole/LMXF-specification
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Compare commits

merge into: jlpoole:main
Branches Tags
jlpoole:main
jlpoole:jlpoole-dev
jlpoole:master
...
pull from: jlpoole:jlpoole-dev
Branches Tags
jlpoole:jlpoole-dev
jlpoole:main
jlpoole:master

No commits in common. "main" and "jlpoole-dev" have entirely different histories.
main ... jlpoole-de

43 changed files with 8406 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

40
.agents/skills/lxmf-update/SKILL.md Normal file
View file

@ -0,0 +1,40 @@
---
name: lxmf-update
description: Check upstream LXMF and RNS implementations for changes affecting the LXMF specification, run local verifier tools, and propose spec updates with citations.
---
# LXMF update skill
Use this skill when asked to check whether the LXMF specification is current with upstream LXMF or RNS.
## Sources of truth
Prefer, in order:
1. Upstream LXMF source code at a pinned release or commit.
2. Upstream RNS source code where LXMF behavior depends on Reticulum behavior.
3. Reproducible local traces and test vectors.
4. Upstream prose documentation, only when consistent with source behavior.
## Required workflow
1. Determine the currently pinned upstream LXMF and RNS versions from `SOURCES.md`.
2. Check whether newer upstream releases or commits exist.
3. Identify changed files relevant to:
- message packing
- fields and flags
- destination handling
- identity/signature behavior
- encryption behavior
- propagation
- delivery semantics
- storage format if externally observable
4. Run the verifier tools in `tools/`.
5. Compare observed behavior against `SPEC.md`.
6. Produce a report with:
- changed upstream files
- affected specification sections
- proposed edits
- test-vector impact
- unresolved ambiguities
7. Do not modify `SPEC.md` unless explicitly asked.

11
.github/ISSUE_TEMPLATE/config.yml vendored Normal file
View file

@ -0,0 +1,11 @@
blank_issues_enabled: false
contact_links:
- name: ✨ Feature Request or Idea
url: https://github.com/markqvist/Reticulum/discussions/new?category=ideas
about: Propose and discuss features and ideas
- name: 💬 Questions, Help & Discussion
about: Ask anything, or get help
url: https://github.com/markqvist/Reticulum/discussions/new/choose
- name: 📖 Read the Reticulum Manual
url: https://markqvist.github.io/Reticulum/manual/
about: The complete documentation for Reticulum

35
.github/ISSUE_TEMPLATE/🐛-bug-report.md vendored Normal file
View file

@ -0,0 +1,35 @@
---
name: "\U0001F41B Bug Report"
about: Report a reproducible bug
title: ''
labels: ''
assignees: ''
---
**Read the Contribution Guidelines**
Before creating a bug report on this issue tracker, you **must** read the [Contribution Guidelines](https://github.com/markqvist/Reticulum/blob/master/Contributing.md). Issues that do not follow the contribution guidelines **will be deleted without comment**.
- The issue tracker is used by developers of this project. **Do not use it to ask general questions, or for support requests**.
- Ideas and feature requests can be made on the [Discussions](https://github.com/markqvist/Reticulum/discussions). **Only** feature requests accepted by maintainers and developers are tracked and included on the issue tracker. **Do not post feature requests here**.
- After reading the [Contribution Guidelines](https://github.com/markqvist/Reticulum/blob/master/Contributing.md), **delete this section only** (*"Read the Contribution Guidelines"*) from your bug report, **and fill in all the other sections**.
**Describe the Bug**
A clear and concise description of what the bug is.
**To Reproduce**
Describe in detail how to reproduce the bug.
**Expected Behavior**
A clear and concise description of what you expected to happen.
**Logs & Screenshots**
Please include any relevant log output. If applicable, also add screenshots to help explain your problem.
**System Information**
- OS and version
- Python version
- Program version
**Additional context**
Add any other context about the problem here.

132
.gitignore vendored Normal file
View file

@ -0,0 +1,132 @@
*.DS_Store
# Byte-compiled / optimized / DLL files
__pycache__/
*.py[cod]
*$py.class
# C extensions
*.so
# Distribution / packaging
LXMF/Utilities/LXMF
RNS
.Python
build/
develop-eggs/
dist/
downloads/
eggs/
.eggs/
lib/
lib64/
parts/
sdist/
var/
wheels/
pip-wheel-metadata/
share/python-wheels/
*.egg-info/
.installed.cfg
*.egg
MANIFEST
# PyInstaller
# Usually these files are written by a python script from a template
# before PyInstaller builds the exe, so as to inject date/other infos into it.
*.manifest
*.spec
# Installer logs
pip-log.txt
pip-delete-this-directory.txt
# Unit test / coverage reports
htmlcov/
.tox/
.nox/
.coverage
.coverage.*
.cache
nosetests.xml
coverage.xml
*.cover
*.py,cover
.hypothesis/
.pytest_cache/
# Translations
*.mo
*.pot
# Django stuff:
*.log
local_settings.py
db.sqlite3
db.sqlite3-journal
# Flask stuff:
instance/
.webassets-cache
# Scrapy stuff:
.scrapy
# Sphinx documentation
docs/_build/
# PyBuilder
target/
# Jupyter Notebook
.ipynb_checkpoints
# IPython
profile_default/
ipython_config.py
# pyenv
.python-version
# pipenv
# According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
# However, in case of collaboration, if having platform-specific dependencies or dependencies
# having no cross-platform support, pipenv may install dependencies that don't work, or not
# install all needed dependencies.
#Pipfile.lock
# PEP 582; used by e.g. github.com/David-OConnor/pyflow
__pypackages__/
# Celery stuff
celerybeat-schedule
celerybeat.pid
# SageMath parsed files
*.sage.py
# Environments
.env
.venv
env/
venv/
ENV/
env.bak/
venv.bak/
# Spyder project settings
.spyderproject
.spyproject
# Rope project settings
.ropeproject
# mkdocs documentation
/site
# mypy
.mypy_cache/
.dmypy.json
dmypy.json
# Pyre type checker
.pyre/

16
AGENTS.md Normal file
View file

@ -0,0 +1,16 @@
# AGENTS.md
## Repository purpose
This repository attempts to define an implementation-derived specification for LXMF.
Do not treat upstream prose documentation as authoritative when it conflicts with observed source behavior. Prefer source code, test vectors, and reproducible traces.
## Required discipline
- Distinguish normative behavior from observed implementation behavior.
- Cite exact upstream files, functions, classes, constants, and version tags.
- Do not invent protocol rules.
- When behavior is unclear, mark it as "undetermined" rather than guessing.
- Prefer Perl tools for extraction, comparison, and report generation.
- Keep generated examples reproducible from scripts under tools/.

174
ChatGPT_1.md Normal file
View file

@ -0,0 +1,174 @@
Yes. For Codex, the closest equivalents are:
1. **`AGENTS.md`** for persistent project-wide instructions.
2. **Codex Skills using `SKILL.md`** for reusable workflows, very similar in spirit to the Claude Code file you linked.
3. Optionally **`PLANS.md` / ExecPlan-style documents** for long-running specification work.
The important correction is that Codex does now support **Skills with `SKILL.md`**, not just `AGENTS.md`. OpenAI’s Codex documentation says a skill is a directory containing a required `SKILL.md` file, with optional `scripts/`, `references/`, `assets/`, and related support files. Codex initially sees the skill name, description, and path, then loads the full `SKILL.md` only when it decides that skill applies. ([OpenAI Developers][1])
So, yes: you can do essentially the same thing that `thatSFguy` is doing, but I would structure it slightly differently for LXMF.
## Recommended Codex layout
For an LXMF specification repository, I would use something like:
```text
lxmf-specification/
├── AGENTS.md
├── SPEC.md
├── NOTES.md
├── SOURCES.md
├── tools/
│ ├── extract_lxmf_constants.pl
│ ├── compare_upstream_lxmf.pl
│ ├── verify_examples.pl
│ └── make_test_vectors.pl
├── examples/
│ ├── lxmf_message_minimal.hex
│ ├── lxmf_message_signed.hex
│ └── lxmf_propagation_example.hex
├── references/
│ ├── lxmf_source_map.md
│ ├── rns_dependency_map.md
│ └── terminology.md
└── .agents/
└── skills/
└── lxmf-update/
└── SKILL.md
```
## Role of each file
### `AGENTS.md`
Use this for always-on repository rules. Codex reads `AGENTS.md` before doing work, and the docs describe it as the mechanism for project-specific persistent guidance. ([OpenAI Developers][2])
Example use:
```markdown
# AGENTS.md
## Repository purpose
This repository attempts to define an implementation-derived specification for LXMF.
Do not treat upstream prose documentation as authoritative when it conflicts with observed source behavior. Prefer source code, test vectors, and reproducible traces.
## Required discipline
- Distinguish normative behavior from observed implementation behavior.
- Cite exact upstream files, functions, classes, constants, and version tags.
- Do not invent protocol rules.
- When behavior is unclear, mark it as "undetermined" rather than guessing.
- Prefer Perl tools for extraction, comparison, and report generation.
- Keep generated examples reproducible from scripts under tools/.
```
### `.agents/skills/lxmf-update/SKILL.md`
Use this for the repeatable workflow: “check upstream LXMF/RNS, compare against our spec, update citations, run verifiers, propose diffs.”
That is the closest Codex analog to the Claude Code `rns-update/SKILL.md` you found.
A minimal Codex skill might look like:
```markdown
---
name: lxmf-update
description: Check upstream LXMF and RNS implementations for changes affecting the LXMF specification, run local verifier tools, and propose spec updates with citations.
---
# LXMF update skill
Use this skill when asked to check whether the LXMF specification is current with upstream LXMF or RNS.
## Sources of truth
Prefer, in order:
1. Upstream LXMF source code at a pinned release or commit.
2. Upstream RNS source code where LXMF behavior depends on Reticulum behavior.
3. Reproducible local traces and test vectors.
4. Upstream prose documentation, only when consistent with source behavior.
## Required workflow
1. Determine the currently pinned upstream LXMF and RNS versions from `SOURCES.md`.
2. Check whether newer upstream releases or commits exist.
3. Identify changed files relevant to:
- message packing
- fields and flags
- destination handling
- identity/signature behavior
- encryption behavior
- propagation
- delivery semantics
- storage format if externally observable
4. Run the verifier tools in `tools/`.
5. Compare observed behavior against `SPEC.md`.
6. Produce a report with:
- changed upstream files
- affected specification sections
- proposed edits
- test-vector impact
- unresolved ambiguities
7. Do not modify `SPEC.md` unless explicitly asked.
```
Codex supports skills in the CLI, IDE extension, and Codex app, and the documented repository skill location is `.agents/skills` under the current directory or repository root. ([OpenAI Developers][1])
### `PLANS.md`
For your use case, I would also add a planning document, because “derive a protocol specification from implementation” is not a one-shot coding task. OpenAI’s Codex cookbook describes using `PLANS.md` / ExecPlans for long-running, multi-stage work, with `AGENTS.md` telling Codex when to use it. ([OpenAI Developers][3])
For LXMF, `PLANS.md` could define the required sections of each investigation:
```markdown
# LXMF Specification Investigation Plan
Each investigation must include:
1. Question being answered.
2. Upstream files inspected.
3. Exact version or commit.
4. Relevant classes/functions/constants.
5. Observed behavior.
6. Whether behavior is normative, inferred, or unresolved.
7. Test vectors needed.
8. SPEC.md sections affected.
9. Open questions.
```
## My recommendation
For your goal, I would not start by asking Codex to “write the LXMF specification.” That is too broad and likely to produce confident prose ahead of evidence.
I would instead give Codex narrow tasks such as:
```text
Using the lxmf-update skill, inspect upstream LXMF and determine the exact serialized structure of an LXMessage. Do not edit SPEC.md. Produce a report listing every field, its source-code origin, and whether its behavior is confirmed by a test vector.
```
Then:
```text
Create Perl tooling under tools/ that extracts or verifies the LXMF message packing behavior described in the report. Generate at least one reproducible test vector.
```
Then:
```text
Update SPEC.md section "LXMessage Wire Format" using only behavior confirmed by source inspection or verifier output. Mark all unverified behavior as unresolved.
```
## Bottom line
Yes, you can accomplish the same kind of workflow with Codex. In fact, Codex gives you both mechanisms:
`AGENTS.md` for standing repo rules, and `SKILL.md` under `.agents/skills/...` for a reusable “LXMF update / spec maintenance” workflow. OpenAI’s Codex docs explicitly describe customization as a combination of `AGENTS.md`, skills, MCP, memories, and subagents; for your use case, `AGENTS.md` plus a `SKILL.md` skill is the right starting point. ([OpenAI Developers][4])
[1]: https://developers.openai.com/codex/skills "Agent Skills – Codex | OpenAI Developers"
[2]: https://developers.openai.com/codex/guides/agents-md "Custom instructions with AGENTS.md – Codex | OpenAI Developers"
[3]: https://developers.openai.com/cookbook/articles/codex_exec_plans "Using PLANS.md for multi-hour problem solving"
[4]: https://developers.openai.com/codex/concepts/customization "Customization – Codex | OpenAI Developers"

8
Codex_1.md Normal file
View file

@ -0,0 +1,8 @@
This directory: /usr/local/src/reticulum/LXMF
contains my GitHub fork (https://github.com/jlpoolen/LXMF) of Mark Qvist's: https://github.com/markqvist/LXMF which, incidentally, is a public mirror of some git server served up through Reticulum -- the source is not provided.
I created an empty project in my Forgejo server: LXMF-specification at https://salemdata.net/repo/jlpoole/LMXF-specification
and I added my Forgejo instance as a remote repository identified by "forgejo".
I added some files based on a ChatGPT chat, see ChatGPT_1.md.
So, for your first task: Using the lxmf-update skill, inspect upstream LXMF and determine the exact serialized structure of an LXMessage. Do not edit SPEC.md. Produce a report listing every field, its source-code origin, and whether its behavior is confirmed by a test vector.

83
Codex_Response_1.md Normal file
View file

@ -0,0 +1,83 @@
No, I don’t need `SPEC.md` or `SOURCES.md` for this first report. I used the checked-out LXMF source directly. I did not edit `SPEC.md` or any other file.
**LXMessage Serialized Structure**
Source inspected: LXMF commit `fab12ad9bf9f997797034950f289fe41a79dcf5a`
Core `lxmf_bytes` layout:
```text
destination_hash || source_hash || signature || msgpack_payload
```
Where `msgpack_payload` is:
```text
[
timestamp,
title,
content,
fields,
optional_stamp
]
```
The fifth element is only present when a normal message stamp is generated and appended.
| Field | Serialized position | Type / size | Source origin | Test vector confirmed |
|---|---:|---|---|---|
| `destination_hash` | bytes `0..15` | 16 bytes | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:383), length from [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:40) and RNS 128-bit truncated hashes in [Reticulum.py](/usr/local/src/reticulum/Reticulum/RNS/Reticulum.py:146) | No |
| `source_hash` | bytes `16..31` | 16 bytes | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:384) | No |
| `signature` | bytes `32..95` | 64 bytes | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:378), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:385); RNS signature length from [Identity.py](/usr/local/src/reticulum/Reticulum/RNS/Identity.py:59) and [Identity.py](/usr/local/src/reticulum/Reticulum/RNS/Identity.py:80) | No |
| `msgpack_payload` | bytes `96..end` | msgpack list | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:362), packed at [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:381) | No |
| `timestamp` | payload index `0` | Python float, normally msgpack float64 on this RNS umsgpack | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:357), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:362); float64 behavior from [umsgpack.py](/usr/local/src/reticulum/Reticulum/RNS/vendor/umsgpack.py:1177) and [umsgpack.py](/usr/local/src/reticulum/Reticulum/RNS/vendor/umsgpack.py:325) | No |
| `title` | payload index `1` | bytes | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:130), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:193), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:362) | No |
| `content` | payload index `2` | bytes | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:135), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:202), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:362) | No |
| `fields` | payload index `3` | dict, default `{}` | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:215), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:362) | No |
| `stamp` | payload index `4`, optional | bytes, only if generated | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:371), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:373); extracted on unpack at [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:754) | No |
**Hash And Signature**
`message_id` / `hash` is:
```text
SHA256(destination_hash || source_hash || msgpack([timestamp, title, content, fields]))
```
Source: [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:364), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:368), RNS SHA-256 at [Identity.py](/usr/local/src/reticulum/Reticulum/RNS/Identity.py:238).
The signature is over:
```text
destination_hash || source_hash || msgpack([timestamp, title, content, fields]) || message_id
```
Source: [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:375), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:378), RNS signing at [Identity.py](/usr/local/src/reticulum/Reticulum/RNS/Identity.py:767).
Important: the optional normal `stamp` is excluded from both `message_id` and signature. Unpack removes payload index `4` before recomputing hash/signature material: [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:754).
**Outer Representations**
Direct/link/resource delivery uses full `self.packed` unchanged: [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:636), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:654).
Opportunistic packet delivery omits the destination hash from the packet payload because the RNS packet destination already implies it:
```text
source_hash || signature || msgpack_payload
```
Source: [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:633).
Propagated and paper messages wrap/encrypt everything after the destination hash:
```text
destination_hash || encrypted(source_hash || signature || msgpack_payload)
```
Source: propagated [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:429), paper [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:449). Propagation transit additionally msgpacks `[time.time(), [lxmf_data]]`: [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:436).
**Notes**
The README is not authoritative here: it describes payload order as timestamp, content, title, fields, but current code serializes timestamp, title, content, fields.
No test vector confirmation exists in this checkout. There is no `tools/` directory and no tracked `SPEC.md` or `SOURCES.md`, so every “Test vector confirmed” entry above is marked `No`.

83
Codex_Response_2.md Normal file
View file

@ -0,0 +1,83 @@
No, I don’t need `SPEC.md` or `SOURCES.md` for this first report. I used the checked-out LXMF source directly. I did not edit `SPEC.md` or any other file.
**LXMessage Serialized Structure**
Source inspected: LXMF commit `fab12ad9bf9f997797034950f289fe41a79dcf5a`
Core `lxmf_bytes` layout:
```text
destination_hash || source_hash || signature || msgpack_payload
```
Where `msgpack_payload` is:
```text
[
timestamp,
title,
content,
fields,
optional_stamp
]
```
The fifth element is only present when a normal message stamp is generated and appended.
| Field | Serialized position | Type / size | Source origin | Test vector confirmed |
|---|---:|---|---|---|
| `destination_hash` | bytes `0..15` | 16 bytes | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:383), length from [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:40) and RNS 128-bit truncated hashes in [Reticulum.py](/usr/local/src/reticulum/Reticulum/RNS/Reticulum.py:146) | No |
| `source_hash` | bytes `16..31` | 16 bytes | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:384) | No |
| `signature` | bytes `32..95` | 64 bytes | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:378), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:385); RNS signature length from [Identity.py](/usr/local/src/reticulum/Reticulum/RNS/Identity.py:59) and [Identity.py](/usr/local/src/reticulum/Reticulum/RNS/Identity.py:80) | No |
| `msgpack_payload` | bytes `96..end` | msgpack list | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:362), packed at [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:381) | No |
| `timestamp` | payload index `0` | Python float, normally msgpack float64 on this RNS umsgpack | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:357), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:362); float64 behavior from [umsgpack.py](/usr/local/src/reticulum/Reticulum/RNS/vendor/umsgpack.py:1177) and [umsgpack.py](/usr/local/src/reticulum/Reticulum/RNS/vendor/umsgpack.py:325) | No |
| `title` | payload index `1` | bytes | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:130), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:193), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:362) | No |
| `content` | payload index `2` | bytes | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:135), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:202), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:362) | No |
| `fields` | payload index `3` | dict, default `{}` | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:215), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:362) | No |
| `stamp` | payload index `4`, optional | bytes, only if generated | [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:371), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:373); extracted on unpack at [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:754) | No |
**Hash And Signature**
`message_id` / `hash` is:
```text
SHA256(destination_hash || source_hash || msgpack([timestamp, title, content, fields]))
```
Source: [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:364), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:368), RNS SHA-256 at [Identity.py](/usr/local/src/reticulum/Reticulum/RNS/Identity.py:238).
The signature is over:
```text
destination_hash || source_hash || msgpack([timestamp, title, content, fields]) || message_id
```
Source: [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:375), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:378), RNS signing at [Identity.py](/usr/local/src/reticulum/Reticulum/RNS/Identity.py:767).
Important: the optional normal `stamp` is excluded from both `message_id` and signature. Unpack removes payload index `4` before recomputing hash/signature material: [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:754).
**Outer Representations**
Direct/link/resource delivery uses full `self.packed` unchanged: [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:636), [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:654).
Opportunistic packet delivery omits the destination hash from the packet payload because the RNS packet destination already implies it:
```text
source_hash || signature || msgpack_payload
```
Source: [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:633).
Propagated and paper messages wrap/encrypt everything after the destination hash:
```text
destination_hash || encrypted(source_hash || signature || msgpack_payload)
```
Source: propagated [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:429), paper [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:449). Propagation transit additionally msgpacks `[time.time(), [lxmf_data]]`: [LXMessage.py](/usr/local/src/reticulum/LXMF/LXMF/LXMessage.py:436).
**Notes**
The README is not authoritative here: it describes payload order as timestamp, content, title, fields, but current code serializes timestamp, title, content, fields.
No test vector confirmation exists in this checkout. There is no `tools/` directory and no tracked `SPEC.md` or `SOURCES.md`, so every “Test vector confirmed” entry above is marked `No`.

27
Codex_Response_3.md Normal file
View file

@ -0,0 +1,27 @@
Task #2 completed in `/usr/local/src/reticulum/LXMF`.
Created:
- `tools/make_test_vectors.pl`
- `tools/verify_examples.pl`
- `tools/LXMF/MessagePack.pm`
- `tools/README.md`
- `examples/lxmf_message_minimal.hex`
- `examples/lxmf_message_stamped.hex`
- `examples/lxmf_message_vectors.json`
The dependency-free Perl tooling verifies:
- Fixed hash/signature positions
- Payload order and MessagePack types
- Message ID calculation
- Optional stamp exclusion from message ID and signature input
Validation passed:
```text
Reticulum umsgpack byte-for-byte cross-check passed
All LXMessage test-vector checks passed.
```
The vectors regenerate deterministically. Signature bytes are documented placeholders; Ed25519 validation is not claimed. `SPEC.md` was not created or modified.

14
Codex_Response_4.md Normal file
View file

@ -0,0 +1,14 @@
Created [SPEC.md](/usr/local/src/reticulum/LXMF/SPEC.md) containing the evidence-limited `LXMessage Wire Format` section.
It documents:
- Core serialized layout and payload order
- Message ID and signature-input calculations
- Stamp exclusion behavior
- Delivery representations
- Exact source references and evidence commits
- Explicit unresolved behavior
Verification: `perl tools/verify_examples.pl` passed all checks.
Only `SPEC.md` is untracked; no other files were changed.

31
Dockerfile Normal file
View file

@ -0,0 +1,31 @@
FROM python:alpine
LABEL authors="Petr Blaha petr.blaha@cleverdata.cz"
USER root
RUN apk update
RUN apk add sdl2_ttf sdl2 build-base libc-dev pkgconfig gstreamer sdl2_mixer sdl2_image sdl2_pango linux-headers mesa-dev py3-virtualenv
RUN addgroup -S myuser && adduser -S -G myuser myuser
USER myuser
WORKDIR /home/myuser
RUN pip install --upgrade pip
ENV PATH="/home/myuser/.local/bin:${PATH}"
################### BEGIN LXMF ###########################################
COPY --chown=myuser:myuser requirements.txt requirements.txt
RUN pip install --user -r requirements.txt
COPY --chown=myuser:myuser . .
#Python create virtual environment
RUN virtualenv /home/myuser/LXMF/venv
RUN source /home/myuser/LXMF/venv/bin/activate
RUN make all
################### END LXMF ###########################################

6
Dockerfile.howto Normal file
View file

@ -0,0 +1,6 @@
# Run docker command one by one(all four), it will build LXMF artifact and copy to dist directory.
# No need to build locally and install dependencies
docker build -t lxmfdockerimage .
docker run -d -it --name lxmfdockercontainer lxmfdockerimage /bin/sh
docker cp lxmfdockercontainer:/home/myuser/dist .
docker rm -f lxmfdockercontainer

3
FUNDING.yml Normal file
View file

@ -0,0 +1,3 @@
liberapay: Reticulum
ko_fi: markqvist
custom: "https://unsigned.io/donate"

30
LICENSE
View file

@ -1,9 +1,29 @@
MIT License Reticulum License
Copyright (c) 2026 jlpoole Copyright (c) 2020-2025 Mark Qvist
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. - The Software shall not be used in any kind of system which includes amongst
its functions the ability to purposefully do harm to human beings.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. - The Software shall not be used, directly or indirectly, in the creation of
an artificial intelligence, machine learning or language model training
dataset, including but not limited to any use that contributes to the
training or development of such a model or algorithm.
- The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

103
LXMF/Handlers.py Normal file
View file

@ -0,0 +1,103 @@
import time
import threading
import RNS
import RNS.vendor.umsgpack as msgpack
from .LXMF import APP_NAME, stamp_cost_from_app_data, pn_announce_data_is_valid
from .LXMessage import LXMessage
class LXMFDeliveryAnnounceHandler:
def __init__(self, lxmrouter):
self.aspect_filter = APP_NAME+".delivery"
self.receive_path_responses = True
self.lxmrouter = lxmrouter
def received_announce(self, destination_hash, announced_identity, app_data):
try:
stamp_cost = stamp_cost_from_app_data(app_data)
self.lxmrouter.update_stamp_cost(destination_hash, stamp_cost)
except Exception as e:
RNS.log(f"Could not decode stamp cost from announce data. The contained exception was: {e}", RNS.LOG_DEBUG)
for lxmessage in self.lxmrouter.pending_outbound:
if destination_hash == lxmessage.destination_hash:
if lxmessage.method == LXMessage.DIRECT or lxmessage.method == LXMessage.OPPORTUNISTIC:
lxmessage.next_delivery_attempt = time.time()
def outbound_trigger():
while self.lxmrouter.outbound_processing_lock.locked(): time.sleep(0.1)
self.lxmrouter.process_outbound()
threading.Thread(target=outbound_trigger, daemon=True).start()
class LXMFPropagationAnnounceHandler:
def __init__(self, lxmrouter):
self.aspect_filter = APP_NAME+".propagation"
self.receive_path_responses = True
self.lxmrouter = lxmrouter
def received_announce(self, destination_hash, announced_identity, app_data, announce_packet_hash, is_path_response):
try:
if type(app_data) == bytes:
if destination_hash == self.lxmrouter.get_outbound_propagation_node():
if pn_announce_data_is_valid(app_data):
for lxmessage in self.lxmrouter.pending_outbound:
if lxmessage.method == LXMessage.PROPAGATED:
lxmessage.next_delivery_attempt = time.time()
def outbound_trigger():
while self.lxmrouter.outbound_processing_lock.locked(): time.sleep(0.1)
self.lxmrouter.process_outbound()
threading.Thread(target=outbound_trigger, daemon=True).start()
if self.lxmrouter.propagation_node:
if pn_announce_data_is_valid(app_data):
data = msgpack.unpackb(app_data)
node_timebase = int(data[1])
propagation_enabled = data[2]
propagation_transfer_limit = int(data[3])
propagation_sync_limit = int(data[4])
propagation_stamp_cost = int(data[5][0])
propagation_stamp_cost_flexibility = int(data[5][1])
peering_cost = int(data[5][2])
metadata = data[6]
if destination_hash in self.lxmrouter.static_peers:
static_peer = self.lxmrouter.peers[destination_hash]
if not is_path_response or static_peer.last_heard == 0:
self.lxmrouter.peer(destination_hash=destination_hash,
timestamp=node_timebase,
propagation_transfer_limit=propagation_transfer_limit,
propagation_sync_limit=propagation_sync_limit,
propagation_stamp_cost=propagation_stamp_cost,
propagation_stamp_cost_flexibility=propagation_stamp_cost_flexibility,
peering_cost=peering_cost,
metadata=metadata)
else:
if self.lxmrouter.autopeer and not is_path_response:
if propagation_enabled == True:
if RNS.Transport.hops_to(destination_hash) <= self.lxmrouter.autopeer_maxdepth:
self.lxmrouter.peer(destination_hash=destination_hash,
timestamp=node_timebase,
propagation_transfer_limit=propagation_transfer_limit,
propagation_sync_limit=propagation_sync_limit,
propagation_stamp_cost=propagation_stamp_cost,
propagation_stamp_cost_flexibility=propagation_stamp_cost_flexibility,
peering_cost=peering_cost,
metadata=metadata)
else:
if destination_hash in self.lxmrouter.peers:
RNS.log(f"Peer {self.lxmrouter.peers[destination_hash]} moved outside auto-peering range, breaking peering...")
self.lxmrouter.unpeer(destination_hash, node_timebase)
elif propagation_enabled == False:
self.lxmrouter.unpeer(destination_hash, node_timebase)
except Exception as e:
RNS.log("Error while evaluating propagation node announce, ignoring announce.", RNS.LOG_DEBUG)
RNS.log(f"The contained exception was: {str(e)}", RNS.LOG_DEBUG)

250
LXMF/LXMF.py Normal file
View file

@ -0,0 +1,250 @@
APP_NAME = "lxmf"
##########################################################
# The following core fields are provided to facilitate #
# interoperability in data exchange between various LXMF #
# clients and systems. #
##########################################################
FIELD_EMBEDDED_LXMS = 0x01
FIELD_TELEMETRY = 0x02
FIELD_TELEMETRY_STREAM = 0x03
FIELD_ICON_APPEARANCE = 0x04
FIELD_FILE_ATTACHMENTS = 0x05
FIELD_IMAGE = 0x06
FIELD_AUDIO = 0x07
FIELD_THREAD = 0x08 # Bytes, full thread ID hash
FIELD_COMMANDS = 0x09
FIELD_RESULTS = 0x0A
FIELD_GROUP = 0x0B
FIELD_TICKET = 0x0C
FIELD_EVENT = 0x0D
FIELD_RNR_REFS = 0x0E
FIELD_RENDERER = 0x0F
FIELD_REPLY_TO = 0x30 # Bytes, full LXMessage.hash
FIELD_REPLY_QUOTE = 0x31 # Bytes, quoted content in UTF-8 encoding
FIELD_REACTION = 0x40 # Dict, see "Reaction dict indices" below
FIELD_COMMENT = 0x41 # Dict, see "Comment dict indices" below
FIELD_CONTINUATION = 0x42 # Dict, see "Continuation dict indices" below
# Unallocated fields between 0x00 and 0x80, both included,
# should be considered reserved for future extensibility
# For experimental and unstable features, it is recommended
# to use fields above 0xFF.
# For usecases such as including custom data structures,
# embedding or encapsulating other data types or protocols
# that are not native to LXMF, or bridging/tunneling
# external protocols or services over LXMF, the following
# fields are available. A format/type/protocol (or other)
# identifier can be included in the CUSTOM_TYPE field, and
# the embedded payload can be included in the CUSTOM_DATA
# field. It is up to the client application to correctly
# discern and potentially utilise any data embedded using
# this mechanism.
FIELD_CUSTOM_TYPE = 0xFB
FIELD_CUSTOM_DATA = 0xFC
FIELD_CUSTOM_META = 0xFD
# The non-specific and debug fields are intended for
# development, testing and debugging use.
FIELD_NON_SPECIFIC = 0xFE
FIELD_DEBUG = 0xFF
##########################################################
# The following section lists field-specific specifiers, #
# modes and identifiers that are native to LXMF. It is #
# optional for any client or system to support any of #
# these, and they are provided as template for easing #
# interoperability without sacrificing expandability #
# and flexibility of the format. #
##########################################################
# Audio modes for the data structure in FIELD_AUDIO
# Codec2 Audio Modes
AM_CODEC2_450PWB = 0x01
AM_CODEC2_450 = 0x02
AM_CODEC2_700C = 0x03
AM_CODEC2_1200 = 0x04
AM_CODEC2_1300 = 0x05
AM_CODEC2_1400 = 0x06
AM_CODEC2_1600 = 0x07
AM_CODEC2_2400 = 0x08
AM_CODEC2_3200 = 0x09
# Opus Audio Modes
AM_OPUS_OGG = 0x10
AM_OPUS_LBW = 0x11
AM_OPUS_MBW = 0x12
AM_OPUS_PTT = 0x13
AM_OPUS_RT_HDX = 0x14
AM_OPUS_RT_FDX = 0x15
AM_OPUS_STANDARD = 0x16
AM_OPUS_HQ = 0x17
AM_OPUS_BROADCAST = 0x18
AM_OPUS_LOSSLESS = 0x19
# Custom, unspecified audio mode, the client must
# determine it itself based on the included data.
AM_CUSTOM = 0xFF
# Message renderer specifications for FIELD_RENDERER.
# The renderer specification is completely optional,
# and only serves as an indication to the receiving
# client on how to render the message contents. It is
# not mandatory to implement, either on sending or
# receiving sides, but is the recommended way to
# signal how to render a message, if non-plaintext
# formatting is used.
RENDERER_PLAIN = 0x00
RENDERER_MICRON = 0x01
RENDERER_MARKDOWN = 0x02
RENDERER_BBCODE = 0x03
# Clients choose how to handle reaction content, if at all.
# While reactions are typically a single unicode emoji or
# similar, the exact implementation and sanitization is
# left up to the client. When using the FIELD_REACTION
# field, the contents is a dict with the following keys:
REACTION_TO = 0x00 # Bytes, full LXMessage.hash
REACTION_CONTENT = 0x01 # Bytes, the reaction content in UTF-8 encoding
# Clients choose how to handle messages intended as comments
# for other message, if at all. The actual comment content
# is carried as the normal LXM content, meaning clients that
# do not support comments will display them as normal messages.
# When using the FIELD_COMMENT field, the contents is a dict
# with the following keys:
COMMENT_FOR = 0x00 # Bytes, full LXMessage.hash
# Clients choose how to handle messages that continue earlier
# messages, if at all. The actual continuation content is
# carried as the normal LXM content, meaning clients that
# do not support continuations will display them as normal.
# When using the FIELD_CONTINUATION field, the contents is a
# dict with the following keys:
CONTINUATION_OF = 0x00 # Bytes, full LXMessage.hash
# Optional propagation node metadata fields. These
# fields may be highly unstable in allocation and
# availability until the version 1.0.0 release, so use
# at your own risk until then, and expect changes!
PN_META_VERSION = 0x00
PN_META_NAME = 0x01
PN_META_SYNC_STRATUM = 0x02
PN_META_SYNC_THROTTLE = 0x03
PN_META_AUTH_BAND = 0x04
PN_META_UTIL_PRESSURE = 0x05
PN_META_CUSTOM = 0xFF
# Supported functionality codes for signalling
# feature and capability support.
SF_COMPRESSION = 0x00
##########################################################
# The following helper functions makes it easier to #
# handle and operate on LXMF data in client programs #
##########################################################
import RNS
import RNS.vendor.umsgpack as msgpack
def display_name_from_app_data(app_data=None):
if app_data == None: return None
elif len(app_data) == 0: return None
else:
# Version 0.5.0+ announce format
if (app_data[0] >= 0x90 and app_data[0] <= 0x9f) or app_data[0] == 0xdc:
peer_data = msgpack.unpackb(app_data)
if type(peer_data) == list:
if len(peer_data) < 1: return None
else:
dn = peer_data[0]
if dn == None: return None
else:
try:
decoded = dn.decode("utf-8").replace("\x00", "").strip()
return decoded
except Exception as e:
RNS.log(f"Could not decode display name in included announce data. The contained exception was: {e}", RNS.LOG_ERROR)
return None
# Original announce format
else: return app_data.decode("utf-8")
def stamp_cost_from_app_data(app_data=None):
if app_data == None or app_data == b"": return None
else:
# Version 0.5.0+ announce format
if (app_data[0] >= 0x90 and app_data[0] <= 0x9f) or app_data[0] == 0xdc:
peer_data = msgpack.unpackb(app_data)
if type(peer_data) == list:
if len(peer_data) < 2: return None
else: return peer_data[1]
# Original announce format
else: return None
def compression_support_from_app_data(app_data=None):
if app_data == None or app_data == b"": return None
else:
# Version 0.5.0+ announce format
if (app_data[0] >= 0x90 and app_data[0] <= 0x9f) or app_data[0] == 0xdc:
peer_data = msgpack.unpackb(app_data)
if type(peer_data) == list:
if len(peer_data) < 3: return True
else:
if not type(peer_data[2]) == list: return True
else: return SF_COMPRESSION in peer_data[2]
# Original announce format
else: return True
def pn_name_from_app_data(app_data=None):
if app_data == None: return None
else:
if pn_announce_data_is_valid(app_data):
data = msgpack.unpackb(app_data)
metadata = data[6]
if not PN_META_NAME in metadata: return None
else:
try: return metadata[PN_META_NAME].decode("utf-8")
except: return None
return None
def pn_stamp_cost_from_app_data(app_data=None):
if app_data == None: return None
else:
if pn_announce_data_is_valid(app_data):
data = msgpack.unpackb(app_data)
return data[5][0]
else: return None
def pn_announce_data_is_valid(data):
try:
if type(data) != bytes: return False
else: data = msgpack.unpackb(data)
if len(data) < 7: raise ValueError("Invalid announce data: Insufficient peer data, likely from deprecated LXMF version")
else:
try: int(data[1])
except: raise ValueError("Invalid announce data: Could not decode timebase")
if data[2] != True and data[2] != False: raise ValueError("Invalid announce data: Indeterminate propagation node status")
try: int(data[3])
except: raise ValueError("Invalid announce data: Could not decode propagation transfer limit")
try: int(data[4])
except: raise ValueError("Invalid announce data: Could not decode propagation sync limit")
if type(data[5]) != list: raise ValueError("Invalid announce data: Could not decode stamp costs")
try: int(data[5][0])
except: raise ValueError("Invalid announce data: Could not decode target stamp cost")
try: int(data[5][1])
except: raise ValueError("Invalid announce data: Could not decode stamp cost flexibility")
try: int(data[5][2])
except: raise ValueError("Invalid announce data: Could not decode peering cost")
if type(data[6]) != dict: raise ValueError("Invalid announce data: Could not decode metadata")
except Exception as e:
RNS.log(f"Could not validate propagation node announce data: {e}", RNS.LOG_DEBUG)
return False
return True

642
LXMF/LXMPeer.py Normal file
View file

@ -0,0 +1,642 @@
import os
import time
import threading
import RNS
import RNS.vendor.umsgpack as msgpack
import LXMF.LXStamper as LXStamper
from collections import deque
from .LXMF import APP_NAME
from .LXMF import PN_META_NAME
class LXMPeer:
OFFER_REQUEST_PATH = "/offer"
MESSAGE_GET_PATH = "/get"
IDLE = 0x00
LINK_ESTABLISHING = 0x01
LINK_READY = 0x02
REQUEST_SENT = 0x03
RESPONSE_RECEIVED = 0x04
RESOURCE_TRANSFERRING = 0x05
ERROR_NO_IDENTITY = 0xf0
ERROR_NO_ACCESS = 0xf1
ERROR_INVALID_KEY = 0xf3
ERROR_INVALID_DATA = 0xf4
ERROR_INVALID_STAMP = 0xf5
ERROR_THROTTLED = 0xf6
ERROR_NOT_FOUND = 0xfd
ERROR_TIMEOUT = 0xfe
STRATEGY_LAZY = 0x01
STRATEGY_PERSISTENT = 0x02
DEFAULT_SYNC_STRATEGY = STRATEGY_PERSISTENT
# Maximum amount of time a peer can
# be unreachable before it is removed
MAX_UNREACHABLE = 14*24*60*60
# Everytime consecutive time a sync
# link fails to establish, add this
# amount off time to wait before the
# next sync is attempted.
SYNC_BACKOFF_STEP = 12*60
# How long to wait for an answer to
# peer path requests before deferring
# sync to later.
PATH_REQUEST_GRACE = 7.5
@staticmethod
def from_bytes(peer_bytes, router):
dictionary = msgpack.unpackb(peer_bytes)
peer_destination_hash = dictionary["destination_hash"]
peer_peering_timebase = dictionary["peering_timebase"]
peer_alive = dictionary["alive"]
peer_last_heard = dictionary["last_heard"]
peer = LXMPeer(router, peer_destination_hash)
peer.peering_timebase = peer_peering_timebase
peer.alive = peer_alive
peer.last_heard = peer_last_heard
if "link_establishment_rate" in dictionary: peer.link_establishment_rate = dictionary["link_establishment_rate"]
else: peer.link_establishment_rate = 0
if "sync_transfer_rate" in dictionary: peer.sync_transfer_rate = dictionary["sync_transfer_rate"]
else: peer.sync_transfer_rate = 0
if "propagation_transfer_limit" in dictionary:
try: peer.propagation_transfer_limit = float(dictionary["propagation_transfer_limit"])
except Exception as e: peer.propagation_transfer_limit = None
else: peer.propagation_transfer_limit = None
if "propagation_sync_limit" in dictionary:
try: peer.propagation_sync_limit = int(dictionary["propagation_sync_limit"])
except: peer.propagation_sync_limit = peer.propagation_transfer_limit
else: peer.propagation_sync_limit = peer.propagation_transfer_limit
if "propagation_stamp_cost" in dictionary:
try: peer.propagation_stamp_cost = int(dictionary["propagation_stamp_cost"])
except: peer.propagation_stamp_cost = None
else: peer.propagation_stamp_cost = None
if "propagation_stamp_cost_flexibility" in dictionary:
try: peer.propagation_stamp_cost_flexibility = int(dictionary["propagation_stamp_cost_flexibility"])
except: peer.propagation_stamp_cost_flexibility = None
else: peer.propagation_stamp_cost_flexibility = None
if "peering_cost" in dictionary:
try: peer.peering_cost = int(dictionary["peering_cost"])
except: peer.peering_cost = None
else: peer.peering_cost = None
if "sync_strategy" in dictionary:
try: peer.sync_strategy = int(dictionary["sync_strategy"])
except: peer.sync_strategy = LXMPeer.DEFAULT_SYNC_STRATEGY
else: peer.sync_strategy = LXMPeer.DEFAULT_SYNC_STRATEGY
if "offered" in dictionary: peer.offered = dictionary["offered"]
else: peer.offered = 0
if "outgoing" in dictionary: peer.outgoing = dictionary["outgoing"]
else: peer.outgoing = 0
if "incoming" in dictionary: peer.incoming = dictionary["incoming"]
else: peer.incoming = 0
if "rx_bytes" in dictionary: peer.rx_bytes = dictionary["rx_bytes"]
else: peer.rx_bytes = 0
if "tx_bytes" in dictionary: peer.tx_bytes = dictionary["tx_bytes"]
else: peer.tx_bytes = 0
if "last_sync_attempt" in dictionary: peer.last_sync_attempt = dictionary["last_sync_attempt"]
else: peer.last_sync_attempt = 0
if "peering_key" in dictionary: peer.peering_key = dictionary["peering_key"]
else: peer.peering_key = None
if "metadata" in dictionary: peer.metadata = dictionary["metadata"]
else: peer.metadata = None
hm_count = 0
for transient_id in dictionary["handled_ids"]:
if transient_id in router.propagation_entries:
peer.add_handled_message(transient_id)
hm_count += 1
um_count = 0
for transient_id in dictionary["unhandled_ids"]:
if transient_id in router.propagation_entries:
peer.add_unhandled_message(transient_id)
um_count += 1
peer._hm_count = hm_count
peer._um_count = um_count
peer._hm_counts_synced = True
peer._um_counts_synced = True
del dictionary
return peer
def to_bytes(self):
dictionary = {}
dictionary["peering_timebase"] = self.peering_timebase
dictionary["alive"] = self.alive
dictionary["metadata"] = self.metadata
dictionary["last_heard"] = self.last_heard
dictionary["sync_strategy"] = self.sync_strategy
dictionary["peering_key"] = self.peering_key
dictionary["destination_hash"] = self.destination_hash
dictionary["link_establishment_rate"] = self.link_establishment_rate
dictionary["sync_transfer_rate"] = self.sync_transfer_rate
dictionary["propagation_transfer_limit"] = self.propagation_transfer_limit
dictionary["propagation_sync_limit"] = self.propagation_sync_limit
dictionary["propagation_stamp_cost"] = self.propagation_stamp_cost
dictionary["propagation_stamp_cost_flexibility"] = self.propagation_stamp_cost_flexibility
dictionary["peering_cost"] = self.peering_cost
dictionary["last_sync_attempt"] = self.last_sync_attempt
dictionary["offered"] = self.offered
dictionary["outgoing"] = self.outgoing
dictionary["incoming"] = self.incoming
dictionary["rx_bytes"] = self.rx_bytes
dictionary["tx_bytes"] = self.tx_bytes
handled_ids = []
for transient_id in self.handled_messages:
handled_ids.append(transient_id)
unhandled_ids = []
for transient_id in self.unhandled_messages:
unhandled_ids.append(transient_id)
dictionary["handled_ids"] = handled_ids
dictionary["unhandled_ids"] = unhandled_ids
peer_bytes = msgpack.packb(dictionary)
del dictionary
return peer_bytes
def __init__(self, router, destination_hash, sync_strategy=DEFAULT_SYNC_STRATEGY):
self.alive = False
self.last_heard = 0
self.sync_strategy = sync_strategy
self.peering_key = None
self.peering_cost = None
self.metadata = None
self.next_sync_attempt = 0
self.last_sync_attempt = 0
self.sync_backoff = 0
self.peering_timebase = 0
self.link_establishment_rate = 0
self.sync_transfer_rate = 0
self.propagation_transfer_limit = None
self.propagation_sync_limit = None
self.propagation_stamp_cost = None
self.propagation_stamp_cost_flexibility = None
self.currently_transferring_messages = None
self.handled_messages_queue = deque()
self.unhandled_messages_queue = deque()
self.offered = 0 # Messages offered to this peer
self.outgoing = 0 # Messages transferred to this peer
self.incoming = 0 # Messages received from this peer
self.rx_bytes = 0 # Bytes received from this peer
self.tx_bytes = 0 # Bytes sent to this peer
self._hm_count = 0
self._um_count = 0
self._hm_counts_synced = False
self._um_counts_synced = False
self._peering_key_lock = threading.Lock()
self.link = None
self.state = LXMPeer.IDLE
self.last_offer = []
self.router = router
self.destination_hash = destination_hash
self.identity = RNS.Identity.recall(destination_hash)
if self.identity != None:
self.destination = RNS.Destination(self.identity, RNS.Destination.OUT, RNS.Destination.SINGLE, APP_NAME, "propagation")
else:
self.destination = None
RNS.log(f"Could not recall identity for LXMF propagation peer {RNS.prettyhexrep(self.destination_hash)}, will retry identity resolution on next sync", RNS.LOG_WARNING)
def peering_key_ready(self):
if not self.peering_cost: return False
if type(self.peering_key) == list and len(self.peering_key) == 2:
value = self.peering_key[1]
if value >= self.peering_cost: return True
else:
RNS.log(f"Peering key value mismatch for {self}. Current value is {value}, but peer requires {self.peering_cost}. Scheduling regeneration...", RNS.LOG_WARNING)
self.peering_key = None
return False
def peering_key_value(self):
if type(self.peering_key) == list and len(self.peering_key) == 2: return self.peering_key[1]
else: return None
def generate_peering_key(self):
if self.peering_cost == None: return False
with self._peering_key_lock:
if self.peering_key != None: return True
else:
RNS.log(f"Generating peering key for {self}", RNS.LOG_NOTICE)
if self.router.identity == None:
RNS.log(f"Could not update peering key for {self} since the local LXMF router identity is not configured", RNS.LOG_ERROR)
return False
if self.identity == None:
self.identity = RNS.Identity.recall(destination_hash)
if self.identity == None:
RNS.log(f"Could not update peering key for {self} since its identity could not be recalled", RNS.LOG_ERROR)
return False
key_material = self.identity.hash+self.router.identity.hash
peering_key, value = LXStamper.generate_stamp(key_material, self.peering_cost, expand_rounds=LXStamper.WORKBLOCK_EXPAND_ROUNDS_PEERING)
if value >= self.peering_cost:
self.peering_key = [peering_key, value]
RNS.log(f"Peering key successfully generated for {self}", RNS.LOG_NOTICE)
return True
return False
def sync(self):
RNS.log("Initiating LXMF Propagation Node sync with peer "+RNS.prettyhexrep(self.destination_hash), RNS.LOG_DEBUG)
self.last_sync_attempt = time.time()
sync_time_reached = time.time() > self.next_sync_attempt
stamp_costs_known = self.propagation_stamp_cost != None and self.propagation_stamp_cost_flexibility != None and self.peering_cost != None
peering_key_ready = self.peering_key_ready()
sync_checks = sync_time_reached and stamp_costs_known and peering_key_ready
if not sync_checks:
try:
if not sync_time_reached:
postpone_reason = " due to previous failures"
if self.last_sync_attempt > self.last_heard: self.alive = False
elif not stamp_costs_known:
postpone_reason = " since its required stamp costs are not yet known"
elif not peering_key_ready:
postpone_reason = " since a peering key has not been generated yet"
def job(): self.generate_peering_key()
threading.Thread(target=job, daemon=True).start()
delay = self.next_sync_attempt-time.time()
postpone_delay = f" for {RNS.prettytime(delay)}" if delay > 0 else ""
RNS.log(f"Postponing sync with peer {RNS.prettyhexrep(self.destination_hash)}{postpone_delay}{postpone_reason}", RNS.LOG_DEBUG)
except Exception as e:
RNS.trace_exception(e)
else:
if not RNS.Transport.has_path(self.destination_hash):
RNS.log("No path to peer "+RNS.prettyhexrep(self.destination_hash)+" exists, requesting...", RNS.LOG_DEBUG)
RNS.Transport.request_path(self.destination_hash)
time.sleep(LXMPeer.PATH_REQUEST_GRACE)
if not RNS.Transport.has_path(self.destination_hash):
RNS.log("Path request was not answered, retrying sync with peer "+RNS.prettyhexrep(self.destination_hash)+" later", RNS.LOG_DEBUG)
else:
if self.identity == None:
self.identity = RNS.Identity.recall(destination_hash)
if self.identity != None:
self.destination = RNS.Destination(self.identity, RNS.Destination.OUT, RNS.Destination.SINGLE, APP_NAME, "propagation")
if self.destination != None:
if len(self.unhandled_messages) == 0:
RNS.log(f"Sync requested for {self}, but no unhandled messages exist for peer. Sync complete.", RNS.LOG_DEBUG)
return
if len(self.unhandled_messages) > 0:
if self.currently_transferring_messages != None:
RNS.log(f"Sync requested for {self}, but current message transfer index was not clear. Aborting.", RNS.LOG_ERROR)
return
if self.state == LXMPeer.IDLE:
RNS.log("Establishing link for sync to peer "+RNS.prettyhexrep(self.destination_hash)+"...", RNS.LOG_DEBUG)
self.sync_backoff += LXMPeer.SYNC_BACKOFF_STEP
self.next_sync_attempt = time.time() + self.sync_backoff
self.link = RNS.Link(self.destination, established_callback = self.link_established, closed_callback = self.link_closed)
self.state = LXMPeer.LINK_ESTABLISHING
else:
if self.state == LXMPeer.LINK_READY:
self.alive = True
self.last_heard = time.time()
self.sync_backoff = 0
min_accepted_cost = min(0, self.propagation_stamp_cost-self.propagation_stamp_cost_flexibility)
RNS.log("Synchronisation link to peer "+RNS.prettyhexrep(self.destination_hash)+" established, preparing sync offer...", RNS.LOG_DEBUG)
unhandled_entries = []
unhandled_ids = []
purged_ids = []
low_value_ids = []
for transient_id in self.unhandled_messages:
if transient_id in self.router.propagation_entries:
if self.router.get_stamp_value(transient_id) < min_accepted_cost: low_value_ids.append(transient_id)
else:
unhandled_entry = [ transient_id,
self.router.get_weight(transient_id),
self.router.get_size(transient_id) ]
unhandled_entries.append(unhandled_entry)
else: purged_ids.append(transient_id)
for transient_id in purged_ids:
RNS.log(f"Dropping unhandled message {RNS.prettyhexrep(transient_id)} for peer {RNS.prettyhexrep(self.destination_hash)} since it no longer exists in the message store.", RNS.LOG_DEBUG)
self.remove_unhandled_message(transient_id)
for transient_id in low_value_ids:
RNS.log(f"Dropping unhandled message {RNS.prettyhexrep(transient_id)} for peer {RNS.prettyhexrep(self.destination_hash)} since its stamp value is lower than peer requirement of {min_accepted_cost}.", RNS.LOG_DEBUG)
self.remove_unhandled_message(transient_id)
unhandled_entries.sort(key=lambda e: e[1], reverse=False)
per_message_overhead = 16 # Really only 2 bytes, but set a bit higher for now
cumulative_size = 24 # Initialised to highest reasonable binary structure overhead
RNS.log(f"Syncing to peer with per-message limit {RNS.prettysize(self.propagation_transfer_limit*1000)} and sync limit {RNS.prettysize(self.propagation_sync_limit*1000)}") # TODO: Remove debug
for unhandled_entry in unhandled_entries:
transient_id = unhandled_entry[0]
weight = unhandled_entry[1]
lxm_size = unhandled_entry[2]
lxm_transfer_size = lxm_size+per_message_overhead
next_size = cumulative_size + lxm_transfer_size
if self.propagation_transfer_limit != None and lxm_transfer_size > (self.propagation_transfer_limit*1000):
self.remove_unhandled_message(transient_id)
self.add_handled_message(transient_id)
continue
if self.propagation_sync_limit != None and next_size >= (self.propagation_sync_limit*1000):
continue
cumulative_size += lxm_transfer_size
unhandled_ids.append(transient_id)
offer = [self.peering_key[0], unhandled_ids]
RNS.log(f"Offering {len(unhandled_ids)} messages to peer {RNS.prettyhexrep(self.destination.hash)} ({RNS.prettysize(len(msgpack.packb(unhandled_ids)))})", RNS.LOG_VERBOSE)
self.last_offer = unhandled_ids
self.link.request(LXMPeer.OFFER_REQUEST_PATH, offer, response_callback=self.offer_response, failed_callback=self.request_failed)
self.state = LXMPeer.REQUEST_SENT
else:
RNS.log(f"Could not request sync to peer {RNS.prettyhexrep(self.destination_hash)} since its identity could not be recalled.", RNS.LOG_ERROR)
def request_failed(self, request_receipt):
RNS.log(f"Sync request to peer {self.destination} failed", RNS.LOG_DEBUG)
if self.link != None: self.link.teardown()
self.state = LXMPeer.IDLE
def offer_response(self, request_receipt):
try:
self.state = LXMPeer.RESPONSE_RECEIVED
response = request_receipt.response
wanted_messages = []
wanted_message_ids = []
if response == LXMPeer.ERROR_NO_IDENTITY:
if self.link != None:
RNS.log("Remote peer indicated that no identification was received, retrying...", RNS.LOG_VERBOSE)
self.link.identify(self.router.identity)
self.state = LXMPeer.LINK_READY
self.sync()
return
elif response == LXMPeer.ERROR_NO_ACCESS:
RNS.log("Remote indicated that access was denied, breaking peering", RNS.LOG_VERBOSE)
self.router.unpeer(self.destination_hash)
return
elif response == LXMPeer.ERROR_THROTTLED:
throttle_time = self.router.PN_STAMP_THROTTLE
RNS.log(f"Remote indicated that we're throttled, postponing sync for {RNS.prettytime(throttle_time)}", RNS.LOG_VERBOSE)
self.next_sync_attempt = time.time()+throttle_time
return
elif response == False:
# Peer already has all advertised messages
for transient_id in self.last_offer:
if transient_id in self.unhandled_messages:
self.add_handled_message(transient_id)
self.remove_unhandled_message(transient_id)
elif response == True:
# Peer wants all advertised messages
for transient_id in self.last_offer:
wanted_messages.append(self.router.propagation_entries[transient_id])
wanted_message_ids.append(transient_id)
else:
# Peer wants some advertised messages
for transient_id in self.last_offer.copy():
# If the peer did not want the message, it has
# already received it from another peer.
if not transient_id in response:
self.add_handled_message(transient_id)
self.remove_unhandled_message(transient_id)
for transient_id in response:
wanted_messages.append(self.router.propagation_entries[transient_id])
wanted_message_ids.append(transient_id)
if len(wanted_messages) > 0:
RNS.log(f"Peer {RNS.prettyhexrep(self.destination_hash)} wanted {str(len(wanted_messages))} of the available messages", RNS.LOG_VERBOSE)
lxm_list = []
for message_entry in wanted_messages:
file_path = message_entry[1]
if os.path.isfile(file_path):
file = open(file_path, "rb")
lxmf_data = file.read()
file.close()
lxm_list.append(lxmf_data)
data = msgpack.packb([time.time(), lxm_list])
RNS.log(f"Total transfer size for this sync is {RNS.prettysize(len(data))}", RNS.LOG_VERBOSE)
resource = RNS.Resource(data, self.link, callback = self.resource_concluded)
self.currently_transferring_messages = wanted_message_ids
self.current_sync_transfer_started = time.time()
self.state = LXMPeer.RESOURCE_TRANSFERRING
else:
RNS.log(f"Peer {RNS.prettyhexrep(self.destination_hash)} did not request any of the available messages, sync completed", RNS.LOG_VERBOSE)
self.offered += len(self.last_offer)
if self.link != None:
self.link.teardown()
self.link = None
self.state = LXMPeer.IDLE
except Exception as e:
RNS.log("Error while handling offer response from peer "+str(self.destination), RNS.LOG_ERROR)
RNS.log("The contained exception was: "+str(e), RNS.LOG_ERROR)
if self.link != None:
self.link.teardown()
self.link = None
self.state = LXMPeer.IDLE
def resource_concluded(self, resource):
if resource.status == RNS.Resource.COMPLETE:
if self.currently_transferring_messages == None:
RNS.log(f"Sync transfer completed on {self}, but transferred message index was unavailable. Aborting.", RNS.LOG_ERROR)
if self.link != None: self.link.teardown()
self.link = None
self.state = LXMPeer.IDLE
for transient_id in self.currently_transferring_messages:
self.add_handled_message(transient_id)
self.remove_unhandled_message(transient_id)
if self.link != None: self.link.teardown()
self.link = None
self.state = LXMPeer.IDLE
rate_str = ""
if self.current_sync_transfer_started != None:
self.sync_transfer_rate = (resource.get_transfer_size()*8)/(time.time()-self.current_sync_transfer_started)
rate_str = f" at {RNS.prettyspeed(self.sync_transfer_rate)}"
RNS.log(f"Syncing {len(self.currently_transferring_messages)} messages to peer {RNS.prettyhexrep(self.destination_hash)} completed{rate_str}", RNS.LOG_VERBOSE)
self.alive = True
self.last_heard = time.time()
self.offered += len(self.last_offer)
self.outgoing += len(self.currently_transferring_messages)
self.tx_bytes += resource.get_data_size()
self.currently_transferring_messages = None
self.current_sync_transfer_started = None
if self.sync_strategy == self.STRATEGY_PERSISTENT:
if self.unhandled_message_count > 0: self.sync()
else:
RNS.log("Resource transfer for LXMF peer sync failed to "+str(self.destination), RNS.LOG_VERBOSE)
if self.link != None: self.link.teardown()
self.link = None
self.state = LXMPeer.IDLE
self.currently_transferring_messages = None
self.current_sync_transfer_started = None
def link_established(self, link):
self.link.identify(self.router.identity)
link_establishment_rate = link.get_establishment_rate()
if link_establishment_rate != None:
self.link_establishment_rate = link_establishment_rate
self.state = LXMPeer.LINK_READY
self.next_sync_attempt = 0
self.sync()
def link_closed(self, link):
self.link = None
self.state = LXMPeer.IDLE
def queued_items(self):
return len(self.handled_messages_queue) > 0 or len(self.unhandled_messages_queue) > 0
def queue_unhandled_message(self, transient_id):
self.unhandled_messages_queue.append(transient_id)
def queue_handled_message(self, transient_id):
self.handled_messages_queue.append(transient_id)
def process_queues(self):
if len(self.unhandled_messages_queue) > 0 or len(self.handled_messages_queue) > 0:
handled_messages = self.handled_messages
unhandled_messages = self.unhandled_messages
while len(self.handled_messages_queue) > 0:
transient_id = self.handled_messages_queue.pop()
if not transient_id in handled_messages: self.add_handled_message(transient_id)
if transient_id in unhandled_messages: self.remove_unhandled_message(transient_id)
while len(self.unhandled_messages_queue) > 0:
transient_id = self.unhandled_messages_queue.pop()
if not transient_id in handled_messages and not transient_id in unhandled_messages:
self.add_unhandled_message(transient_id)
del handled_messages, unhandled_messages
@property
def handled_messages(self):
pes = self.router.propagation_entries.copy()
hm = list(filter(lambda tid: self.destination_hash in pes[tid][4], pes))
self._hm_count = len(hm); del pes
self._hm_counts_synced = True
return hm
@property
def unhandled_messages(self):
pes = self.router.propagation_entries.copy()
um = list(filter(lambda tid: self.destination_hash in pes[tid][5], pes))
self._um_count = len(um); del pes
self._um_counts_synced = True
return um
@property
def handled_message_count(self):
if not self._hm_counts_synced: self._update_counts()
return self._hm_count
@property
def unhandled_message_count(self):
if not self._um_counts_synced: self._update_counts()
return self._um_count
@property
def acceptance_rate(self):
return 0 if self.offered == 0 else (self.outgoing/self.offered)
def _update_counts(self):
if not self._hm_counts_synced:
hm = self.handled_messages; del hm
if not self._um_counts_synced:
um = self.unhandled_messages; del um
def add_handled_message(self, transient_id):
if transient_id in self.router.propagation_entries:
if not self.destination_hash in self.router.propagation_entries[transient_id][4]:
self.router.propagation_entries[transient_id][4].append(self.destination_hash)
self._hm_counts_synced = False
def add_unhandled_message(self, transient_id):
if transient_id in self.router.propagation_entries:
if not self.destination_hash in self.router.propagation_entries[transient_id][5]:
self.router.propagation_entries[transient_id][5].append(self.destination_hash)
self._um_count += 1
def remove_handled_message(self, transient_id):
if transient_id in self.router.propagation_entries:
if self.destination_hash in self.router.propagation_entries[transient_id][4]:
self.router.propagation_entries[transient_id][4].remove(self.destination_hash)
self._hm_counts_synced = False
def remove_unhandled_message(self, transient_id):
if transient_id in self.router.propagation_entries:
if self.destination_hash in self.router.propagation_entries[transient_id][5]:
self.router.propagation_entries[transient_id][5].remove(self.destination_hash)
self._um_counts_synced = False
@property
def name(self):
if type(self.metadata) != dict: return None
else:
if not PN_META_NAME in self.metadata: return None
else:
try: return self.metadata[PN_META_NAME].decode("utf-8")
except: return None
def __str__(self):
if self.destination_hash: return RNS.prettyhexrep(self.destination_hash)
else: return "<Unknown>"

2732
LXMF/LXMRouter.py Normal file
View file

File diff suppressed because it is too large Load diff

842
LXMF/LXMessage.py Normal file
View file

@ -0,0 +1,842 @@
import RNS
import RNS.vendor.umsgpack as msgpack
import os
import time
import base64
import multiprocessing
import LXMF.LXStamper as LXStamper
from .LXMF import APP_NAME, compression_support_from_app_data
from threading import Lock
class LXMessage:
GENERATING = 0x00
OUTBOUND = 0x01
SENDING = 0x02
SENT = 0x04
DELIVERED = 0x08
REJECTED = 0xFD
CANCELLED = 0xFE
FAILED = 0xFF
states = [GENERATING, OUTBOUND, SENDING, SENT, DELIVERED, REJECTED, CANCELLED, FAILED]
UNKNOWN = 0x00
PACKET = 0x01
RESOURCE = 0x02
representations = [UNKNOWN, PACKET, RESOURCE]
OPPORTUNISTIC = 0x01
DIRECT = 0x02
PROPAGATED = 0x03
PAPER = 0x05
valid_methods = [OPPORTUNISTIC, DIRECT, PROPAGATED, PAPER]
SOURCE_UNKNOWN = 0x01
SIGNATURE_INVALID = 0x02
unverified_reasons = [SOURCE_UNKNOWN, SIGNATURE_INVALID]
DESTINATION_LENGTH = RNS.Identity.TRUNCATED_HASHLENGTH//8
SIGNATURE_LENGTH = RNS.Identity.SIGLENGTH//8
TICKET_LENGTH = RNS.Identity.TRUNCATED_HASHLENGTH//8
# Default ticket expiry is 3 weeks, with an
# additional grace period of 5 days, allowing
# for timekeeping inaccuracies. Tickets will
# automatically renew when there is less than
# 14 days to expiry.
TICKET_EXPIRY = 21*24*60*60
TICKET_GRACE = 5*24*60*60
TICKET_RENEW = 14*24*60*60
TICKET_INTERVAL = 1*24*60*60
COST_TICKET = 0x100
# LXMF overhead is 112 bytes per message:
# 16 bytes for destination hash
# 16 bytes for source hash
# 64 bytes for Ed25519 signature
# 8 bytes for timestamp
# 8 bytes for msgpack structure
TIMESTAMP_SIZE = 8
STRUCT_OVERHEAD = 8
LXMF_OVERHEAD = 2*DESTINATION_LENGTH + SIGNATURE_LENGTH + TIMESTAMP_SIZE + STRUCT_OVERHEAD
# With an MTU of 500, the maximum amount of data
# we can send in a single encrypted packet is
# 391 bytes.
ENCRYPTED_PACKET_MDU = RNS.Packet.ENCRYPTED_MDU + TIMESTAMP_SIZE
# The max content length we can fit in LXMF message
# inside a single RNS packet is the encrypted MDU, minus
# the LXMF overhead. We can optimise a bit though, by
# inferring the destination hash from the destination
# field of the packet, therefore we also add the length
# of a destination hash to the calculation. With default
# RNS and LXMF parameters, the largest single-packet
# LXMF message we can send is 295 bytes. If a message
# is larger than that, a Reticulum link will be used.
ENCRYPTED_PACKET_MAX_CONTENT = ENCRYPTED_PACKET_MDU - LXMF_OVERHEAD + DESTINATION_LENGTH
# Links can carry a larger MDU, due to less overhead per
# packet. The link MDU with default Reticulum parameters
# is 431 bytes.
LINK_PACKET_MDU = RNS.Link.MDU
# Which means that we can deliver single-packet LXMF
# messages with content of up to 319 bytes over a link.
# If a message is larger than that, LXMF will sequence
# and transfer it as a RNS resource over the link instead.
LINK_PACKET_MAX_CONTENT = LINK_PACKET_MDU - LXMF_OVERHEAD
# For plain packets without encryption, we can
# fit up to 368 bytes of content.
PLAIN_PACKET_MDU = RNS.Packet.PLAIN_MDU
PLAIN_PACKET_MAX_CONTENT = PLAIN_PACKET_MDU - LXMF_OVERHEAD + DESTINATION_LENGTH
# Descriptive strings regarding transport encryption
ENCRYPTION_DESCRIPTION_AES = "AES-128"
ENCRYPTION_DESCRIPTION_EC = "Curve25519"
ENCRYPTION_DESCRIPTION_UNENCRYPTED = "Unencrypted"
# Constants for QR/URI encoding LXMs
URI_SCHEMA = "lxm"
QR_ERROR_CORRECTION = "ERROR_CORRECT_L"
QR_MAX_STORAGE = 2953
PAPER_MDU = ((QR_MAX_STORAGE-(len(URI_SCHEMA)+len("://")))*6)//8
def __str__(self):
if self.hash != None:
return "<LXMessage "+RNS.hexrep(self.hash, delimit=False)+">"
else:
return "<LXMessage>"
def __init__(self, destination, source, content = "", title = "", fields = None, desired_method = None, destination_hash = None, source_hash = None, stamp_cost=None, include_ticket=False):
if isinstance(destination, RNS.Destination) or destination == None:
self.__destination = destination
if destination != None: self.destination_hash = destination.hash
else: self.destination_hash = destination_hash
else: raise ValueError("LXMessage initialised with invalid destination")
if isinstance(source, RNS.Destination) or source == None:
self.__source = source
if source != None: self.source_hash = source.hash
else: self.source_hash = source_hash
else: raise ValueError("LXMessage initialised with invalid source")
if title == None: title = ""
if type(title) == bytes: self.set_title_from_bytes(title)
else: self.set_title_from_string(title)
if type(content) == bytes: self.set_content_from_bytes(content)
else: self.set_content_from_string(content)
self.set_fields(fields)
self.payload = None
self.timestamp = None
self.signature = None
self.hash = None
self.transient_id = None
self.packed = None
self.auto_compress = True
self.state = LXMessage.GENERATING
self.method = LXMessage.UNKNOWN
self.progress = 0.0
self.rssi = None
self.snr = None
self.q = None
self.stamp = None
self.stamp_cost = stamp_cost
self.stamp_value = None
self.stamp_valid = False
self.stamp_checked = False
self.propagation_stamp = None
self.propagation_stamp_value = None
self.propagation_stamp_valid = False
self.propagation_target_cost = None
self.defer_stamp = True
self.defer_propagation_stamp = True
self.outbound_ticket = None
self.include_ticket = include_ticket
self.propagation_packed = None
self.paper_packed = None
self.incoming = False
self.source_blackholed = False
self.signature_validated = False
self.unverified_reason = None
self.ratchet_id = None
self.representation = LXMessage.UNKNOWN
self.desired_method = desired_method
self.delivery_attempts = 0
self.transport_encrypted = False
self.transport_encryption = None
self.ratchet_id = None
self.packet_representation = None
self.resource_representation = None
self.__delivery_destination = None
self.__delivery_callback = None
self.__pn_encrypted_data = None
self.__persist_lock = Lock()
self.failed_callback = None
self.deferred_stamp_generating = False
def set_title_from_string(self, title_string):
self.title = title_string.encode("utf-8")
def set_title_from_bytes(self, title_bytes):
self.title = title_bytes
def title_as_string(self):
return self.title.decode("utf-8")
def set_content_from_string(self, content_string):
self.content = content_string.encode("utf-8")
def set_content_from_bytes(self, content_bytes):
self.content = content_bytes
def content_as_string(self):
try:
return self.content.decode("utf-8")
except Exception as e:
RNS.log(f"{self} could not decode message content as string: {e}")
return None
def set_fields(self, fields):
if isinstance(fields, dict) or fields == None:
self.fields = fields or {}
else:
raise ValueError("LXMessage property \"fields\" can only be dict or None")
def get_fields(self):
return self.fields
@property
def destination(self):
return self.__destination
@destination.setter
def destination(self, destination):
self.set_destination(destination)
def get_destination(self):
return self.destination
def set_destination(self, destination):
if self.destination == None:
if isinstance(destination, RNS.Destination):
self.__destination = destination
else:
raise ValueError("Invalid destination set on LXMessage")
else:
raise ValueError("Cannot reassign destination on LXMessage")
@property
def source(self):
return self.__source
@source.setter
def source(self, source):
self.set_source(source)
def get_source(self):
return self.source
def set_source(self, source):
if self.source == None:
if isinstance(source, RNS.Destination):
self.__source = source
else:
raise ValueError("Invalid source set on LXMessage")
else:
raise ValueError("Cannot reassign source on LXMessage")
def set_delivery_destination(self, delivery_destination):
self.__delivery_destination = delivery_destination
def register_delivery_callback(self, callback):
self.__delivery_callback = callback
def register_failed_callback(self, callback):
self.failed_callback = callback
def validate_stamp(self, target_cost, tickets=None):
if tickets != None:
for ticket in tickets:
try:
if self.stamp == RNS.Identity.truncated_hash(ticket+self.message_id):
RNS.log(f"Stamp on {self} validated by inbound ticket", RNS.LOG_DEBUG) # TODO: Remove at some point
self.stamp_value = LXMessage.COST_TICKET
return True
except Exception as e:
RNS.log(f"Error while validating ticket: {e}", RNS.LOG_ERROR)
RNS.trace_exception(e)
if self.stamp == None:
return False
else:
workblock = LXStamper.stamp_workblock(self.message_id)
if LXStamper.stamp_valid(self.stamp, target_cost, workblock):
RNS.log(f"Stamp on {self} validated", RNS.LOG_DEBUG) # TODO: Remove at some point
self.stamp_value = LXStamper.stamp_value(workblock, self.stamp)
return True
else:
return False
def get_stamp(self, timeout=None):
# If an outbound ticket exists, use this for
# generating a valid stamp.
if self.outbound_ticket != None and type(self.outbound_ticket) == bytes and len(self.outbound_ticket) == LXMessage.TICKET_LENGTH:
generated_stamp = RNS.Identity.truncated_hash(self.outbound_ticket+self.message_id)
self.stamp_value = LXMessage.COST_TICKET
RNS.log(f"Generated stamp with outbound ticket {RNS.hexrep(self.outbound_ticket)} for {self}", RNS.LOG_DEBUG) # TODO: Remove at some point
return generated_stamp
# If no stamp cost is required, we can just
# return immediately.
elif self.stamp_cost == None:
self.stamp_value = None
return None
# If a stamp was already generated, return
# it immediately.
elif self.stamp != None:
return self.stamp
# Otherwise, we will need to generate a
# valid stamp according to the cost that
# the receiver has specified.
else:
generated_stamp, value = LXStamper.generate_stamp(self.message_id, self.stamp_cost)
if generated_stamp:
self.stamp_value = value
self.stamp_valid = True
return generated_stamp
else:
return None
def get_propagation_stamp(self, target_cost, timeout=None):
# If a stamp was already generated, return
# it immediately.
if self.propagation_stamp != None:
return self.propagation_stamp
# Otherwise, we will need to generate a
# valid stamp according to the cost that
# the propagation node has specified.
else:
self.propagation_target_cost = target_cost
if self.propagation_target_cost == None:
raise ValueError("Cannot generate propagation stamp without configured target propagation cost")
if not self.transient_id: self.pack()
generated_stamp, value = LXStamper.generate_stamp(self.transient_id, target_cost, expand_rounds=LXStamper.WORKBLOCK_EXPAND_ROUNDS_PN)
if generated_stamp:
self.propagation_stamp = generated_stamp
self.propagation_stamp_value = value
self.propagation_stamp_valid = True
return generated_stamp
else:
return None
def pack(self, payload_updated=False):
if not self.packed:
if self.timestamp == None: self.timestamp = time.time()
self.propagation_packed = None
self.paper_packed = None
self.payload = [self.timestamp, self.title, self.content, self.fields]
hashed_part = b""
hashed_part += self.__destination.hash
hashed_part += self.__source.hash
hashed_part += msgpack.packb(self.payload)
self.hash = RNS.Identity.full_hash(hashed_part)
self.message_id = self.hash
if not self.defer_stamp:
self.stamp = self.get_stamp()
if self.stamp != None: self.payload.append(self.stamp)
signed_part = b""
signed_part += hashed_part
signed_part += self.hash
self.signature = self.__source.sign(signed_part)
self.signature_validated = True
packed_payload = msgpack.packb(self.payload)
self.packed = b""
self.packed += self.__destination.hash
self.packed += self.__source.hash
self.packed += self.signature
self.packed += packed_payload
self.packed_size = len(self.packed)
content_size = len(packed_payload)-LXMessage.TIMESTAMP_SIZE-LXMessage.STRUCT_OVERHEAD
# If no desired delivery method has been defined,
# one will be chosen according to these rules:
if self.desired_method == None:
self.desired_method = LXMessage.DIRECT
# If opportunistic delivery was requested, check
# that message will fit within packet size limits
if self.desired_method == LXMessage.OPPORTUNISTIC:
if self.__destination.type == RNS.Destination.SINGLE:
if content_size > LXMessage.ENCRYPTED_PACKET_MAX_CONTENT:
RNS.log(f"Opportunistic delivery was requested for {self}, but content of length {content_size} exceeds packet size limit. Falling back to link-based delivery.", RNS.LOG_DEBUG)
self.desired_method = LXMessage.DIRECT
# Set delivery parameters according to delivery method
if self.desired_method == LXMessage.OPPORTUNISTIC:
if self.__destination.type == RNS.Destination.SINGLE:
single_packet_content_limit = LXMessage.ENCRYPTED_PACKET_MAX_CONTENT
elif self.__destination.type == RNS.Destination.PLAIN:
single_packet_content_limit = LXMessage.PLAIN_PACKET_MAX_CONTENT
if content_size > single_packet_content_limit:
raise TypeError(f"LXMessage desired opportunistic delivery method, but content of length {content_size} exceeds single-packet content limit of {single_packet_content_limit}.")
else:
self.method = LXMessage.OPPORTUNISTIC
self.representation = LXMessage.PACKET
self.__delivery_destination = self.__destination
elif self.desired_method == LXMessage.DIRECT:
single_packet_content_limit = LXMessage.LINK_PACKET_MAX_CONTENT
if content_size <= single_packet_content_limit:
self.method = self.desired_method
self.representation = LXMessage.PACKET
else:
self.method = self.desired_method
self.representation = LXMessage.RESOURCE
elif self.desired_method == LXMessage.PROPAGATED:
single_packet_content_limit = LXMessage.LINK_PACKET_MAX_CONTENT
if self.__pn_encrypted_data == None or payload_updated:
self.__pn_encrypted_data = self.__destination.encrypt(self.packed[LXMessage.DESTINATION_LENGTH:])
self.ratchet_id = self.__destination.latest_ratchet_id
lxmf_data = self.packed[:LXMessage.DESTINATION_LENGTH]+self.__pn_encrypted_data
self.transient_id = RNS.Identity.full_hash(lxmf_data)
if self.propagation_stamp != None: lxmf_data += self.propagation_stamp
self.propagation_packed = msgpack.packb([time.time(), [lxmf_data]])
content_size = len(self.propagation_packed)
if content_size <= single_packet_content_limit:
self.method = self.desired_method
self.representation = LXMessage.PACKET
else:
self.method = self.desired_method
self.representation = LXMessage.RESOURCE
elif self.desired_method == LXMessage.PAPER:
paper_content_limit = LXMessage.PAPER_MDU
encrypted_data = self.__destination.encrypt(self.packed[LXMessage.DESTINATION_LENGTH:])
self.ratchet_id = self.__destination.latest_ratchet_id
self.paper_packed = self.packed[:LXMessage.DESTINATION_LENGTH]+encrypted_data
content_size = len(self.paper_packed)
if content_size <= paper_content_limit:
self.method = self.desired_method
self.representation = LXMessage.PAPER
else:
raise TypeError("LXMessage desired paper delivery method, but content exceeds paper message maximum size.")
else:
raise ValueError("Attempt to re-pack LXMessage "+str(self)+" that was already packed")
def send(self):
self.determine_transport_encryption()
self.determine_compression_support()
if self.method == LXMessage.OPPORTUNISTIC:
lxm_packet = self.__as_packet()
lxm_packet.send().set_delivery_callback(self.__mark_delivered)
self.progress = 0.50
self.ratchet_id = lxm_packet.ratchet_id
self.state = LXMessage.SENT
elif self.method == LXMessage.DIRECT:
self.state = LXMessage.SENDING
if self.representation == LXMessage.PACKET:
lxm_packet = self.__as_packet()
receipt = lxm_packet.send()
self.ratchet_id = self.__delivery_destination.link_id
if receipt:
receipt.set_delivery_callback(self.__mark_delivered)
receipt.set_timeout_callback(self.__link_packet_timed_out)
self.progress = 0.50
else:
if self.__delivery_destination:
self.__delivery_destination.teardown()
elif self.representation == LXMessage.RESOURCE:
self.resource_representation = self.__as_resource()
self.ratchet_id = self.__delivery_destination.link_id
self.progress = 0.10
elif self.method == LXMessage.PROPAGATED:
self.state = LXMessage.SENDING
if self.representation == LXMessage.PACKET:
receipt = self.__as_packet().send()
if receipt:
receipt.set_delivery_callback(self.__mark_propagated)
receipt.set_timeout_callback(self.__link_packet_timed_out)
self.progress = 0.50
else:
self.__delivery_destination.teardown()
elif self.representation == LXMessage.RESOURCE:
self.resource_representation = self.__as_resource()
self.progress = 0.10
def determine_compression_support(self):
app_data = RNS.Identity.recall_app_data(self.destination_hash)
if app_data: self.auto_compress = compression_support_from_app_data(app_data)
else: self.auto_compress = True
###### TODO: Remove debug logging
if app_data: RNS.log(f"Set compression support from app data to: {self.auto_compress}", RNS.LOG_DEBUG)
else: RNS.log(f"Defaulting compression support to {self.auto_compress}", RNS.LOG_DEBUG)
######
def determine_transport_encryption(self):
# TODO: These descriptions are old and outdated.
# Update the transport encryption descriptions to
# account for ratchets and other changes.
if self.method == LXMessage.OPPORTUNISTIC:
if self.__destination.type == RNS.Destination.SINGLE:
self.transport_encrypted = True
self.transport_encryption = LXMessage.ENCRYPTION_DESCRIPTION_EC
elif self.__destination.type == RNS.Destination.GROUP:
self.transport_encrypted = True
self.transport_encryption = LXMessage.ENCRYPTION_DESCRIPTION_AES
else:
self.transport_encrypted = False
self.transport_encryption = LXMessage.ENCRYPTION_DESCRIPTION_UNENCRYPTED
elif self.method == LXMessage.DIRECT:
self.transport_encrypted = True
self.transport_encryption = LXMessage.ENCRYPTION_DESCRIPTION_EC
elif self.method == LXMessage.PROPAGATED:
if self.__destination.type == RNS.Destination.SINGLE:
self.transport_encrypted = True
self.transport_encryption = LXMessage.ENCRYPTION_DESCRIPTION_EC
elif self.__destination.type == RNS.Destination.GROUP:
self.transport_encrypted = True
self.transport_encryption = LXMessage.ENCRYPTION_DESCRIPTION_AES
else:
self.transport_encrypted = False
self.transport_encryption = LXMessage.ENCRYPTION_DESCRIPTION_UNENCRYPTED
elif self.method == LXMessage.PAPER:
if self.__destination.type == RNS.Destination.SINGLE:
self.transport_encrypted = True
self.transport_encryption = LXMessage.ENCRYPTION_DESCRIPTION_EC
elif self.__destination.type == RNS.Destination.GROUP:
self.transport_encrypted = True
self.transport_encryption = LXMessage.ENCRYPTION_DESCRIPTION_AES
else:
self.transport_encrypted = False
self.transport_encryption = LXMessage.ENCRYPTION_DESCRIPTION_UNENCRYPTED
else:
self.transport_encrypted = False
self.transport_encryption = LXMessage.ENCRYPTION_DESCRIPTION_UNENCRYPTED
def __mark_delivered(self, receipt = None):
RNS.log("Received delivery notification for "+str(self), RNS.LOG_DEBUG)
self.state = LXMessage.DELIVERED
self.progress = 1.0
if self.__delivery_callback != None and callable(self.__delivery_callback):
try:
self.__delivery_callback(self)
except Exception as e:
RNS.log("An error occurred in the external delivery callback for "+str(self), RNS.LOG_ERROR)
RNS.trace_exception(e)
def __mark_propagated(self, receipt = None):
RNS.log("Received propagation success notification for "+str(self), RNS.LOG_DEBUG)
self.state = LXMessage.SENT
self.progress = 1.0
if self.__delivery_callback != None and callable(self.__delivery_callback):
try:
self.__delivery_callback(self)
except Exception as e:
RNS.log("An error occurred in the external delivery callback for "+str(self), RNS.LOG_ERROR)
RNS.trace_exception(e)
def __mark_paper_generated(self, receipt = None):
RNS.log("Paper message generation succeeded for "+str(self), RNS.LOG_DEBUG)
self.state = LXMessage.PAPER
self.progress = 1.0
if self.__delivery_callback != None and callable(self.__delivery_callback):
try:
self.__delivery_callback(self)
except Exception as e:
RNS.log("An error occurred in the external delivery callback for "+str(self), RNS.LOG_ERROR)
RNS.trace_exception(e)
def __resource_concluded(self, resource):
if resource.status == RNS.Resource.COMPLETE:
self.__mark_delivered()
else:
if resource.status == RNS.Resource.REJECTED:
self.state = LXMessage.REJECTED
elif self.state != LXMessage.CANCELLED:
resource.link.teardown()
self.state = LXMessage.OUTBOUND
def __propagation_resource_concluded(self, resource):
if resource.status == RNS.Resource.COMPLETE:
self.__mark_propagated()
else:
if self.state != LXMessage.CANCELLED:
resource.link.teardown()
self.state = LXMessage.OUTBOUND
def __link_packet_timed_out(self, packet_receipt):
if self.state != LXMessage.CANCELLED:
if packet_receipt:
packet_receipt.destination.teardown()
self.state = LXMessage.OUTBOUND
def __update_transfer_progress(self, resource):
self.progress = 0.10 + (resource.get_progress()*0.90)
def __as_packet(self):
if not self.packed:
self.pack()
if not self.__delivery_destination:
raise ValueError("Can't synthesize packet for LXMF message before delivery destination is known")
if self.method == LXMessage.OPPORTUNISTIC:
return RNS.Packet(self.__delivery_destination, self.packed[LXMessage.DESTINATION_LENGTH:])
elif self.method == LXMessage.DIRECT:
return RNS.Packet(self.__delivery_destination, self.packed)
elif self.method == LXMessage.PROPAGATED:
return RNS.Packet(self.__delivery_destination, self.propagation_packed)
def __as_resource(self):
if not self.packed:
self.pack()
if not self.__delivery_destination:
raise ValueError("Can't synthesize resource for LXMF message before delivery destination is known")
if not self.__delivery_destination.type == RNS.Destination.LINK:
raise TypeError("Tried to synthesize resource for LXMF message on a delivery destination that was not a link")
if not self.__delivery_destination.status == RNS.Link.ACTIVE:
raise ConnectionError("Tried to synthesize resource for LXMF message on a link that was not active")
if self.method == LXMessage.DIRECT:
return RNS.Resource(self.packed, self.__delivery_destination, callback = self.__resource_concluded, progress_callback = self.__update_transfer_progress, auto_compress=self.auto_compress)
elif self.method == LXMessage.PROPAGATED:
return RNS.Resource(self.propagation_packed, self.__delivery_destination, callback = self.__propagation_resource_concluded, progress_callback = self.__update_transfer_progress)
else:
return None
def packed_container(self):
if not self.packed:
self.pack()
container = {
"state": self.state,
"lxmf_bytes": self.packed,
"transport_encrypted": self.transport_encrypted,
"transport_encryption": self.transport_encryption,
"method": self.method
}
return msgpack.packb(container)
def write_to_directory(self, directory_path):
file_name = RNS.hexrep(self.hash, delimit=False)
file_path = directory_path+"/"+file_name
tmp_path = file_path+".tmp."+str(os.getpid() or time.time())+"."+RNS.hexrep(os.urandom(8), delimit=False)
with self.__persist_lock:
try:
with open(tmp_path, "wb") as file:
file.write(self.packed_container())
file.flush()
try: os.fsync(file.fileno())
except OSError as e: RNS.log(f"Error while waiting for persist fsync for {self}: {e}", RNS.LOG_WARNING)
os.replace(tmp_path, file_path)
return file_path
except Exception as e:
try:
if os.path.exists(tmp_path): os.unlink(tmp_path)
except Exception as e: RNS.log(f"Error while cleaning temporary file {tmp_path} for {self}: {e}", RNS.LOG_ERROR)
RNS.log(f"Error while writing LXMF message to file \"{file_path}\". The contained exception was: {e}", RNS.LOG_ERROR)
return None
def as_uri(self, finalise=True):
if not self.packed:
self.pack()
if self.desired_method == LXMessage.PAPER and self.paper_packed != None:
# Encode packed LXM with URL-safe base64 and remove padding
encoded_bytes = base64.urlsafe_b64encode(self.paper_packed)
# Add protocol specifier and return
lxm_uri = LXMessage.URI_SCHEMA+"://"+encoded_bytes.decode("utf-8").replace("=","")
if finalise:
self.determine_transport_encryption()
self.__mark_paper_generated()
return lxm_uri
else:
raise TypeError("Attempt to represent LXM with non-paper delivery method as URI")
def as_qr(self):
if not self.packed:
self.pack()
if self.desired_method == LXMessage.PAPER and self.paper_packed != None:
import importlib
if importlib.util.find_spec('qrcode') != None:
import qrcode
qr = qrcode.make(
error_correction = qrcode.constants.__dict__[LXMessage.QR_ERROR_CORRECTION],
border = 1,
data = self.as_uri(finalise=False),
)
self.determine_transport_encryption()
self.__mark_paper_generated()
return qr
else:
RNS.log("Generating QR-code representanions of LXMs requires the \"qrcode\" module to be installed.", RNS.LOG_CRITICAL)
RNS.log("You can install it with the command: python3 -m pip install qrcode", RNS.LOG_CRITICAL)
return None
else:
raise TypeError("Attempt to represent LXM with non-paper delivery method as QR-code")
@staticmethod
def unpack_from_bytes(lxmf_bytes, original_method = None):
destination_hash = lxmf_bytes[:LXMessage.DESTINATION_LENGTH]
source_hash = lxmf_bytes[LXMessage.DESTINATION_LENGTH:2*LXMessage.DESTINATION_LENGTH]
signature = lxmf_bytes[2*LXMessage.DESTINATION_LENGTH:2*LXMessage.DESTINATION_LENGTH+LXMessage.SIGNATURE_LENGTH]
packed_payload = lxmf_bytes[2*LXMessage.DESTINATION_LENGTH+LXMessage.SIGNATURE_LENGTH:]
unpacked_payload = msgpack.unpackb(packed_payload)
# Extract stamp from payload if included
if len(unpacked_payload) > 4:
stamp = unpacked_payload[4]
unpacked_payload = unpacked_payload[:4]
packed_payload = msgpack.packb(unpacked_payload)
else:
stamp = None
hashed_part = b"" + destination_hash + source_hash + packed_payload
message_hash = RNS.Identity.full_hash(hashed_part)
signed_part = b"" + hashed_part + message_hash
timestamp = unpacked_payload[0]
title_bytes = unpacked_payload[1]
content_bytes = unpacked_payload[2]
fields = unpacked_payload[3]
destination_identity = RNS.Identity.recall(destination_hash, _no_use=True)
if destination_identity != None:
destination = RNS.Destination(destination_identity, RNS.Destination.OUT, RNS.Destination.SINGLE, APP_NAME, "delivery")
else:
destination = None
source_identity = RNS.Identity.recall(source_hash, _no_use=True)
if source_identity != None:
source = RNS.Destination(source_identity, RNS.Destination.OUT, RNS.Destination.SINGLE, APP_NAME, "delivery")
else:
source = None
message = LXMessage(
destination = destination,
source = source,
content = "",
title = "",
fields = fields,
destination_hash = destination_hash,
source_hash = source_hash,
desired_method = original_method)
message.hash = message_hash
message.message_id = message.hash
message.signature = signature
message.stamp = stamp
message.incoming = True
message.timestamp = timestamp
message.packed = lxmf_bytes
message.packed_size = len(lxmf_bytes)
message.set_title_from_bytes(title_bytes)
message.set_content_from_bytes(content_bytes)
try:
if source_identity != None: message.source_blackholed = RNS.Reticulum.get_instance().is_blackholed(source_identity)
except Exception as e: RNS.log(f"Could not determine message source blackhole status: {e}", RNS.LOG_WARNING)
try:
if source:
if source.identity.validate(signature, signed_part):
message.signature_validated = True
else:
message.signature_validated = False
message.unverified_reason = LXMessage.SIGNATURE_INVALID
else:
signature_validated = False
message.unverified_reason = LXMessage.SOURCE_UNKNOWN
RNS.log("Unpacked LXMF message signature could not be validated, since source identity is unknown", RNS.LOG_DEBUG)
except Exception as e:
message.signature_validated = False
RNS.log("Error while validating LXMF message signature. The contained exception was: "+str(e), RNS.LOG_ERROR)
return message
@staticmethod
def unpack_from_file(lxmf_file_handle):
try:
container = msgpack.unpackb(lxmf_file_handle.read())
lxm = LXMessage.unpack_from_bytes(container["lxmf_bytes"])
if "state" in container:
lxm.state = container["state"]
if "transport_encrypted" in container:
lxm.transport_encrypted = container["transport_encrypted"]
if "transport_encryption" in container:
lxm.transport_encryption = container["transport_encryption"]
if "method" in container:
lxm.method = container["method"]
return lxm
except Exception as e:
RNS.log("Could not unpack LXMessage from file. The contained exception was: "+str(e), RNS.LOG_ERROR)
return None

523
LXMF/LXStamper.py Normal file
View file

@ -0,0 +1,523 @@
import RNS
import RNS.vendor.umsgpack as msgpack
import os
import sys
import time
import math
import itertools
import contextlib
import multiprocessing
WORKBLOCK_EXPAND_ROUNDS = 3000
WORKBLOCK_EXPAND_ROUNDS_PN = 1000
WORKBLOCK_EXPAND_ROUNDS_PEERING = 25
STAMP_SIZE = RNS.Identity.HASHLENGTH//8
PN_VALIDATION_POOL_MIN_SIZE = 256
USE_WORKER_MANAGER = False
active_jobs = {}
if sys.version_info[0] >= 3 and sys.version_info[1] >= 14:
USE_WORKER_MANAGER = True
@contextlib.contextmanager
def worker_context(ctx, stamp_cost, workblock, message_id):
stop_event = ctx.Event()
result_queue = ctx.Queue(1)
rounds_queue = ctx.Queue()
active_jobs[message_id] = [stop_event, result_queue]
job_procs = []
try: yield stop_event, result_queue, rounds_queue, job_procs
finally:
if message_id in active_jobs: del active_jobs[message_id]
stop_event.set()
for p in job_procs:
if p.is_alive(): p.terminate()
for p in job_procs: p.join(timeout=0.5)
result_queue.close()
result_queue.join_thread()
rounds_queue.close()
rounds_queue.join_thread()
stop_event.clear()
def stamp_workblock(material, expand_rounds=WORKBLOCK_EXPAND_ROUNDS):
wb_st = time.time()
workblock = b""
for n in range(expand_rounds):
workblock += RNS.Cryptography.hkdf(length=256,
derive_from=material,
salt=RNS.Identity.full_hash(material+msgpack.packb(n)),
context=None)
wb_time = time.time() - wb_st
# RNS.log(f"Stamp workblock size {RNS.prettysize(len(workblock))}, generated in {round(wb_time*1000,2)}ms", RNS.LOG_DEBUG)
return workblock
def stamp_value(workblock, stamp):
value = 0
bits = 256
material = RNS.Identity.full_hash(workblock+stamp)
i = int.from_bytes(material, byteorder="big")
while ((i & (1 << (bits - 1))) == 0):
i = (i << 1)
value += 1
return value
def stamp_valid(stamp, target_cost, workblock):
target = 0b1 << 256-target_cost
result = RNS.Identity.full_hash(workblock+stamp)
if int.from_bytes(result, byteorder="big") > target: return False
else: return True
def validate_peering_key(peering_id, peering_key, target_cost):
workblock = stamp_workblock(peering_id, expand_rounds=WORKBLOCK_EXPAND_ROUNDS_PEERING)
if not stamp_valid(peering_key, target_cost, workblock): return False
else: return True
def validate_pn_stamp(transient_data, target_cost):
from .LXMessage import LXMessage
if len(transient_data) <= LXMessage.LXMF_OVERHEAD+STAMP_SIZE: return None, None, None, None
else:
lxm_data = transient_data[:-STAMP_SIZE]
stamp = transient_data[-STAMP_SIZE:]
transient_id = RNS.Identity.full_hash(lxm_data)
workblock = stamp_workblock(transient_id, expand_rounds=WORKBLOCK_EXPAND_ROUNDS_PN)
if not stamp_valid(stamp, target_cost, workblock): return None, None, None, None
else:
value = stamp_value(workblock, stamp)
return transient_id, lxm_data, value, stamp
def validate_pn_stamps_job_simple(transient_list, target_cost):
validated_messages = []
for transient_data in transient_list:
transient_id, lxm_data, value, stamp_data = validate_pn_stamp(transient_data, target_cost)
if transient_id: validated_messages.append([transient_id, lxm_data, value, stamp_data])
return validated_messages
def validate_pn_stamps_job_multip(transient_list, target_cost):
cores = multiprocessing.cpu_count()
pool_count = min(cores, math.ceil(len(transient_list) / PN_VALIDATION_POOL_MIN_SIZE))
RNS.log(f"Validating {len(transient_list)} stamps using {pool_count} processes...", RNS.LOG_VERBOSE)
with multiprocessing.get_context("spawn").Pool(pool_count) as p:
validated_entries = p.starmap(validate_pn_stamp, zip(transient_list, itertools.repeat(target_cost)))
RNS.log(f"Validation pool completed for {len(transient_list)} stamps", RNS.LOG_VERBOSE)
return [e for e in validated_entries if e[0] != None]
def validate_pn_stamps(transient_list, target_cost):
non_mp_platform = RNS.vendor.platformutils.is_android()
if len(transient_list) <= PN_VALIDATION_POOL_MIN_SIZE or non_mp_platform: return validate_pn_stamps_job_simple(transient_list, target_cost)
else: return validate_pn_stamps_job_multip(transient_list, target_cost)
def generate_stamp(message_id, stamp_cost, expand_rounds=WORKBLOCK_EXPAND_ROUNDS):
RNS.log(f"Generating stamp with cost {stamp_cost} for {RNS.prettyhexrep(message_id)}...", RNS.LOG_DEBUG)
workblock = stamp_workblock(message_id, expand_rounds=expand_rounds)
start_time = time.time()
stamp = None
rounds = 0
value = 0
if RNS.vendor.platformutils.is_windows() or RNS.vendor.platformutils.is_darwin(): stamp, rounds = job_simple(stamp_cost, workblock, message_id)
elif RNS.vendor.platformutils.is_android(): stamp, rounds = job_android(stamp_cost, workblock, message_id)
else:
if USE_WORKER_MANAGER: stamp, rounds = job_linux_managed(stamp_cost, workblock, message_id)
else: stamp, rounds = job_linux(stamp_cost, workblock, message_id)
duration = time.time() - start_time
speed = rounds/duration
if stamp != None: value = stamp_value(workblock, stamp)
RNS.log(f"Stamp with value {value} generated in {RNS.prettytime(duration)}, {rounds} rounds, {int(speed)} rounds per second", RNS.LOG_DEBUG)
return stamp, value
def cancel_work(message_id):
if RNS.vendor.platformutils.is_windows() or RNS.vendor.platformutils.is_darwin():
try:
if message_id in active_jobs:
active_jobs[message_id] = True
except Exception as e:
RNS.log("Error while terminating stamp generation workers: {e}", RNS.LOG_ERROR)
RNS.trace_exception(e)
elif RNS.vendor.platformutils.is_android():
try:
if message_id in active_jobs:
active_jobs[message_id] = True
except Exception as e:
RNS.log("Error while terminating stamp generation workers: {e}", RNS.LOG_ERROR)
RNS.trace_exception(e)
else:
try:
if message_id in active_jobs:
stop_event = active_jobs[message_id][0]
result_queue = active_jobs[message_id][1]
stop_event.set()
result_queue.put(None)
active_jobs.pop(message_id)
except Exception as e:
RNS.log("Error while terminating stamp generation workers: {e}", RNS.LOG_ERROR)
RNS.trace_exception(e)
def job_simple(stamp_cost, workblock, message_id):
# A simple, single-process stamp generator.
# should work on any platform, and is used
# as a fall-back, in case of limited multi-
# processing and/or acceleration support.
platform = RNS.vendor.platformutils.get_platform()
RNS.log(f"Running stamp generation on {platform}, work limited to single CPU core. This will be slower than ideal.", RNS.LOG_WARNING)
rounds = 0
pstamp = os.urandom(256//8)
st = time.time()
active_jobs[message_id] = False;
def sv(s, c, w):
target = 0b1<<256-c; m = w+s
result = RNS.Identity.full_hash(m)
if int.from_bytes(result, byteorder="big") > target: return False
else: return True
while not sv(pstamp, stamp_cost, workblock) and not active_jobs[message_id]:
pstamp = os.urandom(256//8); rounds += 1
if rounds % 2500 == 0:
speed = rounds / (time.time()-st)
RNS.log(f"Stamp generation running. {rounds} rounds completed so far, {int(speed)} rounds per second", RNS.LOG_DEBUG)
if active_jobs[message_id] == True:
pstamp = None
active_jobs.pop(message_id)
return pstamp, rounds
def job_linux_managed(stamp_cost, workblock, message_id):
ctx = multiprocessing.get_context("fork")
cores = multiprocessing.cpu_count()
jobs = cores if cores <= 12 else int(cores/2)
allow_kill = True
stamp = None
total_rounds = 0
with worker_context(ctx, stamp_cost, workblock, message_id) as (stop_event, result_queue, rounds_queue, job_procs):
RNS.log(f"Starting {jobs} stamp generation workers", RNS.LOG_DEBUG)
def job(stop_event, sc, wb, worker_id):
terminated = False
rounds = 0
pstamp = os.urandom(256//8)
def sv(s, c, w):
target = 0b1<<256-c; m = w+s
result = RNS.Identity.full_hash(m)
if int.from_bytes(result, byteorder="big") > target: return False
else: return True
while not stop_event.is_set() and not sv(pstamp, sc, wb):
pstamp = os.urandom(256//8); rounds += 1
if not stop_event.is_set():
stop_event.set()
try: result_queue.put_nowait(pstamp)
except: pass
try: rounds_queue.put_nowait(rounds)
except: pass
for jpn in range(jobs):
p = ctx.Process(target=job, args=(stop_event, stamp_cost, workblock, jpn), daemon=True)
job_procs.append(p)
p.start()
try: stamp = result_queue.get(timeout=None)
except Exception as e:
RNS.log(f"Failed to get result from workers: {e}", RNS.LOG_ERROR)
stamp = None
# Collect any potential spurious
# results from worker queue.
try:
while True: result_queue.get_nowait()
except: pass
for j in range(jobs):
nrounds = 0
try:
nrounds = rounds_queue.get(timeout=2)
except Exception as e:
RNS.log(f"Failed to get round stats part {j}: {e}", RNS.LOG_ERROR)
total_rounds += nrounds
all_exited = False
exit_timeout = time.time() + 5
while time.time() < exit_timeout:
if not any(p.is_alive() for p in job_procs):
all_exited = True
break
time.sleep(0.1)
if not all_exited:
RNS.log("Stamp generation IPC timeout, possible worker deadlock. Terminating remaining processes.", RNS.LOG_ERROR)
if allow_kill:
for j in range(jobs):
process = job_procs[j]
process.kill()
else:
return None
else:
for j in range(jobs):
process = job_procs[j]
process.join()
# RNS.log(f"Joined {j} / {process}", RNS.LOG_DEBUG) # TODO: Remove
return stamp, total_rounds
def job_linux(stamp_cost, workblock, message_id):
allow_kill = True
stamp = None
total_rounds = 0
cores = multiprocessing.cpu_count()
jobs = cores if cores <= 12 else int(cores/2)
stop_event = multiprocessing.Event()
result_queue = multiprocessing.Queue(1)
rounds_queue = multiprocessing.Queue()
def job(stop_event, pn, sc, wb):
terminated = False
rounds = 0
pstamp = os.urandom(256//8)
def sv(s, c, w):
target = 0b1<<256-c; m = w+s
result = RNS.Identity.full_hash(m)
if int.from_bytes(result, byteorder="big") > target:
return False
else:
return True
while not stop_event.is_set() and not sv(pstamp, sc, wb):
pstamp = os.urandom(256//8); rounds += 1
if not stop_event.is_set():
stop_event.set()
result_queue.put(pstamp)
rounds_queue.put(rounds)
job_procs = []
RNS.log(f"Starting {jobs} stamp generation workers", RNS.LOG_DEBUG)
for jpn in range(jobs):
process = multiprocessing.get_context("fork").Process(target=job, kwargs={"stop_event": stop_event, "pn": jpn, "sc": stamp_cost, "wb": workblock}, daemon=True)
job_procs.append(process)
process.start()
active_jobs[message_id] = [stop_event, result_queue]
stamp = result_queue.get()
# Collect any potential spurious
# results from worker queue.
try:
while True: result_queue.get_nowait()
except: pass
for j in range(jobs):
nrounds = 0
try:
nrounds = rounds_queue.get(timeout=2)
except Exception as e:
RNS.log(f"Failed to get round stats part {j}: {e}", RNS.LOG_ERROR)
total_rounds += nrounds
all_exited = False
exit_timeout = time.time() + 5
while time.time() < exit_timeout:
if not any(p.is_alive() for p in job_procs):
all_exited = True
break
time.sleep(0.1)
if not all_exited:
RNS.log("Stamp generation IPC timeout, possible worker deadlock. Terminating remaining processes.", RNS.LOG_ERROR)
if allow_kill:
for j in range(jobs):
process = job_procs[j]
process.kill()
else:
return None
else:
for j in range(jobs):
process = job_procs[j]
process.join()
# RNS.log(f"Joined {j} / {process}", RNS.LOG_DEBUG) # TODO: Remove
return stamp, total_rounds
def job_android(stamp_cost, workblock, message_id):
# Semaphore support is flaky to non-existent on
# Android, so we need to manually dispatch and
# manage workloads here, while periodically
# checking in on the progress.
stamp = None
start_time = time.time()
total_rounds = 0
rounds_per_worker = 1000
use_nacl = False
try:
import nacl.encoding
import nacl.hash
use_nacl = True
except:
pass
if use_nacl:
def full_hash(m):
return nacl.hash.sha256(m, encoder=nacl.encoding.RawEncoder)
else:
def full_hash(m):
return RNS.Identity.full_hash(m)
def sv(s, c, w):
target = 0b1<<256-c
m = w+s
result = full_hash(m)
if int.from_bytes(result, byteorder="big") > target:
return False
else:
return True
wm = multiprocessing.Manager()
jobs = multiprocessing.cpu_count()
def job(procnum=None, results_dict=None, wb=None, sc=None, jr=None):
# RNS.log(f"Worker {procnum} starting for {jr} rounds...") # TODO: Remove
try:
rounds = 0
found_stamp = None
while True:
pstamp = os.urandom(256//8)
rounds += 1
if sv(pstamp, sc, wb):
found_stamp = pstamp
break
if rounds >= jr:
# RNS.log(f"Worker {procnum} found no result in {rounds} rounds") # TODO: Remove
break
results_dict[procnum] = [found_stamp, rounds]
except Exception as e:
RNS.log(f"Stamp generation worker error: {e}", RNS.LOG_ERROR)
RNS.trace_exception(e)
active_jobs[message_id] = False;
RNS.log(f"Dispatching {jobs} workers for stamp generation...", RNS.LOG_DEBUG) # TODO: Remove
results_dict = wm.dict()
while stamp == None and active_jobs[message_id] == False:
job_procs = []
try:
for pnum in range(jobs):
pargs = {"procnum":pnum, "results_dict": results_dict, "wb": workblock, "sc":stamp_cost, "jr":rounds_per_worker}
process = multiprocessing.Process(target=job, kwargs=pargs)
job_procs.append(process)
process.start()
for process in job_procs:
process.join()
for j in results_dict:
r = results_dict[j]
total_rounds += r[1]
if r[0] != None:
stamp = r[0]
if stamp == None:
elapsed = time.time() - start_time
speed = total_rounds/elapsed
RNS.log(f"Stamp generation running. {total_rounds} rounds completed so far, {int(speed)} rounds per second", RNS.LOG_DEBUG)
except Exception as e:
RNS.log(f"Stamp generation job error: {e}")
RNS.trace_exception(e)
active_jobs.pop(message_id)
return stamp, total_rounds
# def stamp_value_linear(workblock, stamp):
# value = 0
# bits = 256
# material = RNS.Identity.full_hash(workblock+stamp)
# s = int.from_bytes(material, byteorder="big")
# return s.bit_count()
if __name__ == "__main__":
import sys
if len(sys.argv) < 2:
RNS.log("No cost argument provided", RNS.LOG_ERROR)
exit(1)
else:
try: cost = int(sys.argv[1])
except Exception as e:
RNS.log(f"Invalid cost argument provided: {e}", RNS.LOG_ERROR)
exit(1)
rounds = 1
if len(sys.argv) > 2:
try: rounds = int(sys.argv[2])
except Exception as e:
RNS.log(f"Invalid rounds argument provided: {e}", RNS.LOG_ERROR)
exit(1)
RNS.loglevel = RNS.LOG_DEBUG
for i in range(rounds):
RNS.log("", RNS.LOG_DEBUG)
RNS.log(f"Round {i+1} of {rounds}", RNS.LOG_DEBUG)
RNS.log("Testing LXMF stamp generation", RNS.LOG_DEBUG)
message_id = os.urandom(32)
generate_stamp(message_id, cost)
RNS.log("", RNS.LOG_DEBUG)
RNS.log("Testing propagation stamp generation", RNS.LOG_DEBUG)
message_id = os.urandom(32)
generate_stamp(message_id, cost, expand_rounds=WORKBLOCK_EXPAND_ROUNDS_PN)
RNS.log("", RNS.LOG_DEBUG)
RNS.log("Testing peering key generation", RNS.LOG_DEBUG)
message_id = os.urandom(32)
generate_stamp(message_id, cost, expand_rounds=WORKBLOCK_EXPAND_ROUNDS_PEERING)
# transient_list = []
# st = time.time(); count = 10000
# for i in range(count): transient_list.append(os.urandom(256))
# validate_pn_stamps(transient_list, 5)
# dt = time.time()-st; mps = count/dt
# RNS.log(f"Validated {count} PN stamps in {RNS.prettytime(dt)}, {round(mps,1)} m/s", RNS.LOG_DEBUG)

1126
LXMF/Utilities/lxmd.py Normal file
View file

File diff suppressed because it is too large Load diff

10
LXMF/__init__.py Normal file
View file

@ -0,0 +1,10 @@
import os
import glob
from .LXMessage import LXMessage
from .LXMRouter import LXMRouter
from .LXMF import *
from ._version import __version__
modules = glob.glob(os.path.dirname(__file__)+"/*.py")
__all__ = [ os.path.basename(f)[:-3] for f in modules if not f.endswith('__init__.py')]

1
LXMF/_version.py Normal file
View file

@ -0,0 +1 @@
__version__ = "1.0.1"

33
MIRROR.md Normal file
View file

@ -0,0 +1,33 @@
This repository is a public mirror. All potential future development is happening elsewhere.
I am stepping back from all public-facing interaction with this project. Reticulum has always been primarily my work, and continuing in the current public, internet-facing model is no longer sustainable.
The software remains available for use as-is. Occasional updates may appear at unpredictable intervals, but there will be no support, no responses to issues, no discussions, and no community management in this or any other public venue. If it doesn't work for you, it doesn't work. That is the entire extent of available troubleshooting assistance I can offer you.
If you've followed this project for a while, you already know what this means. You know who designed, wrote and tested this, and you know how many years of my life it took. You'll also know about both my particular challenges and strengths, and how I believe anything worth building needs to be built and maintained with our own hands.
Seven months ago, I said I needed to step back, that I was exhausted, and that I needed to recover. I believed a public resolve would be enough to effectuate that, but while striving to get just a few more useful features and protocols out, the unproductive requests and demands also ramped up, and I got pulled back into the same patterns and draining interactions that I'd explicitly said I couldn't sustain anymore.
So here's what you might have already guessed: I'm done playing the game by rules I can't win at.
Everything you need is right here, and by any sensible measure, it's done. Anyone who wants to invest the time, skill and persistence can build on it, or completely re-imagine it with different priorities. That was always the point.
The people who actually contributed - you know who you are, and you know I mean it when I say: Thank you. All of you who've used this to build something real - that was the goal, and you did it without needing me to hold your hand.
The rest of you: You have what you need. Use it or don't. I am not going to be the person who explains it to you anymore.
This is not a temporary break. It's not "see you after some rest", but a recognition that the current model is fundamentally incompatible with my life, my health, and my reality.
If you want to support continued work, you can do so at the donation links listed in this repository. But please understand, that this is not purchasing support or guaranteeing updates. It is support for work that happens on my timeline, according to my capacity, which at the moment is not what it was.
If you want Reticulum to continue evolving, you have the power to make that happen. The protocol is public domain. The code is open source. Everything you need is right here. I've provided the tools, but building what comes next is not my responsibility anymore. It's yours.
To the small group of people who has actually been here, and understood what this work was and what it cost - you already know where to find me if it actually matters.
To everyone else: This is where we part ways. No hard feelings. It's just time.
---
असतो मा सद्गमय
तमसो मा ज्योतिर्गमय
मृत्योर्मा अमृतं गमय

35
Makefile Normal file
View file

@ -0,0 +1,35 @@
all: release
clean:
@echo Cleaning...
-rm -r ./build
-rm -r ./dist
remove_symlinks:
@echo Removing symlinks for build...
-rm ./RNS
-rm ./LXMF/Utilities/LXMF
create_symlinks:
@echo Creating symlinks...
-ln -s ../Reticulum/RNS ./
-ln -s ../../LXMF ./LXMF/Utilities/LXMF
build_wheel:
python3 setup.py bdist_wheel
build_sdist:
python3 setup.py sdist
build_spkg: remove_symlinks build_sdist create_symlinks
release: remove_symlinks build_wheel build_spkg create_symlinks
upload:
@echo Ready to publish release over Reticulum
@read VOID
rngit release rns://7649a50d84610232d1416b41d2896aff/reticulum/lxmf create $$(python setup.py --getversion):dist --name lxmf
upload-pip:
@echo Uploading to PyPi...
twine upload dist/*.whl dist/*.tar.gz

13
PLANS.md Normal file
View file

@ -0,0 +1,13 @@
# LXMF Specification Investigation Plan
Each investigation must include:
1. Question being answered.
2. Upstream files inspected.
3. Exact version or commit.
4. Relevant classes/functions/constants.
5. Observed behavior.
6. Whether behavior is normative, inferred, or unresolved.
7. Test vectors needed.
8. SPEC.md sections affected.
9. Open questions.

219
README.md
View file

@ -1,3 +1,218 @@
# LMXF-specification # Lightweight Extensible Message Format
Based on Mark Qvist's LMXF, this project attempts to distill a specification from the Python code which is an implementation of the LMXF "protocol" LXMF is a simple and flexible messaging format and delivery protocol that allows a wide variety of implementations, while using as little bandwidth as possible. It is built on top of [Reticulum](a8d24177d946de4f1f0a0fe1af9a1338:/page/repo.mu`g=reticulum|r=reticulum) and offers zero-conf message routing, end-to-end encryption and Forward Secrecy, and can be transported over any kind of medium that Reticulum supports.
LXMF is efficient enough that it can deliver messages over extremely low-bandwidth systems such as packet radio or LoRa. Encrypted LXMF messages can also be encoded as QR-codes or text-based URIs, allowing completely analog *paper message* transport.
User-facing clients built on LXMF include:
- [Sideband](https://unsigned.io/sideband)
- [MeshChat](https://github.com/liamcottle/reticulum-meshchat)
- [Nomad Network](https://unsigned.io/nomadnet)
Community-provided tools and utilities for LXMF include:
- [LXMFy](https://lxmfy.quad4.io/)
- [LXMF-Bot](https://github.com/randogoth/lxmf-bot)
- [LXMF Messageboard](https://github.com/chengtripp/lxmf_messageboard)
- [LXMEvent](https://github.com/faragher/LXMEvent)
- [RangeMap](https://github.com/faragher/RangeMap)
- [LXMF Tools](https://github.com/SebastianObi/LXMF-Tools)
## Structure
LXMF messages are stored in a simple and efficient format, that's easy to parse and write.
### The format follows this general structure:
- Destination
- Source
- Ed25519 Signature
- Payload
- Timestamp
- Content
- Title
- Fields
### And these rules:
1. A LXMF message is identified by its __message-id__, which is a SHA-256 hash of the __Destination__, __Source__ and __Payload__. The message-id is never included directly in the message, since it can always be inferred from the message itself.
In some cases the actual message-id cannot be inferred, for example when a Propagation Node is storing an encrypted message for an offline user. In these cases a _transient-id_ is used to identify the message while in storage or transit.
2. __Destination__, __Source__, __Signature__ and __Payload__ parts are mandatory, as is the __Timestamp__ part of the payload.
- The __Destination__ and __Source__ fields are 16-byte Reticulum destination hashes
- The __Signature__ field is a 64-byte Ed25519 signature of the __Destination__, __Source__, __Payload__ and __message-id__
- The __Payload__ part is a [msgpacked](https://msgpack.org) list containing four items:
1. The __Timestamp__ is a double-precision floating point number representing the number of seconds since the UNIX epoch.
2. The __Content__ is the optional content or body of the message
3. The __Title__ is an optional title for the message
4. The __Fields__ is an optional dictionary
3. The __Content__, __Title__ and __Fields__ parts must be included in the message structure, but can be left empty.
4. The __Fields__ part can be left empty, or contain a dictionary of any structure or depth.
## Usage Examples
LXMF offers flexibility to implement many different messaging schemes, ranging from human communication to machine control and sensor monitoring. Here are a few examples:
- A messaging system for passing short, simple messages between human users, akin to SMS can be implemented using only the __Content__ field, and leaving all other optional fields empty.
- For sending full-size mail, an email-like system can be implemented using the __Title__ and __Content__ fields to store "subject" and "body" parts of the message, and optionally the __Fields__ part can be used to store attachments or other metadata.
- Machine-control messages or sensor readings can be implemented using command structures embedded in the __Fields__ dictionary.
- Distributed discussion or news-groups, akin to USENET or similar systems, can be implemented using the relevant fields and LXMF Propagation Nodes. Broadcast bulletins can be implemented in a similar fashion.
## Propagation Nodes
LXM Propagation Nodes offer a way to store and forward messages to users or endpoints that are not directly reachable at the time of message emission. Propagation Nodes can also provide infrastructure for distributed bulletin, news or discussion boards.
When Propagation Nodes exist on a Reticulum network, they will by default peer with each other and synchronise messages, automatically creating an encrypted, distributed message store. Users and other endpoints can retrieve messages destined for them from any available Propagation Nodes on the network.
## The LXM Router
The LXM Router handles transporting messages over a Reticulum network, managing delivery receipts, outbound and inbound queues, and is the point of API interaction for client programs. The LXM Router also implements functionality for acting as an LXMF Propagation Node.
Programatically, using the LXM Router to send a message is as simple as:
```python
import LXMF
lxm_router = LXMF.LXMRouter()
message = LXMF.LXMessage(destination, source, "This is a short, simple message.")
lxm_router.handle_outbound(message)
```
The LXM Router then handles the heavy lifting, such as message packing, encryption, delivery confirmation, path lookup, routing, retries and failure notifications.
## Transport Encryption
LXMF uses encryption provided by [Reticulum](https://reticulum.network), and thus uses end-to-end encryption by default. The delivery method of a message will influence which transport encryption scheme is used.
- If a message is delivered over a Reticulum link (which is the default method), the message will be encrypted with ephemeral AES-128 keys derived with ECDH on Curve25519. This mode offers forward secrecy.
- A message can be delivered opportunistically, embedded in a single Reticulum packet. In this cases the message will be opportunistically routed through the network, and will be encrypted with per-packet AES-128 keys derived with ECDH on Curve25519.
- If a message is delivered to the Reticulum GROUP destination type, the message will be encrypted using the symmetric AES-128 key of the GROUP destination.
## Wire Format & Overhead
Assuming the default Reticulum configuration, the binary wire-format is as follows:
- 16 bytes destination hash
- 16 bytes source hash
- 64 bytes Ed25519 signature
- Remaining bytes of [msgpack](https://msgpack.org) payload data, in accordance with the structure defined above
The complete message overhead for LXMF is only 111 bytes, which in return gives you timestamped, digitally signed, infinitely extensible, end-to-end encrypted, zero-conf routed, minimal-infrastructure messaging that's easy to use and build applications with.
## Code Examples
Before writing your own programs using LXMF, you need to have a basic understanding of how the [Reticulum](https://reticulum.network) protocol and API works. Please see the [Reticulum Manual](https://reticulum.network/manual/). For a few simple examples of how to send and receive messages with LXMF, please see the [receiver example](./docs/example_receiver.py) and the [sender example](./docs/example_sender.py) included in this repository.
## Example Paper Message
You can try out the paper messaging functionality by using the following QR code. It is a paper message sent to the LXMF address `6b3362bd2c1dbf87b66a85f79a8d8c75`. To be able to decrypt and read the message, you will need to import the following Reticulum Identity to an LXMF messaging app:
`3BPTDTQCRZPKJT3TXAJCMQFMOYWIM3OCLKPWMG4HCF2T4CH3YZHVNHNRDU6QAZWV2KBHMWBNT2C62TQEVC5GLFM4MN25VLZFSK3ADRQ=`
The [Sideband](https://unsigned.io/sideband) application allows you to do this easily. After you have imported the identity into an app of your choice, you can scan the following QR code and open it in the app, where it will be decrypted and added as a message.
[Paper message QR code](a8d24177d946de4f1f0a0fe1af9a1338:/page/blob.mu`g=reticulum|r=lxmf|ref=HEAD|path=docs/paper_msg_test.png)
You can also find the entire message in this link:
```
lxm://azNivSwdv4e2aoX3mo2MdTAozuI7BlzrLlHULmnVgpz3dNT9CMPVwgywzCJP8FVogj5j_kU7j7ywuvBNcr45kRTrd19c3iHenmnSDe4VEd6FuGsAiT0Khzl7T81YZHPTDhRNp0FdhDE9AJ7uphw7zKMyqhHHxOxqrYeBeKF66gpPxDceqjsOApvsSwggjcuHBx9OxOBy05XmnJxA1unCKgvNfOFYc1T47luxoY3c0dLOJnJPwZuFRytx2TXlQNZzOJ28yTEygIfkDqEO9mZi5lgev7XZJ0DvgioQxMIyoCm7lBUzfq66zW3SQj6vHHph7bhr36dLOCFgk4fZA6yia2MlTT9KV66Tn2l8mPNDlvuSAJhwDA_xx2PN9zKadCjo9sItkAp8r-Ss1CzoUWZUAyT1oDw7ly6RrzGBG-e3eM3CL6u1juIeFiHby7_3cON-6VTUuk4xR5nwKlFTu5vsYMVXe5H3VahiDSS4Q1aqX7I
```
On operating systems that allow for registering custom URI-handlers, you can click the link, and it will be decoded directly in your LXMF client. This works with Sideband on Android.
## Installation
If you want to try out LXMF, you can install it with pip:
```bash
pip install lxmf
```
If you are using an operating system that blocks normal user package installation via `pip`,
you can return `pip` to normal behaviour by editing the `~/.config/pip/pip.conf` file,
and adding the following directive in the `[global]` section:
```text
[global]
break-system-packages = true
```
Alternatively, you can use the `pipx` tool to install Reticulum in an isolated environment:
```bash
pipx install lxmf
```
## Daemon Included
The `lxmf` package comes with the `lxmd` program, a fully functional (but lightweight) LXMF message router and propagation node daemon. After installing the `lxmf` package, you can run `lxmd --help` to learn more about the command-line options:
```text
$ lxmd --help
usage: lxmd [-h] [--config CONFIG] [--rnsconfig RNSCONFIG] [-p] [-i PATH] [-v] [-q] [-s] [--exampleconfig] [--version]
Lightweight Extensible Messaging Daemon
options:
-h, --help show this help message and exit
--config CONFIG path to alternative lxmd config directory
--rnsconfig RNSCONFIG
path to alternative Reticulum config directory
-p, --propagation-node
run an LXMF Propagation Node
-i PATH, --on-inbound PATH
executable to run when a message is received
-v, --verbose
-q, --quiet
-s, --service lxmd is running as a service and should log to file
--exampleconfig print verbose configuration example to stdout and exit
--version show program's version number and exit
```
Or run `lxmd --exampleconfig` to generate a commented example configuration documenting all the available configuration directives.
## Support LXMF Development
You can help support the continued development of open, free and private communications systems by donating via one of the following channels:
- Monero:
```
84FpY1QbxHcgdseePYNmhTHcrgMX4nFfBYtz2GKYToqHVVhJp8Eaw1Z1EedRnKD19b3B8NiLCGVxzKV17UMmmeEsCrPyA5w
```
- Bitcoin
```
bc1pgqgu8h8xvj4jtafslq396v7ju7hkgymyrzyqft4llfslz5vp99psqfk3a6
```
- Ethereum
```
0x91C421DdfB8a30a49A71d63447ddb54cEBe3465E
```
- Liberapay: https://liberapay.com/Reticulum/
- Ko-Fi: https://ko-fi.com/markqvist
## Caveat Emptor
LXMF is beta software, and should be considered experimental. While it has been built with cryptography best practices very foremost in mind, it _has not_ been externally security audited, and there could very well be privacy-breaking bugs. If you want to help out, or help sponsor an audit, please do get in touch.
## Development Roadmap
LXMF is actively being developed, and the following improvements and features are currently planned for implementation:
- Sneakernet and physical transport functionality
- Content Destinations, and easy to use API for group messaging and discussion threads
- Write and release full API and protocol documentation
- Documenting and possibly expanding LXMF limits and priorities

153
SPEC.md Normal file
View file

@ -0,0 +1,153 @@
# LXMF Implementation-Derived Specification
## LXMessage Wire Format
This section describes behavior confirmed in the checked-in LXMF implementation
and by the reproducible vectors under `examples/`. It does not assign normative
requirements to independent implementations.
Evidence baseline:
- LXMF commit `d483d40d5cb994c6027802af1bf284dfeccc623b`
- Reticulum commit `2646f673ee061135d7c351ef44dce290ccd7e06e`
### Evidence status
- **Source-confirmed** means the behavior is directly implemented by the
checked-in LXMF or Reticulum source.
- **Vector-confirmed** means `perl tools/verify_examples.pl` confirms the
behavior for the deterministic minimal and stamped vectors.
- Behavior not covered by either form of evidence is marked **Unresolved**.
### Core serialized message
`LXMessage.pack()` constructs `lxmf_bytes` by concatenating:
```text
destination_hash || source_hash || signature || packed_payload
```
| Part | Position | Confirmed representation | Evidence |
|---|---:|---|---|
| `destination_hash` | bytes `0..15` | 16 bytes | Source-confirmed: `LXMF/LXMessage.py:40`, `LXMF/LXMessage.py:383`; Vector-confirmed |
| `source_hash` | bytes `16..31` | 16 bytes | Source-confirmed: `LXMF/LXMessage.py:384`; Vector-confirmed |
| `signature` | bytes `32..95` | 64 bytes | Source-confirmed: `LXMF/LXMessage.py:41`, `LXMF/LXMessage.py:385`; Vector-confirmed position and length only |
| `packed_payload` | bytes `96..end` | MessagePack array | Source-confirmed: `LXMF/LXMessage.py:362`, `LXMF/LXMessage.py:381-386`; Vector-confirmed |
The 16-byte hash length derives from
`RNS.Identity.TRUNCATED_HASHLENGTH // 8`. The checked-in Reticulum source sets
the truncated hash length to 128 bits. The 64-byte signature length derives
from `RNS.Identity.SIGLENGTH // 8`. See `RNS/Reticulum.py:146-147` and
`RNS/Identity.py:80-83` in the evidence-baseline Reticulum checkout.
### Payload
For messages generated by `LXMessage.pack()`, the payload is:
```text
[
timestamp,
title,
content,
fields
]
```
If a message stamp is available when packing, it is appended:
```text
[
timestamp,
title,
content,
fields,
stamp
]
```
| Index | Name | Confirmed behavior | Evidence |
|---:|---|---|---|
| `0` | `timestamp` | Set from `time.time()` when not already set, then packed as the first payload value | Source-confirmed: `LXMF/LXMessage.py:357`, `LXMF/LXMessage.py:362`; Vector-confirmed as MessagePack float64 for the deterministic vectors |
| `1` | `title` | Constructor string input is UTF-8 encoded to bytes; byte input is retained | Source-confirmed: `LXMF/LXMessage.py:130-133`, `LXMF/LXMessage.py:193-197`; Vector-confirmed as MessagePack binary |
| `2` | `content` | Constructor string input is UTF-8 encoded to bytes; byte input is retained | Source-confirmed: `LXMF/LXMessage.py:135-136`, `LXMF/LXMessage.py:202-206`; Vector-confirmed as MessagePack binary |
| `3` | `fields` | Constructor input must be a dictionary or `None`; `None` becomes an empty dictionary | Source-confirmed: `LXMF/LXMessage.py:138`, `LXMF/LXMessage.py:215-219`; Vector-confirmed for an empty MessagePack map |
| `4` | `stamp` | Optional value appended after the four base payload values | Source-confirmed: `LXMF/LXMessage.py:371-373`; Vector-confirmed as MessagePack binary in the stamped vector |
The payload order above is source- and vector-confirmed. It differs from the
order stated in the upstream `README.md`, which lists content before title.
### Message ID
The message ID, also stored as `LXMessage.hash`, is:
```text
SHA-256(
destination_hash ||
source_hash ||
msgpack([timestamp, title, content, fields])
)
```
The optional stamp is excluded from the MessagePack payload used to calculate
the message ID. This is source-confirmed by both packing and unpacking behavior,
and vector-confirmed by the minimal and stamped vectors producing the same
message ID. See `LXMF/LXMessage.py:362-369` and
`LXMF/LXMessage.py:754-764`. Reticulum defines `full_hash()` as SHA-256 at
`RNS/Identity.py:238-246`.
The message ID is not included in `lxmf_bytes`.
### Signature input
`LXMessage.pack()` requests a signature over:
```text
destination_hash ||
source_hash ||
msgpack([timestamp, title, content, fields]) ||
message_id
```
The optional stamp is excluded from the signature input. The construction of
the signature input is source-confirmed at `LXMF/LXMessage.py:375-378` and
`LXMF/LXMessage.py:762-764`, and vector-confirmed.
The deterministic vectors contain placeholder signature bytes. They confirm
the signature position and input bytes, but do not confirm signing or signature
validation.
### Delivery representations
The core `lxmf_bytes` representation is modified or wrapped for some delivery
methods:
| Delivery representation | Confirmed serialized data | Evidence |
|---|---|---|
| Direct packet or resource | Full `lxmf_bytes` | Source-confirmed: `LXMF/LXMessage.py:635-636`, `LXMF/LXMessage.py:653-654` |
| Opportunistic packet | `lxmf_bytes` without the leading destination hash | Source-confirmed: `LXMF/LXMessage.py:633-634` |
| Propagated message data | `destination_hash || encrypt(lxmf_bytes after destination_hash)` | Source-confirmed: `LXMF/LXMessage.py:429-436` |
| Propagation transfer wrapper | MessagePack `[time.time(), [propagated_message_data]]` | Source-confirmed: `LXMF/LXMessage.py:436` |
| Paper message data | `destination_hash || encrypt(lxmf_bytes after destination_hash)` | Source-confirmed: `LXMF/LXMessage.py:446-451` |
| Paper URI | URL-safe Base64 of paper message data, without `=` padding, prefixed by `lxm://` | Source-confirmed: `LXMF/LXMessage.py:698-707` |
These delivery representations are not covered by the current deterministic
test vectors.
### Unresolved behavior
The following behavior is not established by the current source inspection and
test-vector coverage:
- Normative requirements for independent LXMF implementations.
- Cross-implementation Ed25519 signature generation and validation.
- Accepted signature encodings beyond the fixed 64-byte position generated by
the checked-in implementation.
- A universal MessagePack float width for timestamps on every supported
platform. The deterministic vectors confirm float64 only for those vectors.
- Interoperable constraints on the contents, key types, value types, ordering,
and nesting depth of `fields`.
- Interoperable constraints on stamp length and encoding.
- Required handling of malformed, truncated, non-canonical, or payload arrays
containing fewer than four or more than five entries.
- Byte-for-byte vectors for opportunistic, propagated, paper, URI, encrypted,
or persisted-container representations.

415
Zen of Reticulum.md Normal file
View file

@ -0,0 +1,415 @@
# Zen of Reticulum
## I: The Illusion Of The Center
For the better part of a generation, we have been taught to visualize the digital world through the lens of hierarchy. The mental maps we carry are dominated by a single, misleading image: **The Cloud**.
We imagine the network as a vast, ethereal space "up there" or "out there". A centralized repository of services and data to which we, the lowly clients, must connect. We build our software with this assumption hardcoded into our logic: *There is a server. The server has the authority. The server knows the way. I must find the server to function*.
This is the Client-Server mental model, and it is the primary obstacle to understanding Reticulum.
### Fallacy Of The Cloud
The first step in the Zen of Reticulum is to realize that *there is no cloud*. There is only other people's computers. When you build for the cloud, you are building *for* a landlord. You are accepting that your application's existence is conditional on the permission, uptime, and continued goodwill of a central authority.
In Reticulum, you must shift your thinking from "connecting to" to "being among". Reticulum is not a service you subscribe to - *it is a fabric you inhabit*. There is no "up there". There is only *here* and *there*, and the space between them is peer-to-peer.
### Decentralization Or Uncentralizability?
It is common to hear the word "decentralized" thrown around in modern tech circles. But often, this is merely a marketing term for "slightly distributed centralization". A blockchain with a few dominant miners, or a federated protocol with a few giant servers. *In practice*, it's still centralized. It simply has a few centers instead of one.
Reticulum goes further. It wants **Uncentralizability**.
This is not a wishful political stance, but a foundational mathematical characteristic of the protocol, onto which everything else has been built. Reticulum assumes that every peer on the network is potentially hostile, and every link is potentially compromised. It is designed with no "privileged" nodes. While some nodes may act as Transport Instances - forwarding traffic for others - they do so *blindly*, and they only know about their immediate surroundings, and nothing more. They route based on cryptographic proofs, not on administrative privilege. They cannot see who is talking to whom, nor can they selectively manipulate traffic without breaking their own ability to route entirely.
The system is designed to make hierarchy structurally impossible. You cannot hijack an address, because there is no central registry to hijack. You cannot block a user, because there is no central switch to flip. You can offer paths through the network, but you can't force anyone to use them.
### Death To The Address
To break free of the center, you must also let go of the concept of the "Address".
In the IP world, an address is a location. It is a coordinate in a *deeply hierarchical* and static grid. If you move your computer to a different house, your address changes. If your router reboots, your address might change. Your *identity* is bound to your *location*, and therefore, it is fragile, and easily controlled.
Reticulum abolishes this link between *Identity* and *Location*.
In Reticulum, an address is not a place; it is a **Hash of an Identity**. It is a cryptographic representation of *who* you are, not *where* you are. Because of this, your address is portable. You can take a laptop from a WiFi cafe in Berlin, to a LoRa mesh in the mountains, to a packet radio link on a boat, and your "address" - your *Destination Hash* - never changes.
The network does not route to a place; it routes to a *person* (or a machine). When you send a packet, you are not targeting a coordinate in a grid; you are encrypting a message for a specific entity. The network dynamically discovers where that entity currently resides, and it does so in a way where no one really knows where that entity is actually located physically.
**Consider:**
- **The Old Way:** *"I am at `192.168.1.5`. Come find me"*.
- **The Zen Way:** *"I am `<327c1b2f87c9353e01769b01090b18f2>`. Wherever I am, my peers can reach me"*.
Once you stop thinking about servers and start thinking about portable identities, where everyone can always reach everyone else directly, the illusion of the center fades away. You realize there *is* no center holding the network together. No coordinators or bureaucrats required. The network is simply the sum of its peers, communicating directly, sovereignly, and without a master.
## II: Physics Of Trust
*Paranoia Is A Great Design Principle*
If we accept that there is no center - that the network is a chaotic, peer-to-peer mesh - we are forced to confront a terrifying reality: **There is no one guarding the door**.
In the traditional networking mindset, we rely on the concept of the "trusted core". We assume our local coffee shop WiFi is safe, or that the backbone providers are neutral custodians. We build our security like a castle: strong walls on the outside, soft and trusting on the inside. We use encryption only when we step out into the "wild" internet.
### Hostile Environments
The Zen of Reticulum requires you to invert this. You must assume that *every* environment is hostile. This isn't cynicism, just uncaring physics.
When you transmit information over radio waves, you are shouting into a crowded room. Anyone can listen. When you traverse the internet, your packets pass through routers controlled by strangers, corporations, and state actors. Assuming privacy in this environment without cryptographic protection is not optimism but gross negligence.
Reticulum is built on the premise that every link is tapped, and every peer is a potential adversary. If your system cannot survive an adversary owning the physical layer, it cannot survive at all.
But this is the paradox: By assuming the network is hostile, you make it safe. When you accept the dangers for what they are, they become manageable. When you stop trusting the infrastructure and start trusting the math, you eliminate the single point of failure: Human integrity.
### Encryption Is Not A Feature
In the world of TCP/IP, encryption is an afterthought. It is a layer we slap on top of the protocol (HTTPS, TLS) to patch the security holes of the original design. It is a "feature" you sometimes *enable* for "sensitive data". This is fundamentally flawed, since all data is sensitive.
In Reticulum, encryption is **gravity**.
It is not optional. It is not a plugin. It is the *fundamental force that allows the network to exist*. If you were to strip the encryption from Reticulum, the routing would break. The Transport system uses cryptographic signatures and entropy to verify paths and pass information. If packets were plaintext, intermediate nodes could not prove that a route was valid, nor could endpoints prevent spoofing or tampering.
In Reticulum, the entropy of the encrypted packet *is* the routing logic.
To ask for a version of Reticulum without encryption is like asking for a version of the ocean without liquid. You are not asking for a feature change; you're asking for a different physical universe. We design for a universe where information has mass, structure, and integrity.
### Zero-Trust Architectures
We must unlearn our reliance on **Institutional Trust**.
For decades, we have been trained to trust authorities. We trust a website because a chain of Certificate Authorities (companies we don't know) vouches for it. We trust an app because it is in an app store (run by a corporation we don't control). We trust a message because it comes from a phone number assigned by a telecom. Yet, everything in our digital information sphere today is more untrustworthy and risky than a medieval second-hand underwear market.
Reticulum replaces institutional trust with **Cryptographic Proof**.
In Reticulum, you do not trust a node because it has a nice hostname or because it is listed in a directory. You trust it because it holds the private key corresponding to the Destination Hash you are communicating with. This trust is binary, mathematical, and **absolute**. Either the signature matches, or it does not. There is no "maybe".
This shift moves the power from the institution to the individual. You become the ultimate arbiter of your own trust relationships. You decide which keys to accept, which paths to follow, and which identities to recognize.
**Consider:**
- **The Old Way:** *"I trust this site because the browser says the lock icon is green"*.
- **The Zen Way:** *"I trust this destination because I have verified its hash fingerprint out-of-band, and the math confirms the signature"*.
When you internalize the Physics of Trust, you stop looking for protection from firewalls, VPNs, and Terms of Service agreements. You realize that true security comes from the design of the protocol itself. You can stop trusting the cloud, and you start trusting the code - because you can verify it yourself.
## III: Merits Of Scarcity
*Every Bit Counts*
We have grown addicted to abundance. In the modern digital ecosystem, bandwidth is treated as an endless, flat ocean. We stream high-definition video without a thought, we ship entire libraries of code just to render a single button, and we measure performance in gigabits per second. This abundance has hollowed out our craft. When constraints vanish, efficiency dies, and with it, a certain kind of Clarity and Quality.
Reticulum asks you to step out of the ocean and onto the tightrope.
### The Bandwidth Fallacy
The Zen of Reticulum requires the realization that **5 bits per second is a valid speed**.
To a modern developer, this sounds like paralysis. But there is a profound freedom in limits: When you have a gigabit connection, you can be incredibly sloppy. You can be wasteful. You can push your problems onto the infrastructure. *"It’s slow? Get a faster router"*.
But on a high-latency, low-bandwidth link (be it a noisy HF radio channel or a tenuous LoRa hop) you cannot push problems anywhere. You must solve them. The network does not negotiate with waste.
This forces a shift from consumption to interaction. You are no longer, then, consuming a service provided by a fat pipe; you are engaging in a careful negotiation with the physical medium. The medium becomes a partner in the conversation, not just a dumb conduit. You suddenly need to *understand the world to be in it*.
### Cost Of A Byte
In a scarce economy, a byte is not just data, but energy, time, and space.
Every byte you transmit consumes battery life on a solar-powered node. It occupies valuable airtime that could have been used by another peer. It represents a measurable slice of the electromagnetic spectrum.
When you internalize this, you begin to write code differently. You stop asking, "How much data can I send?" and start asking, "What is the *minimum* amount of information required to convey this intent? How can I best utilize my informational entropy?"
This is where the elegance of Reticulum shines. The protocol is designed to strip away the non-essential. A link establishment takes three very small packets. A destination hash fits in 16 bytes. The overhead is vanishingly small, leaving almost the entire channel for the message itself.
**Consider:**
- **The Old Way:** *"I need to send a status update. I'll send a JSON object with metadata, timestamps, and user profile info (15KB)."*
- **The Zen Way:** *"I need to send a status update. I'll send a single byte representing the state code. The context is already known."*
This is of course optimization, but more importantly, *it is a form of respect*. Efficiency in a shared medium is an act of stewardship. By taking only what you need from the network, you leave room for others. The network listens to those who speak with purpose.
### Flow & Time
Scarcity also teaches us about time. We have become addicted to the *synchronous* now - the instant ping, the real-time stream. But Reticulum embraces *asynchronous* time.
When links are intermittent and latency is measured in minutes or hours, "real-time" is an illusion. Reticulum doesn't encourage **Store and Forward** as a mere fallback, but as a primary mode of existence. You write a message, it propagates when it can, and it arrives when it arrives.
This changes the psychological texture of communication. It removes the anxiety of the immediate response. It allows for contemplation. You are not demanding the recipient's attention *right now*; you are placing a gift in their path, to be found when they are ready.
By designing for delay, you design for resilience. You are no longer building a house of cards that collapses when a single packet drops. You are building a stone arch that distributes the load *over time*.
### Liberation From Limits
There is a strange optimism in scarcity. When you are forced to work within strict constraints, you are forced to prioritize. *You* must decide what truly matters. *That* is the real core of agency.
In the infinite fantasy world of The Cloud, everything is urgent, so nothing is. In the economy of Reticulum, the cost of transmission forces you to weigh the value of your message. Do you really need to send that heart beat? Is that photo essential?
When you strip away the noise, what remains is *signal*.
This discipline creates a different kind of developer. It creates a craftsman who understands that the best code is the code you don't have to write. It creates a user who understands that the most powerful message is the one that is *understood*, not the one that is loudest. In the world of Reticulum, you are not a mere consumer of bandwidth; you are an architect of intent.
## IV: Sovereignty Through Infrastructure
**Be Your Own Network**
We live in an era of digital tenancy. We lease our connectivity from ISPs. We rent our storage from cloud providers. We even borrow our identity from social media platforms. We are tenants in a house we did not build, governed by rules we did not write, subject to eviction at the whim of a landlord who has never met us.
The Zen of Reticulum is the realization that you *can* own the house.
### A Carrier-Grade Fallacy
For decades, we have been gaslit into believing that networking is really not just hard, but impossible. It is presented as a dark art reserved for telcos and billionaires, requiring millions of dollars of fiber optics, climate-controlled data centers, and armies of engineers. We are told that building reliable infrastructure is "too complex" for the individual or small organization.
This is a big, fat lie.
Physics is simple. A radio wave needs a transmitter and a receiver. A packet needs a path. The "complexity" of the modern internet is largely bureaucratic - a mountain of billing systems, regulatory capture, and legacy cruft designed to keep the gatekeepers in power.
Reticulum strips away the bureaucracy. It runs on hardware that costs the price of a dinner. It runs on spectrum that is free to use. It demonstrates that a robust, planetary-scale network does not require a Fortune 500 company. It requires only the will to deploy, and the distributed, uncoordinated efforts of many individuals.
### Personal Infrastructure
This is where the rubber meets the road. You can read about Reticulum, you can understand the theory, but the insights only arrive when you plug in a radio and run a Transport Node. Suddenly, you are no longer a consumer. You're an operator.
This shift is subtle but profound. When you run your own infrastructure, the network ceases to be a service that is provided *to* you. It becomes a space that you *inhabit*. You become responsible for the flow of information. You gain an intimate understanding of the medium - the way the weather affects the radio waves, the way the topology changes, the way the packets dance through the ether.
There is a quiet competence that comes from this. You stop asking "Is the internet down?" and start asking "Is *my* links up?" You stop waiting for a technician and start checking the logs. This is a form of strength. To understand the system that carries your words is to be free from the mystery that keeps you dependent.
### The Ability To Disconnect
Why go to the trouble? Why buy the radio, write the config, and leave the Pi running in the corner?
Because the old, centralized network is fragile. And because most of us doesn't even really want to be there anymore.
The internet we rely on today is a chain of single points of failure. Cut the undersea cable, and a continent goes dark. Shut down the power grid, and the cloud evaporates. Deprioritize the "wrong" traffic, and the flow of information is strangled.
Sovereignty is the ability to survive the cut, whether or not that cut was an accident or on purpose.
When you build your own infrastructure, you build a lifeline. Reticulum is designed to function over media that the traditional internet cannot touch - bare wires, battery-powered radios, ad-hoc WiFi meshes. When the grid fails, or the censors arrive, or the bill goes unpaid, your Reticulum network continues to hum.
This is not about "dropping out" of society. It is about building a substrate on which an actual *Society* can function.
**Consider:**
- **The Old Way:** "My connection is slow. I should call my ISP and complain."
- **The Zen Way:** "The path is noisy. I will adjust the antenna or find a better route."
By taking ownership of the infrastructure, you take ownership of your voice. You stop shouting into someone else's megaphone and start building your own. The network is no longer something that happens to you; it is something you make happen.
# V: Identity and Nomadism
**A Fluid Self**
In the old world, you are defined by your coordinates. If you are at `34.109.71.5`, you're *here*. If you unplug the cable and walk down the street, you vanish. Your digital self evaporates because it was tethered to the wall. You are a ghost in the endless machinations of gears, levers and transistors, bound to the hardware, and those that own it.
This creates a subtle, constant anxiety. We are terrified of disconnecting because, in the architecture of the old web, disconnecting is a kind of death.
The Zen of Reticulum offers a different way to be.
### Portable Existence
In Reticulum, your identity is not a location, or a username granted by a service. It is a cryptographic key - a complex, unique mathematical signature that exists independently of the physical world. You can carry it only in your mind, if you want to.
Think of it less like a street address and more like a name. *A true name*.
If you travel from Berlin to Tokyo, you do not change your name. You are still you. The people who know you can still recognize you. Reticulum applies this principle to the network layer. Your Destination Hash is **invariant**. It travels with you, stored securely on your device, *immutable as a stone*.
This changes the relationship between you and the machine. You are not "logged into" the network via a specific gateway. You *are* the endpoint. The network does not connect to a place; *it converges on you*.
### Roaming Nodes
This freedom introduces a new concept of time and space: **Nomadism**.
Because your identity is portable, your connectivity can be fluid. You can be sitting at a desk connected to a fiber backbone one moment, and walking through a field connected only to a long-range LoRa mesh the next. To the rest of the network, nothing has changed. Your friends do not need to update your contact info. The messages they send do not bounce back. The network senses the shift in the medium and reroutes the flow of data automatically.
You are no longer a stationary node in a fixed grid. You are a wanderer in a fluid medium.
The interfaces - whether it is WiFi, Ethernet, Packet Radio, or a physical wire - is merely the clothing your node wears. You change it to suit the environment. Underneath, you remain the same. This is the liberation of the protocol. It treats the physical medium as a transient circumstance, not a definition of self.
**Consider:**
- **The Old Way:** *"I lost connection. I have to reconnect to the VPN to tell them where I am now."*
- **The Zen Way:** *"I moved. The network subtly bends to accomodate this new reality."*
### Announcing Presence
How does the network find a wanderer? It listens.
In the IP world, we query directories. We ask a server, "Where is Mark?" The server checks its database and gives us a coordinate. This means that someone, somewhere, is keeping track of you. It assumes and *requires* surveillance.
Reticulum replaces surveillance with **Announces**.
Instead of asking a central authority where you are, you simply state your presence. You broadcast a cryptographic proof: "I am here, and I am who I say I am". This ripples out through the mesh. Your neighbors hear it, update their path tables, and pass it on.
This is a quiet, organic process. It is the digital equivalent of lighting lanterns in the dark. You do not need to chase the light; you let the light find you. It respects your autonomy. You choose when to announce, how often to speak, and to whom. You also choose when to disappear - for but a moment or perpetually.
### Anchor In The Flow
There is a deep peace in this nomadism. It teaches you that stability does not come from standing still. Stability comes from *internal coherence*.
By holding your own private key, you hold your own center of gravity. The world around you; the infrastructure, the topography and the availability of links can all shift chaotically. Storms can knock out towers. Cables can be cut. The internet can go down.
But as long as you possess your key, you possess your identity. The entire infrastructure can be destroyed and rebuilt, and you are still you. Nothing lasts, yet nothing is lost.
You become a sovereign entity moving through the noise, connected not by the rigidity of cables, but by the fluidity of recognition. The network becomes a place you inhabit, rather than a utility you subscribe to: You are at home in the ether.
## VI: Ethics Of The Tool
**Technology With Conscience**
You have unlearned the center. You have accepted the physics of trust. You have embraced the economy of scarcity and the freedom of unbound nomadism. You are standing in a new space. Now, look at the tool in your hand.
In the old world, we were taught that technology is neutral. We are told that "guns don't kill people, people do", or that a component is just a component, indifferent to what its combinatorial potential is. This is a convenient lie. It serves only to allow the builders to wash their hands of responsibility.
But we know better now. We know that **architecture is politics**, and *politics is control*. The way you build a system determines how it will be used. If you build a system optimized for mass surveillance, you *will* get a panopticon. If you build a system optimized for centralized control, you *will* get a dictatorship. If you build a system optimized for extraction, you *will* get a parasite.
The Zen of Reticulum asserts that a tool is never neutral.
On the very contrary: A tool is intent, **crystallized**.
### The Harm Principle
Why does the Reticulum License forbid the software from being used in systems designed to harm humans? Is it not just a restriction on freedom?
It is a restriction on *license*, yes, but it is an expansion of *freedom*.
Building powerful tools without a moral compass is in no way virtuous or commendable, it is plain and simple irresponsibility.
A tool that can easily be used to oppress is a real danger to the user. If you build a network that can be turned against you by a tyrant, you are not free. You are merely waiting for the leash to tighten. By encoding the "Harm Principle" into the legal DNA of the reference implementation, we are building a safeguard. We are stating, clearly and immutably, that *this tool* is for **life**, not for death.
This aligns the software with the interests of humanity. It cements that the network cannot be conscripted into a kill-system, a weaponized drone controller, or a torture device without breaking the license and the law. It is a line drawn in the sand - not by a government or external authority, but by the creators of the tool itself.
**Consider:**
- **The Old Way:** *"It's just software. How people use it is not my problem."*
- **The Zen Way:** *"This software is a habitat. I will not allow it to be used to build a cage."*
It is *your* choice whether to align with this - we are not forcing this stance on anyone. If you choose to align with life over death, with creativity over destruction, we grant you an immensely powerful tool, to own and build with as you please. If you do not, we deny it.
If you do not like this, we most assuredly do not need you here, and you are on your own.
### Public Domain Protocol
This leads to a vital distinction: The difference between the *idea* and the *implementation*.
The protocol - the mathematical rules of how Reticulum works - is dedicated to the Public Domain. It belongs to humanity. **No one can own it**. Anyone can implement it, improve it, or adapt it. This is the core idea of free communication, which itself must be forever free.
But the functional, deployed *reference implementation* - the Python code, the maintenance, the years of labor - has a conscience. This distinction is the engine of sustainability. It allows the protocol to be universal, while ensuring that the specific labor of the builders is not hijacked to undermine the foundational intent of the project itself. From this document, it should be very clear what this intent is.
If you want to build a system with Reticulum that manipulates and damages users for profits or targets missiles, you can use the public domain protocol, and start from scratch. But you cannot take our work. You must do your own. This serves as a pillar of accountability. If you want to build a weapon, *you* go and forge the steel yourself, while the world observes. And when the blood is drawn - it is on **your** hands.
### Preserving Human Agency
We live in an era of predatory extraction. The open-source commons is being scraped, ingested, and regurgitated by machine learning algorithms, whose corporate owners seek to replace the very humans who built those commons. Our code, our words, and our creativity is being used to train systems that are specifically designed to make us obsolete, without offering anything else in return than serfdom and leashes.
Reticulum stands against this.
The license protects the software from being used to feed the beast. It draws a hard line: This tool is for *people*. It is for human-to-human connection. It is not a dataset to be strip-mined for the purpose of building a synthetic overlord, puppeteered by a miniscule conglomerate of controllers.
This is a radical act of preservation. By protecting the code from AI appropriation, we are protecting space for human agency. We are ensuring that there remains a digital realm where the actors are flesh, blood and soul, where decisions are made by minds, not overlords hiding behind models.
When you use Reticulum, you are using a tool that respects you. It does not see you as a product to be tracked. It does not see your data as fuel for an algorithm. It sees you as a sovereign, equal peer.
This changes the foundational premise of using the technology. It restores dignity to the interaction. You are not the user of a service; you are a participant in a mutual covenant. The tool aligns with your autonomy, rather than eroding it.
In this way, ethics is not a restriction, but a foundation. It is the foundation that helps ensure the network will still belong to you tomorrow.
## VII: Design Patterns For Post-IP Systems
**Practical Philosophy for Developers**
The philosophy is useless if it cannot be hammered into code. The metaphors we have explored - nomadism, scarcity, trust - are not just poetry, but real-world engineering constraints. When you sit down to write software for Reticulum, these concepts must shape the very structure of your application.
We are now moving from the *why* to the *how*. This is where the abstract becomes concrete, and where you will see the true depth of the patterns we have been weaving.
### Store & Forward
The web has trained us to be impatient. We write synchronous code. We fire a request and we wait, blocking the UI, holding our breath. If the response doesn't come in 250 milliseconds, we show a spinner. If it doesn't come in five seconds, we show an error. We treat network connectivity as a binary state: either we are "online" or we are "broken".
This is brittle. It is a rejection of reality.
In Reticulum, connectivity is a spectrum, and presence is asynchronous. If at all applicable to your intent, you must design your applications to embrace **Store & Forward**.
Instead of demanding an immediate answer, your application should act as a patient participant. You create a message for someone or something in the mesh. The network holds it. It carries it from node to node, perhaps over hours or days, waiting for the recipient to appear. When they finally surface, the message is delivered. This requires a shift from "request/response" to "event/handler". How exactly you do this is a challenge for you to solve intelligently within your problem domain, but Reticulum-based systems already exist that does this extremely well, and you can use them for inspiration.
**Consider:**
- **The Old Way:** `Connect() -> Send() -> Wait() -> Crash if timeout.`
- **The Zen Way:** `Send() -> Continue living. -> Receive() when it arrives.`
This changes the user experience profoundly. It removes the anxiety of the loading bar. It creates a sense of continuity. The user is not "waiting for the network"; they are interacting with a persistent log of communication that lives in the network itself.
### Naming Is Power
In the IP world, we are slaves to the Domain Name System. We rely on a hierarchy of registrars to map human-readable names to machine-readable addresses. This hierarchy is a choke point. If the registrar revokes your domain, or if the DNS server goes down, you vanish.
Reticulum dissolves this hierarchy with **Hash-based Identity**.
In this design pattern, a name is not a string you look up; it is a cryptographic destination you verify. When you design for Reticulum, you stop asking the user for a URL and start asking for a Destination or Identity Hash.
This feels strange at first. A hash like `<83b7328926fed0d2e6a10a7671f9e237>` looks alien compared to `myfriend.com`. But that alienness is the armor. It **cannot** be spoofed. It **cannot** be censored by a registrar. It is **absolute**.
Designing for this means shifting your UI metaphors. You are no longer browsing a web of pages; you are managing a ledger of keys. You are building an "Address Book" that is actually a keyring. The names are given by the user, and the power stays with them. That hashes look complex is directly analogous to the strengths of the bonds formed by their use. It forces the user to engage in a moment of verification, an out-of-band handshake, which restores the human element of trust that SSL certificates stripped away.
### The Interface Is The Medium
One of the most liberating patterns in Reticulum is **Transport Agnosticism**.
In traditional networking, your code is often littered with transport logic. "Am I on WiFi? Check bandwidth. Am I on Cellular? Check data plan. Am I on Ethernet?". You are constantly micromanaging the pipe.
In Reticulum, you write to the API, and the API writes to the medium. You send a packet to a Destination. You do not care if that packet travels over a TCP tunnel, a LoRa radio wave, or a serial wire interface. That is the stack's concern.
This allows you to write **Universal Applications**.
Imagine a messaging app. You write it once. It works on a laptop connected to fiber. It works on a phone in the city using WiFi. And, without a single line of code changed, it works on a device in the wilderness, talking only to other devices via radio.
The pattern is simple: **Never code to the hardware. Code to the intent.**
**Consider:**
- **The Old Way:** `socket.connect(ip, port)`
- **The Zen Way:** `RNS.Packet(destination, data).send()`
By abstracting the medium, you make your software immortal to changes in infrastructure. The user might switch from a 4G hotspot to a HF modem tomorrow. Your software doesn't need to know. It simply continues the conversation.
### Emergent Patterns
When you combine these patterns - *Store & Forward*, *Hash-based Identity*, and *Transport Agnosticism* - you create software that feels fundamentally different.
It feels *grounded*. It doesn't flicker when the signal drops. It doesn't panic when the server is down. It has weight. It has persistence. It has *relevance*.
You are no longer building a "client" that begs a "server" for attention. You are building an autonomous agent that exists within the mesh. It speaks when it needs to, listens when it can, and carries its identity with it wherever it goes.
This is the culmination of the Zen. The code is not just a set of instructions: It is a behavioral envelope. It is a way of *being* in the network.
## VIII: Fabric Of The Independent
We have stripped away the illusions. We have seen that the center is empty, that trust *must* be hard, that resources are finite, and that we must own our infrastructure. We have seen that tools have ethics and that our identity can move fluidly.
This is a reclaiming of the commons. For too long, we have allowed the most vital substrate of human society - *our ability to speak to one another* - to be colonized by entities that do not share our interests. We have allowed the architecture of our communication to be designed by accountants rather than architects.
We are taking it back. Not by petitioning the masters, but by building the new world within, over, under and around the shell of the old.
### The Work Is Finished
The heavy lifting is done.
The protocol is in the public domain, a gift to humanity that can never be taken away. The software is written, tested, and running on devices scattered across the globe. The manual lies open before you. The source code for the reference implementation is now distributed on hundreds of thousands of devices across the planet. No one can delete or destroy it. The hardware is accessible and abundant.
It was a hard road to get here, but we got here. Now, there is no roadmap committee waiting for approval. There is no venture capital dictating the user experience. There is no CEO to sign off on the next feature release.
There is only you.
The barrier to entry is no longer complexity: It is the mere habit of dependency. You were conditioned to wait. Wait for the app update. Wait for the ISP to fix the line. Wait for the platform to allow the post. Wait for the government to change the policies. Wait for the likes. Wait for the revolution to be televised.
The revolution never was televised.
It is packetized.
### Open Sky
The future of this technology is a construction project.
It looks like a single node on a windowsill, listening to the static. It looks like a message sent to a neighbor, bypassing the noise of the commercial web. It looks like a community mesh that grows, link by link, hop by hop, carried by hands that care more about connection than profit.
You have the blueprints. You have the tools. You have the philosophy. The noise of the old world has fallen away, leaving you with the quiet clarity of the open spectrum.
*Mark, early 2026*

72
docs/example_receiver.py Normal file
View file

@ -0,0 +1,72 @@
import RNS
import LXMF
import time
required_stamp_cost = 8
enforce_stamps = False
def delivery_callback(message):
global my_lxmf_destination, router
time_string = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(message.timestamp))
signature_string = "Signature is invalid, reason undetermined"
if message.signature_validated:
signature_string = "Validated"
else:
if message.unverified_reason == LXMF.LXMessage.SIGNATURE_INVALID:
signature_string = "Invalid signature"
if message.unverified_reason == LXMF.LXMessage.SOURCE_UNKNOWN:
signature_string = "Cannot verify, source is unknown"
if message.stamp_valid:
stamp_string = "Validated"
else:
stamp_string = "Invalid"
RNS.log("\t+--- LXMF Delivery ---------------------------------------------")
RNS.log("\t| Source hash : "+RNS.prettyhexrep(message.source_hash))
RNS.log("\t| Source instance : "+str(message.get_source()))
RNS.log("\t| Destination hash : "+RNS.prettyhexrep(message.destination_hash))
RNS.log("\t| Destination instance : "+str(message.get_destination()))
RNS.log("\t| Transport Encryption : "+str(message.transport_encryption))
RNS.log("\t| Timestamp : "+time_string)
RNS.log("\t| Title : "+str(message.title_as_string()))
RNS.log("\t| Content : "+str(message.content_as_string()))
RNS.log("\t| Fields : "+str(message.fields))
if message.ratchet_id:
RNS.log("\t| Ratchet : "+str(RNS.Identity._get_ratchet_id(message.ratchet_id)))
RNS.log("\t| Message signature : "+signature_string)
RNS.log("\t| Stamp : "+stamp_string)
RNS.log("\t+---------------------------------------------------------------")
# Optionally, send a reply
# source = my_lxmf_destination
# dest = message.source
# lxm = LXMF.LXMessage(dest, source, "Reply", None, desired_method=LXMF.LXMessage.DIRECT, include_ticket=True)
# router.handle_outbound(lxm)
r = RNS.Reticulum()
router = LXMF.LXMRouter(storagepath="./tmp1", enforce_stamps=enforce_stamps)
identity = RNS.Identity()
my_lxmf_destination = router.register_delivery_identity(identity, display_name="Anonymous Peer", stamp_cost=required_stamp_cost)
router.register_delivery_callback(delivery_callback)
RNS.log("Ready to receive on: "+RNS.prettyhexrep(my_lxmf_destination.hash))
# You can set a propagation node address to test receiving
# messages from a propagation node, instead of directly
# router.set_outbound_propagation_node(bytes.fromhex("e75d9b6a69f82b48b6077cf2242d7499"))
# This loop allows you to execute various actions for testing
# and experimenting with the example scripts.
while True:
input()
RNS.log("Announcing lxmf.delivery destination...")
router.announce(my_lxmf_destination.hash)
# input()
# RNS.log("Requesting messages from propagation node...")
# router.request_messages_from_propagation_node(identity)

69
docs/example_sender.py Normal file
View file

@ -0,0 +1,69 @@
import LXMF
import RNS
import time
import random
random_names = ["Tom", "Delilah", "Nancey", "Williams", "Neomi", "Curtis", "Alexa", "Theodora", "Ted", "Dinorah", "Nicol", "Drusilla", "Annalisa", "Verlene", "Latesha", "Tina", "Mia", "Brock", "Timothy", "Philip", "Willian", "Reyna", "Simona", "Mimi", "Stanford", "Ferne", "Catalina", "Lucie", "Jaye", "Natasha", "Willetta", "Isabel", "Esperanza", "Ciara", "Eusebio", "William", "Elma", "Angelica", "Coreen", "Melani", "Jonathan", "Maryland", "Caroline", "Gregg", "Ora", "Jacqui", "Letty", "Roselle", "Oralee", "Angla"]
random_titles = ["Long time", "Hi again", "Hi there", "Test message", "", "", "Something different"]
random_msgs = ["If wishes were horses then beggars might fly. Stuff like that. It's enough to drive you crazy.", "'My ident cards were stolen,' Jason said. 'That fivehundred-dollar bill is yours if you can get me to someone who can replace them. If you're going to do it, do it right now; I'm not going to wait.' Wait to be picked up by a pol or a nat, he thought. Caught here in this rundown dingy hotel.", "A six, no matter what the external circumstances, will always prevail. Because that's the way they genetically defined us.", "'Should be there in an hour,' he called back over his shoulder to Chuck. Then he added, in an afterthought, 'Wonder if the computer’s finished its run. It was due about now.'. Chuck didn’t reply, so George swung round in his saddle. He could just see Chuck’s face, a white oval turned toward the sky."]
def delivery_callback(message):
pass
r = RNS.Reticulum()
router = LXMF.LXMRouter(storagepath="./tmp2")
router.register_delivery_callback(delivery_callback)
ident = RNS.Identity()
source = router.register_delivery_identity(ident, display_name=random_names[random.randint(0,len(random_names)-1)], stamp_cost=8)
router.announce(source.hash)
RNS.log("Source announced")
print("Recipient: ", end=" ")
recipient_hexhash = input()
recipient_hash = bytes.fromhex(recipient_hexhash)
if not RNS.Transport.has_path(recipient_hash):
RNS.log("Destination is not yet known. Requesting path and waiting for announce to arrive...")
RNS.Transport.request_path(recipient_hash)
while not RNS.Transport.has_path(recipient_hash):
time.sleep(0.1)
# Recall the server identity
recipient_identity = RNS.Identity.recall(recipient_hash)
dest = RNS.Destination(recipient_identity, RNS.Destination.OUT, RNS.Destination.SINGLE, "lxmf", "delivery")
# This loop allows you to execute various actions for testing
# and experimenting with the example scripts.
while True:
# Create a message that will be sent directly to the
# destination over a Reticulum link:
lxm = LXMF.LXMessage(dest, source, random_msgs[random.randint(0,len(random_msgs)-1)],
random_titles[random.randint(0,len(random_titles)-1)],
desired_method=LXMF.LXMessage.DIRECT, include_ticket=True)
# Or, create an oppertunistic, single-packet message
# for sending without first establishing a link:
# lxm = LXMF.LXMessage(dest, source, "This is a test",
# random_titles[random.randint(0,len(random_titles)-1)],
# desired_method=LXMF.LXMessage.OPPORTUNISTIC, include_ticket=True)
# Or, try sending the message via a propagation node:
# router.set_outbound_propagation_node(bytes.fromhex("e75d9b6a69f82b48b6077cf2242d7499"))
# lxm = LXMF.LXMessage(dest, source, random_msgs[random.randint(0,len(random_msgs)-1)],
# random_titles[random.randint(0,len(random_titles)-1)],
# desired_method=LXMF.LXMessage.PROPAGATED)
# Finally dispatch the message to the LXMF message
# router, which will handle the delivery according
# to the specified message parameters and options:
router.handle_outbound(lxm)
# Wait for user input before starting over
input()

BIN
docs/paper_msg_test.png Normal file
View file

Binary file not shown.
Side by side

After

Width:  |  Height:  |  Size: 2.6 KiB

1
examples/lxmf_message_minimal.hex Normal file
View file

@ -0,0 +1 @@
000102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f303132333435363738393a3b3c3d3e3f404142434445464748494a4b4c4d4e4f505152535455565758595a5b5c5d5e5f94cb41d954fc40100000c40a54657374207469746c65c41744657465726d696e6973746963204c584d4620626f647980

1
examples/lxmf_message_stamped.hex Normal file
View file

@ -0,0 +1 @@
000102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f303132333435363738393a3b3c3d3e3f404142434445464748494a4b4c4d4e4f505152535455565758595a5b5c5d5e5f95cb41d954fc40100000c40a54657374207469746c65c41744657465726d696e6973746963204c584d4620626f647980c420a0a1a2a3a4a5a6a7a8a9aaabacadaeafb0b1b2b3b4b5b6b7b8b9babbbcbdbebf

26
examples/lxmf_message_vectors.json Normal file
View file

@ -0,0 +1,26 @@
{
"content_hex" : "44657465726d696e6973746963204c584d4620626f6479",
"description" : "Deterministic LXMessage packing vectors; signature bytes are placeholders",
"destination_hash_hex" : "000102030405060708090a0b0c0d0e0f",
"fields" : {},
"format" : "destination_hash || source_hash || signature || msgpack_payload",
"message_id_hex" : "55f497f6f2ea5d858eaff541a9fa77f9d5bdf57be75be4a94df65d1f9a7ffea3",
"minimal_lxmf_hex" : "000102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f303132333435363738393a3b3c3d3e3f404142434445464748494a4b4c4d4e4f505152535455565758595a5b5c5d5e5f94cb41d954fc40100000c40a54657374207469746c65c41744657465726d696e6973746963204c584d4620626f647980",
"payload_order" : [
"timestamp",
"title",
"content",
"fields",
"optional_stamp"
],
"payload_with_stamp_hex" : "95cb41d954fc40100000c40a54657374207469746c65c41744657465726d696e6973746963204c584d4620626f647980c420a0a1a2a3a4a5a6a7a8a9aaabacadaeafb0b1b2b3b4b5b6b7b8b9babbbcbdbebf",
"payload_without_stamp_hex" : "94cb41d954fc40100000c40a54657374207469746c65c41744657465726d696e6973746963204c584d4620626f647980",
"signature_hex" : "202122232425262728292a2b2c2d2e2f303132333435363738393a3b3c3d3e3f404142434445464748494a4b4c4d4e4f505152535455565758595a5b5c5d5e5f",
"signature_valid" : false,
"signed_part_hex" : "000102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1f94cb41d954fc40100000c40a54657374207469746c65c41744657465726d696e6973746963204c584d4620626f64798055f497f6f2ea5d858eaff541a9fa77f9d5bdf57be75be4a94df65d1f9a7ffea3",
"source_hash_hex" : "101112131415161718191a1b1c1d1e1f",
"stamp_hex" : "a0a1a2a3a4a5a6a7a8a9aaabacadaeafb0b1b2b3b4b5b6b7b8b9babbbcbdbebf",
"stamped_lxmf_hex" : "000102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f303132333435363738393a3b3c3d3e3f404142434445464748494a4b4c4d4e4f505152535455565758595a5b5c5d5e5f95cb41d954fc40100000c40a54657374207469746c65c41744657465726d696e6973746963204c584d4620626f647980c420a0a1a2a3a4a5a6a7a8a9aaabacadaeafb0b1b2b3b4b5b6b7b8b9babbbcbdbebf",
"timestamp" : 1700000000.25,
"title_hex" : "54657374207469746c65"
}

39
grep_results.txt Normal file
View file

@ -0,0 +1,39 @@
./LXMF/Utilities/lxmd.py:DEFFERED_JOBS_DELAY = 10
./LXMF/Utilities/lxmd.py:def create_default_config(configpath):
./LXMF/Utilities/lxmd.py:def apply_config():
./LXMF/Utilities/lxmd.py:def lxmf_delivery(lxm):
./LXMF/Utilities/lxmd.py:def program_setup(configdir = None, rnsconfigdir = None, run_pn = False, on_inbound = None, verbosity = 0, quietness = 0, service = False):
./LXMF/Utilities/lxmd.py:def jobs():
./LXMF/Utilities/lxmd.py:def deferred_start_jobs():
./LXMF/Utilities/lxmd.py:def _request_sync(identity, destination_hash, remote_identity, timeout=15, exit_on_fail=False):
./LXMF/Utilities/lxmd.py:def request_sync(target, remote=None, configdir=None, rnsconfigdir=None, verbosity=0, quietness=0, timeout=15, identity_path=None):
./LXMF/Utilities/lxmd.py:def _request_unpeer(identity, destination_hash, remote_identity, timeout=15, exit_on_fail=False):
./LXMF/Utilities/lxmd.py:def request_unpeer(target, remote=None, configdir=None, rnsconfigdir=None, verbosity=0, quietness=0, timeout=15, identity_path=None):
./LXMF/Utilities/lxmd.py:def query_status(identity, remote_identity=None, timeout=5, exit_on_fail=False):
./LXMF/Utilities/lxmd.py:def get_status(remote=None, configdir=None, rnsconfigdir=None, verbosity=0, quietness=0, timeout=5,
./LXMF/Utilities/lxmd.py:def _get_target_identity(remote=None, timeout=5):
./LXMF/Utilities/lxmd.py:def _remote_init(configdir=None, rnsconfigdir=None, verbosity=0, quietness=0, identity_path=None):
./LXMF/Utilities/lxmd.py:def main():
./LXMF/LXMF.py:def display_name_from_app_data(app_data=None):
./LXMF/LXMF.py:def stamp_cost_from_app_data(app_data=None):
./LXMF/LXMF.py:def compression_support_from_app_data(app_data=None):
./LXMF/LXMF.py:def pn_name_from_app_data(app_data=None):
./LXMF/LXMF.py:def pn_stamp_cost_from_app_data(app_data=None):
./LXMF/LXMF.py:def pn_announce_data_is_valid(data):
./LXMF/LXStamper.py:def worker_context(ctx, stamp_cost, workblock, message_id):
./LXMF/LXStamper.py:def stamp_workblock(material, expand_rounds=WORKBLOCK_EXPAND_ROUNDS):
./LXMF/LXStamper.py:def stamp_value(workblock, stamp):
./LXMF/LXStamper.py:def stamp_valid(stamp, target_cost, workblock):
./LXMF/LXStamper.py:def validate_peering_key(peering_id, peering_key, target_cost):
./LXMF/LXStamper.py:def validate_pn_stamp(transient_data, target_cost):
./LXMF/LXStamper.py:def validate_pn_stamps_job_simple(transient_list, target_cost):
./LXMF/LXStamper.py:def validate_pn_stamps_job_multip(transient_list, target_cost):
./LXMF/LXStamper.py:def validate_pn_stamps(transient_list, target_cost):
./LXMF/LXStamper.py:def generate_stamp(message_id, stamp_cost, expand_rounds=WORKBLOCK_EXPAND_ROUNDS):
./LXMF/LXStamper.py:def cancel_work(message_id):
./LXMF/LXStamper.py:def job_simple(stamp_cost, workblock, message_id):
./LXMF/LXStamper.py:def job_linux_managed(stamp_cost, workblock, message_id):
./LXMF/LXStamper.py:def job_linux(stamp_cost, workblock, message_id):
./LXMF/LXStamper.py:def job_android(stamp_cost, workblock, message_id):
./docs/example_receiver.py:def delivery_callback(message):
./docs/example_sender.py:def delivery_callback(message):

2
requirements.txt Normal file
View file

@ -0,0 +1,2 @@
qrcode>=7.4.2
rns>=1.0.0

36
setup.py Normal file
View file

@ -0,0 +1,36 @@
import sys
import setuptools
with open("README.md", "r") as fh:
long_description = fh.read()
exec(open("LXMF/_version.py", "r").read())
if "--getversion" in sys.argv:
print(__version__, end="")
exit(0)
setuptools.setup(
name="lxmf",
version=__version__,
author="Mark Qvist",
author_email="mark@unsigned.io",
description="Lightweight Extensible Message Format for Reticulum",
long_description=long_description,
long_description_content_type="text/markdown",
url="https://github.com/markqvist/lxmf",
packages=["LXMF", "LXMF.Utilities"],
license="Reticulum License",
license_files = ("LICENSE"),
classifiers=[
"Programming Language :: Python :: 3",
"Operating System :: OS Independent",
],
entry_points= {
'console_scripts': [
'lxmd=LXMF.Utilities.lxmd:main',
]
},
install_requires=["rns>=1.3.5"],
python_requires=">=3.7",
)

181
tools/LXMF/MessagePack.pm Normal file
View file

@ -0,0 +1,181 @@
package LXMF::MessagePack;
use strict;
use warnings;
use Exporter 'import';
our @EXPORT_OK = qw(
pack_array
pack_bin
pack_float64
pack_map
parse_item
);
sub pack_float64 {
my ($value) = @_;
return "\xcb" . pack("d>", $value);
}
sub pack_bin {
my ($value) = @_;
my $length = length($value);
return "\xc4" . pack("C", $length) . $value if $length <= 0xff;
return "\xc5" . pack("n", $length) . $value if $length <= 0xffff;
return "\xc6" . pack("N", $length) . $value;
}
sub pack_array {
my (@items) = @_;
my $length = scalar @items;
my $header;
if ($length <= 15) {
$header = pack("C", 0x90 | $length);
}
elsif ($length <= 0xffff) {
$header = "\xdc" . pack("n", $length);
}
else {
$header = "\xdd" . pack("N", $length);
}
return $header . join("", @items);
}
sub pack_map {
my (@entries) = @_;
die "pack_map requires key/value byte-string pairs\n" if @entries % 2;
my $length = @entries / 2;
my $header;
if ($length <= 15) {
$header = pack("C", 0x80 | $length);
}
elsif ($length <= 0xffff) {
$header = "\xde" . pack("n", $length);
}
else {
$header = "\xdf" . pack("N", $length);
}
return $header . join("", @entries);
}
sub _take {
my ($bytes, $offset_ref, $length) = @_;
die "truncated MessagePack item\n" if $$offset_ref + $length > length($bytes);
my $value = substr($bytes, $$offset_ref, $length);
$$offset_ref += $length;
return $value;
}
sub _parse_length {
my ($bytes, $offset_ref, $width) = @_;
my $raw = _take($bytes, $offset_ref, $width);
return unpack("C", $raw) if $width == 1;
return unpack("n", $raw) if $width == 2;
return unpack("N", $raw);
}
sub parse_item {
my ($bytes, $offset_ref) = @_;
$offset_ref //= \my $offset;
$$offset_ref //= 0;
my $start = $$offset_ref;
my $code = unpack("C", _take($bytes, $offset_ref, 1));
my $item = { code => $code };
if ($code <= 0x7f) {
@$item{qw(type value)} = ("integer", $code);
}
elsif ($code >= 0xe0) {
@$item{qw(type value)} = ("integer", $code - 256);
}
elsif (($code & 0xf0) == 0x90) {
_parse_array($bytes, $offset_ref, $item, $code & 0x0f);
}
elsif (($code & 0xf0) == 0x80) {
_parse_map($bytes, $offset_ref, $item, $code & 0x0f);
}
elsif (($code & 0xe0) == 0xa0) {
my $length = $code & 0x1f;
@$item{qw(type value)} = ("string", _take($bytes, $offset_ref, $length));
}
elsif ($code == 0xc0) {
$item->{type} = "nil";
$item->{value} = undef;
}
elsif ($code == 0xc2 || $code == 0xc3) {
@$item{qw(type value)} = ("boolean", $code == 0xc3 ? 1 : 0);
}
elsif ($code == 0xca) {
@$item{qw(type value)} = ("float32", unpack("f>", _take($bytes, $offset_ref, 4)));
}
elsif ($code == 0xcb) {
@$item{qw(type value)} = ("float64", unpack("d>", _take($bytes, $offset_ref, 8)));
}
elsif ($code >= 0xc4 && $code <= 0xc6) {
my $width = 1 << ($code - 0xc4);
my $length = _parse_length($bytes, $offset_ref, $width);
@$item{qw(type value)} = ("binary", _take($bytes, $offset_ref, $length));
}
elsif ($code >= 0xcc && $code <= 0xcf) {
my @formats = ("C", "n", "N", "Q>");
my @widths = (1, 2, 4, 8);
my $index = $code - 0xcc;
@$item{qw(type value)} = (
"integer",
unpack($formats[$index], _take($bytes, $offset_ref, $widths[$index])),
);
}
elsif ($code >= 0xd0 && $code <= 0xd3) {
my @formats = ("c", "s>", "l>", "q>");
my @widths = (1, 2, 4, 8);
my $index = $code - 0xd0;
@$item{qw(type value)} = (
"integer",
unpack($formats[$index], _take($bytes, $offset_ref, $widths[$index])),
);
}
elsif ($code == 0xd9 || $code == 0xda || $code == 0xdb) {
my $width = 1 << ($code - 0xd9);
my $length = _parse_length($bytes, $offset_ref, $width);
@$item{qw(type value)} = ("string", _take($bytes, $offset_ref, $length));
}
elsif ($code == 0xdc || $code == 0xdd) {
my $width = $code == 0xdc ? 2 : 4;
_parse_array($bytes, $offset_ref, $item, _parse_length($bytes, $offset_ref, $width));
}
elsif ($code == 0xde || $code == 0xdf) {
my $width = $code == 0xde ? 2 : 4;
_parse_map($bytes, $offset_ref, $item, _parse_length($bytes, $offset_ref, $width));
}
else {
die sprintf("unsupported MessagePack code 0x%02x at offset %d\n", $code, $start);
}
$item->{start} = $start;
$item->{end} = $$offset_ref;
$item->{raw} = substr($bytes, $start, $$offset_ref - $start);
return $item;
}
sub _parse_array {
my ($bytes, $offset_ref, $item, $length) = @_;
$item->{type} = "array";
$item->{items} = [map { parse_item($bytes, $offset_ref) } 1 .. $length];
}
sub _parse_map {
my ($bytes, $offset_ref, $item, $length) = @_;
$item->{type} = "map";
$item->{entries} = [
map { [parse_item($bytes, $offset_ref), parse_item($bytes, $offset_ref)] } 1 .. $length
];
}
1;

32
tools/README.md Normal file
View file

@ -0,0 +1,32 @@
# LXMessage Test-Vector Tooling
The scripts in this directory verify the implementation-derived LXMessage
packing behavior described in `Code_Response_1.md`.
They use only Perl core modules. The local `LXMF::MessagePack` module implements
the MessagePack subset needed to generate the vectors and parse their complete
payloads.
Generate deterministic vectors:
```sh
perl tools/make_test_vectors.pl
```
Verify the vectors:
```sh
perl tools/verify_examples.pl
```
The vectors confirm:
- the 16-byte destination hash, 16-byte source hash, and 64-byte signature
positions;
- payload order `[timestamp, title, content, fields, optional_stamp]`;
- float64 timestamp and binary title/content/stamp encodings;
- the message ID calculation;
- exclusion of the optional stamp from the message ID and signature input.
The signature bytes are deterministic placeholders. These vectors do not claim
to verify Ed25519 signing or signature validation.

76
tools/make_test_vectors.pl Executable file
View file

@ -0,0 +1,76 @@
#!/usr/bin/env perl
use strict;
use warnings;
use Digest::SHA qw(sha256);
use FindBin qw($Bin);
use File::Path qw(make_path);
use JSON::PP;
use lib "$Bin";
use LXMF::MessagePack qw(pack_array pack_bin pack_float64 pack_map);
my $output_dir = "$Bin/../examples";
make_path($output_dir);
my $destination_hash = pack("C*", 0x00 .. 0x0f);
my $source_hash = pack("C*", 0x10 .. 0x1f);
my $signature = pack("C*", 0x20 .. 0x5f);
my $stamp = pack("C*", 0xa0 .. 0xbf);
my $timestamp = 1_700_000_000.25;
my $title = "Test title";
my $content = "Deterministic LXMF body";
my $fields = pack_map();
my @required_items = (
pack_float64($timestamp),
pack_bin($title),
pack_bin($content),
$fields,
);
my $payload = pack_array(@required_items);
my $stamped_payload = pack_array(@required_items, pack_bin($stamp));
my $message_id = sha256($destination_hash . $source_hash . $payload);
my $signed_part = $destination_hash . $source_hash . $payload . $message_id;
my $minimal = $destination_hash . $source_hash . $signature . $payload;
my $stamped = $destination_hash . $source_hash . $signature . $stamped_payload;
write_hex("$output_dir/lxmf_message_minimal.hex", $minimal);
write_hex("$output_dir/lxmf_message_stamped.hex", $stamped);
my $manifest = {
description => "Deterministic LXMessage packing vectors; signature bytes are placeholders",
format => "destination_hash || source_hash || signature || msgpack_payload",
payload_order => [qw(timestamp title content fields optional_stamp)],
destination_hash_hex => unpack("H*", $destination_hash),
source_hash_hex => unpack("H*", $source_hash),
signature_hex => unpack("H*", $signature),
signature_valid => JSON::PP::false,
timestamp => $timestamp,
title_hex => unpack("H*", $title),
content_hex => unpack("H*", $content),
fields => {},
stamp_hex => unpack("H*", $stamp),
payload_without_stamp_hex => unpack("H*", $payload),
payload_with_stamp_hex => unpack("H*", $stamped_payload),
message_id_hex => unpack("H*", $message_id),
signed_part_hex => unpack("H*", $signed_part),
minimal_lxmf_hex => unpack("H*", $minimal),
stamped_lxmf_hex => unpack("H*", $stamped),
};
open my $json_fh, ">:raw", "$output_dir/lxmf_message_vectors.json"
or die "cannot write vector manifest: $!\n";
print {$json_fh} JSON::PP->new->canonical->pretty->encode($manifest);
close $json_fh;
print "Generated deterministic LXMessage vectors in $output_dir\n";
sub write_hex {
my ($path, $bytes) = @_;
open my $fh, ">:raw", $path or die "cannot write $path: $!\n";
print {$fh} unpack("H*", $bytes), "\n";
close $fh;
}

88
tools/verify_examples.pl Executable file
View file

@ -0,0 +1,88 @@
#!/usr/bin/env perl
use strict;
use warnings;
use Digest::SHA qw(sha256);
use FindBin qw($Bin);
use JSON::PP;
use lib "$Bin";
use LXMF::MessagePack qw(pack_array parse_item);
my $examples_dir = "$Bin/../examples";
my $manifest = read_json("$examples_dir/lxmf_message_vectors.json");
my $minimal = read_hex("$examples_dir/lxmf_message_minimal.hex");
my $stamped = read_hex("$examples_dir/lxmf_message_stamped.hex");
verify_lxmessage("minimal", $minimal, $manifest, 4);
verify_lxmessage("stamped", $stamped, $manifest, 5);
check(unpack("H*", $minimal) eq $manifest->{minimal_lxmf_hex}, "minimal full bytes match manifest");
check(unpack("H*", $stamped) eq $manifest->{stamped_lxmf_hex}, "stamped full bytes match manifest");
print "All LXMessage test-vector checks passed.\n";
sub verify_lxmessage {
my ($name, $bytes, $expected, $expected_count) = @_;
check(length($bytes) > 96, "$name vector has fixed prefix and payload");
my $destination_hash = substr($bytes, 0, 16);
my $source_hash = substr($bytes, 16, 16);
my $signature = substr($bytes, 32, 64);
my $packed_payload = substr($bytes, 96);
check(unpack("H*", $destination_hash) eq $expected->{destination_hash_hex}, "$name destination hash");
check(unpack("H*", $source_hash) eq $expected->{source_hash_hex}, "$name source hash");
check(unpack("H*", $signature) eq $expected->{signature_hex}, "$name signature position and length");
my $offset = 0;
my $payload = parse_item($packed_payload, \$offset);
check($offset == length($packed_payload), "$name payload consumes all remaining bytes");
check($payload->{type} eq "array", "$name payload is a MessagePack array");
check(@{$payload->{items}} == $expected_count, "$name payload has $expected_count entries");
my @items = @{$payload->{items}};
check($items[0]{type} eq "float64", "$name payload[0] timestamp is float64");
check($items[0]{value} == $expected->{timestamp}, "$name payload[0] timestamp value");
check($items[1]{type} eq "binary", "$name payload[1] title is binary");
check(unpack("H*", $items[1]{value}) eq $expected->{title_hex}, "$name payload[1] title value");
check($items[2]{type} eq "binary", "$name payload[2] content is binary");
check(unpack("H*", $items[2]{value}) eq $expected->{content_hex}, "$name payload[2] content value");
check($items[3]{type} eq "map", "$name payload[3] fields is a map");
check(@{$items[3]{entries}} == 0, "$name payload[3] fields map is empty");
my $unstamped_payload = pack_array(map { $_->{raw} } @items[0 .. 3]);
my $message_id = sha256($destination_hash . $source_hash . $unstamped_payload);
my $signed_part = $destination_hash . $source_hash . $unstamped_payload . $message_id;
check(unpack("H*", $unstamped_payload) eq $expected->{payload_without_stamp_hex}, "$name canonical unstamped payload");
check(unpack("H*", $message_id) eq $expected->{message_id_hex}, "$name message ID excludes optional stamp");
check(unpack("H*", $signed_part) eq $expected->{signed_part_hex}, "$name signature input excludes optional stamp");
if ($expected_count == 5) {
check($items[4]{type} eq "binary", "$name payload[4] stamp is binary");
check(unpack("H*", $items[4]{value}) eq $expected->{stamp_hex}, "$name payload[4] stamp value");
}
}
sub read_json {
my ($path) = @_;
open my $fh, "<:raw", $path or die "cannot read $path: $!\n";
local $/;
return JSON::PP->new->decode(<$fh>);
}
sub read_hex {
my ($path) = @_;
open my $fh, "<:raw", $path or die "cannot read $path: $!\n";
local $/;
my $hex = <$fh>;
$hex =~ s/\s+//g;
die "$path contains non-hexadecimal data\n" if $hex =~ /[^0-9a-fA-F]/;
return pack("H*", $hex);
}
sub check {
my ($condition, $description) = @_;
die "not ok - $description\n" unless $condition;
print "ok - $description\n";
}