On 27/10/26 at 08:36 -0600, Gunnar Wolf wrote: > gregor herrmann dijo [Tue, Oct 27, 2026 at 04:39:44AM +0100]: > > |This is an LLM based review using a Debian Inference based workflow. It was > > |not written by a member of the DFSG team and has not been checked by one. > > It > > |is not a decision on the package: it may be wrong, and it may have missed > > |things. > > > > > I do not object in principle to the use of a LLM, but the DFSG team > > > needs to take full responsibility for its output: this kind of > > > disclaimer is not acceptable. > > > > Same here: I don't object to the DFSG team experimenting with LLMs for > > package reviews but dumping them without any human inspections onto > > uploaders seems both disrespectful and contrary to the GR result to me. > > I concur with this feeling that I've seen repeated throughout the thread. I > will not hide that I am an LLM user for several bits of my work, and I > disclosed with my vote in the GR¹ that I am not against using LLM output > for work for Debian. However, I do not think auto-generating a > non-human-mediated LLM response and sending it to any uploader is just > wrong. > > ¹ > https://gwolf.org/2026/08/as-far-as-llms-go-in-debian-i-think-that-936241857.html > > My veeling is that the DFSG team are free and welcome to integrate LLMs > into their workflow, but whatever output they share with fellow developers > should be first and foremost reviewed by the team, and only later > communicated to the uploader.
Since the DFSG team is using a tool to analyze packages at upload time anyway, I think that it makes sense to provide that report to maintainers, so that they can react without waiting for a human review, similarly to how maintainers react to CI results. I don't see the point in hiding those reports. If some maintainers prefer to wait until another human confirmed the diagnosis, that's fine, but why penalize those who welcome early feedback? Lucas
