-
xmr-pr
-
xmr-pr
bsandham opened pull request #11026: simplewallet: show default value in Y/N prompts
-
xmr-pr
-
xmr-pr
maogo opened issue #11027: why wallet cannot refresh by public rpc
-
xmr-pr
-
xmr-pr
tobtoht opened pull request #11028: depends: boost: don't embed absolute paths in cmake config
-
xmr-pr
-
selsta
.merge+ 10989 10988
-
xmr-pr
Added
-
xmr-pr
jpk68 opened pull request #11029: cryptonote_protocol: avoid copying block span in add_blocks
-
xmr-pr
-
xmr-pr
jpk68 opened pull request #11030: wallet2: fix faulty bounds check
-
xmr-pr
-
xmr-pr
tobtoht opened pull request #11031: depends: native_protobuf: don't include debug symbols in protoc binary
-
xmr-pr
-
selsta
tevador: what is the goal of the policy? is it a document that can be cited if a PR is closed and someone complains, or is it something that needs to be followed to the letter?
-
selsta
I mean apart from the copyright stuff of course
-
selsta
but if I as maintainer see something that is clearly useful but it looks like it's written by a bot do I have to close it?
-
tevador
Hopefully, the AI agent will read the policy and act accordingly.
-
br-m
<jpk68:matrix.org> Having to close LLM-authored PRs that fix legit bugs will be nothing but a burden
-
br-m
<jpk68:matrix.org> Why not just keep doing things the way we've already been doing them?
-
selsta
I do not want to play LLM detective, and I do not want to reject legit useful changes.
-
tevador
I would take it as a gate to stop the flood of AI submissions. If an AI agent submits a valid issue, I don't think it needs to be closed.
-
selsta
the current suspected AI agents that submit PRs are not an annoyance in my opinion, and so far also haven't produced a flood
-
selsta
3 months ago I had a different opinion but there was a change in quality
-
tevador
Can you guarantee there won't be another change of quality for the worse?
-
selsta
the same happened on HackerOne, previously AI reports were mostly hallucination while around 4 months ago it changed to legit vulnerability findings
-
br-m
<slowbeardigger:matrix.org> tevador: no one can guarantee that
-
br-m
<slowbeardigger:matrix.org> sure, AI can do crappy stuff, but it is responsibility of the ones who use it
-
br-m
<slowbeardigger:matrix.org> i use it, i am transparent and i take full responsibility for the things i do
-
tevador
-
br-m
<ofrnxmr:xmr.mx> I think the question is about allowing the codebase to be damaged by bloated, unmaintainable garbage
-
selsta
in general I'm okay with the current version of the document
-
selsta
I think there is a large gap between using AI to reproduce another projects copyrighted work and a one line but fix but I assume we don't need to detail that in that document
-
xmr-pr
jpk68 opened pull request #11032: doxyfile: remove comments, improve build performance
-
xmr-pr
-
br-m
<jpk68:matrix.org> Wow, it took less than half an hour this time!
-
br-m
<jpk68:matrix.org> :)