| Message ID | CAA_UwzLJLHFWmiAGxvkwEO0PHzYbQNH834Fm2wv156U2BjTDiw@mail.gmail.com |
|---|---|
| Headers | show |
| Series |
|
| Related | show |
Hi Ray On Thu, Oct 22, 2026 at 01:29:28PM -0400, Ray Strode wrote: > Hi, > > This series adds utils/get_maintainer.py to help contributors find > maintainers and reviewers for their patches, and documents its use in > the contribution guide. > > The script accepts patches or repository paths and uses Git history to > rank review and acknowledgment credits ahead of authorship. It follows > the kernel script's input conventions and supports plain recipient > output for git send-email --cc-cmd. I think we should first formalize if we want maintainers or not in libcamera. I personally think we have reached a size where it would make sense, but that's a discussion larger than just me. However this patch promotes as "maintainer" how has the most contributions on a certain file, something that doesn't imply the person is still active or even interested in doing maintainership. What prompted you to address this issue as your first libcamera contribution ? > > --Ray > > Ray Strode (2): > utils: Find maintainers from Git history > Documentation: contributing: Document the maintainer lookup script
Hi, On Fri, Oct 23, 2026 at 10:15 AM Jacopo Mondi <jacopo.mondi@ideasonboard.com> wrote: > I think we should first formalize if we want maintainers or not in > libcamera. I personally think we have reached a size where it would > make sense, but that's a discussion larger than just me. Sure, just helping out... If it's not necessary we can just ignore this thread (and the v2) > However this patch promotes as "maintainer" how has the most > contributions on a certain file, something that doesn't imply the > person is still active or even interested in doing maintainership. > > What prompted you to address this issue as your first libcamera > contribution ? Well, it's not really my first contribution. I've done several on the list the last few days, and have an update I'm planning to send to the list shortly, but the impetus for this particular patchset was this message: https://lists.libcamera.org/pipermail/libcamera-devel/2026-October/062645.html The heuristics it follows are cribbed from the script the above mail mentioned wanting to reproduce. I think doing things dynamically is better than keeping a hardcoded list because it's more likely to involve people who are currently active. At the end of the day it's just about who gets CC'd on a mail, not who has authority, so the heuristics don't need to be perfect. You have a good point that there's no way for someone who no longer wants to receive libcamera related mails to opt out. We could add support for .get_maintainer.ignore like get_maintainer.pl does, but I guess we should settle whether we want the script at all first before we iterate on it. I think getting Laurent's and Kieran's take on this would be useful. Note I feel no strong desire to get thie script in, it was just a thing to take or leave. --Ray
Quoting Ray Strode (2026-10-23 15:47:08) > Hi, > > On Fri, Oct 23, 2026 at 10:15 AM Jacopo Mondi > <jacopo.mondi@ideasonboard.com> wrote: > > I think we should first formalize if we want maintainers or not in > > libcamera. I personally think we have reached a size where it would > > make sense, but that's a discussion larger than just me. > > Sure, just helping out... If it's not necessary we can just ignore this thread > (and the v2) > > > However this patch promotes as "maintainer" how has the most > > contributions on a certain file, something that doesn't imply the > > person is still active or even interested in doing maintainership. > > > > What prompted you to address this issue as your first libcamera > > contribution ? > > Well, it's not really my first contribution. I've done several on the > list the last few > days, and have an update I'm planning to send to the list shortly, but > the impetus for > this particular patchset was this message: > > https://lists.libcamera.org/pipermail/libcamera-devel/2026-October/062645.html > > The heuristics it follows are cribbed from the script the above mail mentioned > wanting to reproduce. > > I think doing things dynamically is better than keeping a hardcoded > list because it's > more likely to involve people who are currently active. At the end of > the day it's just > about who gets CC'd on a mail, not who has authority, so the > heuristics don't need to > be perfect. > > You have a good point that there's no way for someone who no longer wants to > receive libcamera related mails to opt out. We could add support for > .get_maintainer.ignore like get_maintainer.pl does, but I guess we should settle > whether we want the script at all first before we iterate on it. > > I think getting Laurent's and Kieran's take on this would be useful. > > Note I feel no strong desire to get thie script in, it was just a > thing to take or leave. We discussed this in our weekly. I think having some way to convey who is responsible for specific areas of the code base is useful, but Laurent doesn't want that to be from the git log, so we'll have to make something more like from a MAINTAINERS file or such like the kernel does to be able to explicitly let people decide/state that they are looking after particular code, rather than just by association of having worked on a file in the past. -- Regards Kieran > > --Ray