Re: [PATCH] docs: outline bug expectations wrt automated tools / AI agents
On Mon, Jun 09, 2025 at 03:06:00PM +0200, Michal Prívozník wrote:
On 6/6/25 10:52, Daniel P. Berrangé via Devel wrote:
> From: Daniel P. Berrangé <berrange(a)redhat.com>
>
> Bug reports from automated tools and AI agents are time consuming
Maybe orthogonal topic, but should we also discourage (if not ban)
people from sending patches generated by AI tools? For instance Gentoo
has done so [1] and their foremost reason is possible licensing problem
/ copyright violation.
I've seen some people asking some language models what does this or that
internal function of ours do (when investigating code). And while I
might have preferences on that, it's probably okay. But letting LLMs
generate pieces of code that was trained on who-knows-what might pose
problem once such code is merged.
OTOH - we have Developer Certificate of Origin which should mean that
the author can send given patch.
In QEMU I put forward the viewpoint that contributing under the DCO
is incompatible with the use of common AI content generators today,
given the inability to satisfy any of the DCO clauses for the AI
portion of the patch.
https://lists.nongnu.org/archive/html/qemu-devel/2025-06/msg00453.html
In theory we shouldn't need to state anything, but I expect people
wouldn't be thinking of the implications of the DCO rules when they
decide to use AI tools, hence the suggestion to document it in QEMU.
The QEMU proposal isn't merged yet, but we should an eye on it, as
the position would apply equally to libvirt as QEMU.
With regards,
Daniel
--
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org -o- https://fstop138.berrange.com :|
|: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|