This is an old thread and normally I don’t dig these up, but I figured I’d throw in my 2 cents. I am a maintainer and I’m the guy behind most of the day to day infrastructure changes. Some of this is my own opinion and some of it isn’t, you can reason it out well enough.
@oliver The templates in the forum are a disaster zone. One of my first questions is whether or not people have a POS template from the forum installed as they have all the same problems as the AUR. I generally don’t support such systems unless there is a serious issue that has been discovered in some actual Void system which is triggered by the package that is not currently in tree. As far as putting the templates on the Wiki, that will get you banned for spam.
@Erin We check for build and that they don’t break anything else. There is a level of expectation that the maintainers know what they’re doing. The advantage of being in-repo instead of some other place is that then they get A) hosted somewhere, and B) rebuilt for updates when dependencies are updated. As far as other people setting up a website that hosts the templates/separate repo, I think most people dramatically underestimate how much effort that is. It would also be very aggressively discouraged by the Void team, so I personally wouldn’t expect it to get very far.
@masato Yes, hair goes on end, see my earlier comment. The AUR is a disaster zone for reliability and troubleshooting. Arch is, aside from weirdness with pacman, very stable if you use only the base repositories. If people want Arch without systemd, I encourage them to do so, there are a handful of arch derived distributions with do not use systemd as init. Void is Void, its maintained by a very small team, and right now we’re happy with the direction its going.
@paoloschi While I do not recognize you or any commits you might have made, I do recognize cr6git and know that templates bearing that submit ID are generally quite safe to yolomerge. Perhaps you shouldn’t be so quick to disparage people who are working torwards what the goal of the project is? While cr6git may not be maintaining all of the packages submitted, 90% of the battle is submitting them at all. I would say that 85% of the packages in Void are maintained by ~25 people. It takes me 10 minutes to update something versus perhaps a few hours to set it up from scratch.
@cr6 Keep up the good work.
@hralgmir I always appreciate when people bring up history. If you scroll back through long deserted list archives, you’ll find that the AUR was a highly political thing even when it started. Its a unique thing that has to have the culture to support it. Void does not have this culture. I cannot speak to ABS development, but I will certainly agree that Arch’s explosive growth drove a lot of things. One thing that’s important to note is that we also have an informal concept of Trusted Users, in the form of people who over time the core maintainers start to recognize. These are people that if they approve a PR I know I can generally merge it safely. Its not quite the same thing since we have a lot more security than Arch does and our security model is fundamentally incompatible with an AUR model, but these individuals fulfill a similar role.
As far as the large number of PRs in limbo, here’s kind of what’s going on:
A non-trivial number of Void’s core leadership work on a calendar that is roughly pegged to a US academic calendar. This means that since its the start of September, a lot of people are under high load at the moment. Some patience is appreciated and review will start up again at full speed soon, perhaps in about 2 weeks time. We’re also taking a step back from adding more packages right now to evaluate some infrastructure refactors to try and build more packages faster and with less wasted effort.
repo3 was recently rebuilt, this ate almost all of my time for a week. In addition I’ve been cleaning up the mklive system and getting ready to automate the image release process, this takes time and so I haven’t been merging PRs. @Vaelatern has been working with me on some of the infrastructure refactor and so that’s been time that can’t be used for merging things. We’re moving along and PR review is happening, but there are some other things in the queue right now that are way higher priority than just updating and accepting new packages.
There are also some really outdated PRs that we have to keep an eye on to see about changes that might happen. I’m intending to look at running a filter over void-packages and closing ancient PRs and ancient issues to get the number of PRs and Issues to reflect the current state of Void. A 3 year old Issue for gnome is likely no longer relevant and we have too many of those right now. If you know of such ancient PRs/Issues or you own an ancient PR or Issue, please consider closing it or reaching out on IRC directly.
Void is not Arch. We do not strive to be, we do not intend to be. The AUR is one of the things that makes Arch special and unique. For us its the unified build system of XBPS and xbps-src. We are built on much the same philosophy of things like NetBSD, Suckless, and some of the security ideals of OpenBSD. We aren’t just a ‘systemd-free’ distro and in fact that’s a side-effect of supporting musl, not necessarily a design choice on ideological grounds (go far enough back in the history and you’ll find that Void booted with systemd, arguably before Arch did; some of the infrastructure is even still around for this).
The goal of Void was never to compete with Arch or to be the next Ubuntu. Its a system that the maintainers like working on because of the way it is designed and because we find it genuinely fun and engaging to work on. I always appreciate when I see people using Void, but Void has a fairly narrow vision, if that vision does not align with what you want to do, I strongly encourage you to check out distrowatch.com and see what else is out there. There’s a lot of amazing distros built up by teams that are just as driven as the team behind Void.
The single least productive things you can do though are try to make Void something it isn’t. It will annoy you because the changes won’t go through. It will annoy the maintainers because you will be decreasing the SnR when there is real work to be done. It will annoy the community because you’ll be that guy who just keeps bringing up the same idea over and over that has already been dismissed.
Thanks for pretending to read my wall of text, if you are curious about any of this or why/how I’ve drawn certain conclusions feel free to ping me on IRC, I’m rarely on the forums.