Repository navigation
Conversation
An AI session run through cMeta (task run-ai of cmeta-aops) keeps its memory, skills, conversations and logs inside the artifact it works on, in a folder named !AI. Those belong to whoever ran the session, not to this repository: the folder is ignored (written \!AI/, since a leading "!" negates a pattern), and "git add -f" still adds a file on purpose. The note on session-provenance files no longer names one person's hook. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Signed-off-by: Grigori Fursin <gfursin@gmail.com>
The authorship badge and a paragraph: cMeta was architected and written from scratch by its author and is extended since 2025 together with AI agents that run through cMeta itself, every artifact recording who made it (the generator field). Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Signed-off-by: Grigori Fursin <gfursin@gmail.com>
…AI, powered by cMeta and Agentic Ops Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Signed-off-by: Grigori Fursin <gfursin@gmail.com>
Troubleshooting: what App Control blocks (uv's Python, uv's shims, Git for Windows' DLLs, every compiled program), that turning it off is the fix on a development machine, and the signed-only route for a PC that keeps it on: Python from the Microsoft Store with --python, repositories fetched as archives with curl.exe and cx repo unzip. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Signed-off-by: Grigori Fursin <gfursin@gmail.com>
…nguage, never an error The text mode of subprocess decoded with the locale's encoding, strictly: a byte it does not define (UTF-8 from git on a French Windows, a CP850 letter from a cmd-line tool) raised inside run() and the command came back with exit code -1 and no output. Output is now captured as bytes and decoded by decode_output(): UTF-8 first (a byte-order mark dropped), then the machine's own code pages (the console's OEM page and the ANSI page on Windows, the locale's encoding elsewhere), the whole output first and line by line when tools of different encodings share one output, replacement characters as the last resort. Line endings normalised as before. The same for the timeout paths and for pip's output in packages.py. Tests for the decoder, stand-ins for a French Windows and a Latin-1 Linux, and a child printing mixed bytes through both paths of run(). Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Signed-off-by: Grigori Fursin <gfursin@gmail.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Signed-off-by: Grigori Fursin <gfursin@gmail.com>
…-R locale, not as replacements Seen in Debian containers with fr_FR.ISO-8859-1, de_DE.ISO-8859-1 and ru_RU.KOI8-R: the two bytes are valid there, so the machine's own page decodes them. The test now asks for text of the right length in every case, which is what the decoder promises. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Signed-off-by: Grigori Fursin <gfursin@gmail.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
0.33.1
utils.sys.runcaptured output in the text mode ofsubprocess, which decodes with the locale's encoding and strictly: a byte that encoding does not define (UTF-8 from git or Python on a French Windows, Cyrillic, a CP850 letter from a cmd-line tool) raised insiderun()and the command came back with exit code -1 and no output. The output is now captured as bytes and decoded by the newutils.sys.decode_output: UTF-8 first (a byte-order mark dropped), then the machine's own code pages (the console's OEM page and the ANSI page on Windows, the locale's encoding elsewhere), the whole output first and line by line when tools of different encodings share one output, and UTF-8 with replacement characters as the last resort. Line endings are normalised to\nas before;NonestaysNone. The same decoding applies to the timeout paths and to pip's output inpackages.py.decode_output(data, encoding=...)lets a caller name an encoding it knows.Tested: the engine suite (534) on Windows; CI on Linux, Windows and macOS with Python 3.9 and 3.14; the decoder tests under C, POSIX, C.UTF-8, fr_FR.ISO-8859-1, de_DE.ISO-8859-1, ja_JP.EUC-JP and ru_RU.KOI8-R in Python 3.9 and 3.14 containers, and on a Mac under its fr_FR.ISO8859-1 locale.
See CHANGELOG.md.
🤖 Generated with Claude Code