Hi Phil, Am Mon, May 27, 2024 at 12:15:47AM +0200 schrieb Philip Hands: > This is mostly because I found that I wasn't able to devote the time > required to test things to my satisfaction when my first daughter came > along,
So we will see you back in the team once your youngest child is around 10 or so. ;-P > as I'd be distracted before I completed my tests, so I decided to > do something about automating testing, and I've been down that rabbit > hole ever since. When I'm working on that, I'm pretty happy. Good! Keep on working with that. ;-) > I'd say that that work is now bearing some fruit, finally. I had > originally hoped that I'd then be able to put more effort into D-I > itself, but I suspect that maintaining openQA and the Salsa pipeline > stuff may continue to eat a fair amount of my time. Sounds like a pretty interesting project. Thanks for keeping me informed about this. > > - Do you consider the workload of your team equally shared amongst its > > members and who actually is considered a team member? (I added some > > persons in CC who have recently answered to questions on the mailing > > list.) > > My contributions are pretty-much background noise recently, so I guess > that means that the load is very unequal if you were including me in the > stats. > > Cyril has been responsible for keeping D-I viable in recent times, and > Holger also does _loads_ of (mostly translation related) work too. I highly appreciate all responses from Debian Boot team which are the most extensive so far. > > - Do you have some strategy to gather new contributors for your team? > > One of my intentions with the salsa/openQA work is that I'm trying to > make it possible for people to make simple changes to bits of D-I and > have them receive feedback about whether the result is an improvement. > > Hopefully that will lower the bar to new people contributing. Very nice contribution! > > - Can you give some individual estimation how many hours per week you > > are working on your tasks in youre team? Does this fit the amount of > > time you can really afford for this task? > > My work on D-I is pretty sporadic, because I generally pick some small > thing in D-I to use as a test of the current salsa/openqa setup, and > then spend significantly more time sorting out some new wrinkle that's > revealed in the salsa and/or openqa setup by this new example. > > Often this means that by the time I've finished, someone else has > already dealt with the original bug/patch in D-I. I'm not sure to what > extent that counts as D-I work, but I'm happy with the time I spend on > it. OK > > - I recently had some discussion on Chemnitzer Linuxtage what might > > be the reason for derivatives to write their own installers. While > > I'm personally perfectly happy with the way I can install Debian I'm > > somehow wondering why others are spending time into a problem we > > are considering "solved" and whether we can learn something from this, > > I quite like it as it is, but I'm sure many would not find the installer > particularly pretty, and it is quite hard to work on (being in busybox > shell, and lacking popular things like python), and I personally have no > idea how easy/possible it is to e.g. change its branding (if a > downstream wanted to do that). > > If one doesn't care about installing on our minority architectures, then > it's possible to do something that's much easier to work on by booting a > live image. One can then have something that'll ask all the questions > up-front (especially if one is opinionated about what should be on the > resulting system), and then apply that to the system without further > interaction. Would you agree to the statement I'm drawing from past discussion: Debian has to care for working installer on all architectures. Debian derivatives do not have this requirement and prefer other pretty / fancy / brandable ways over the Debian one? > Some arm64 things certainly can be installed with D-I, because I have > openQA workers running on altra.debian.net testing D-I installs, but I > don't know that much about the details. OK > > - Can I do anything for you? > > I'm currently looking into the options that might be worth exploring for > getting more openqa-workers running. I suppose at some point that might > involve asking for funds to be spent, but I'm not at that stage yet. If you have some ideas whom to ask and reasons to motivate them for X amount to spent I'd happily support you in this. > It probably wouldn't harm to offer some funding to osuosl, because they > let us use their systems for various things and making sure that they > are sustainable would be wise. (that's who host most of what I'm running > openqa on at present, and they also host jenkins and reproducible things > AFAIK) If you want to go into more detail (if you consider in private might be better that's fine) I can talk with treasurers about the options we have. I keep on feeling as greenhorn DPL and have no good idea what options we finally have. But thanks in any case for bringing this up. Kind regards Andreas. -- https://fam-tille.de
