Is epoch increment optional on version scheme change in -git packages?
Dear AUR-maintainers, in this post [¹] I was told to raise this issue on this mailing list, so I do so. Note that I will not stay part of this mailing list; if there is an authorative answer maybe the best is to post it also to [¹] so everyone can notice. The issue is described in full in [²] and some short discussion followed. I also copypaste the issue here (from the Forum -- so forum-syntax): --- regarding [url=https://github.com/chaotic-aur/packages/issues/3151#issuecomment-2230462191]this discussion[/url] about the AUR-package [url=https://aur.archlinux.org/packages/gr-lora_sdr-git]gr-lora_sdr-git[/url] I want to ask about the `$epoch` variable in `-git` packages in the AUR. The maintainer argues that `$epoch` [i]may[/i] be used, but not has to, when the versioning scheme changes so that a newer version of the `PKGBUILD` would result in a lower version according to `pacman`'s version compare logic. In this case, the old `$pkgver` (before adoption and fixing of the AUR package) was `8133965`, the new one now is `0.5.r4.g8133965`. [url=https://github.com/chaotic-aur/packages/issues/3151#issuecomment-2230495522]The maintainer argues as following to not increment `$epoch`[/url]: [quote=xiota] AUR is still build on demand. AUR helper and Chaotic AUR users will receive notification and can force the update if they want. Users unwilling to manually force the upgrade can keep the old version. It's ultimately their choice. [/quote] [url=https://github.com/chaotic-aur/packages/issues/3151#issuecomment-2230462191]I was reasoning that `$epoch` _must_ be incremented in such a case[/url]: [quote=dreirund] [quote=xiota] Also, I don't use epoch on AUR packages because they're build on demand. [/quote] But for the packages in the AUR you still need to set `$epoch`. Since the AUR is independently usable from `chaotic-aur`. The version scheme change you did in `gr-lora_sdr-git` requires an increment of `$epoch` in the AUR package, since the old version was `8133965`, the new one is `0.5.r4.g8133965`, which is lower by version compare logic. So setting `epoch=1` in the AUR package actually is required. [/quote] [url=https://wiki.archlinux.org/title/PKGBUILD#epoch]The wiki[/url] seems not to be clear about it, it may be read in both ways: [url=https://github.com/chaotic-aur/packages/issues/3151#issuecomment-2230607960]One standpoint[/url]: [quote=dreirund] Regardless what opinions anyone has, the guidelines are clear about that: [url]https://wiki.archlinux.org/title/PKGBUILD#epoch[/url] [quote=https://wiki.archlinux.org/title/PKGBUILD#epoch] It is used when the version numbering scheme of a package changes (or is alphanumeric), breaking normal version comparison logic. [/quote] [/quote] and [url=https://github.com/chaotic-aur/packages/issues/3151#issuecomment-2230660924]answer[/url] to that: [quote=xiota] From your own reference: [quote=quote=https://wiki.archlinux.org/title/PKGBUILD#epoch] Warning: epoch should only be used when [b]absolutely required[/b] to do so. [/quote] There is no statement requiring epoch to be used. Only requirements to meet when it is used. [/quote] So, is there any clear official rule for this case? --- [i]EDIT:[/i] After a discussion [url=https://github.com/chaotic-aur/packages/issues/3151#issuecomment-2230462191]over there[/url], the maintainer has moved the discussion thread "[i]to a repository you don't have access to.[/i]". Regards! --- Regards!, @dreieck. [¹] https://bbs.archlinux.org/viewtopic.php?pid=2184346#p2184346 [²] https://bbs.archlinux.org/viewtopic.php?pid=2184286
On Tue, 2024-07-23 at 10:54 +0200, dreieck wrote:
Dear AUR-maintainers,
I don't think there's anything to say that hasn't been said on the forums already. My personal POV, just to add a little more than a confirmation, I would argue that you should use an epoch when your versioning has a break in continuity. Sure, the rules aren't enforced as strictly on the AUR, but that doesn't mean one shouldn't try to provide the right experience for their users.
Jul 24, 2024, 15:03 by bert@bertptrs.nl:
On Tue, 2024-07-23 at 10:54 +0200, dreieck wrote:
Dear AUR-maintainers,
I don't think there's anything to say that hasn't been said on the forums already.
My personal POV, just to add a little more than a confirmation, I would argue that you should use an epoch when your versioning has a break in continuity. Sure, the rules aren't enforced as strictly on the AUR, but that doesn't mean one shouldn't try to provide the right experience for their users.
I'm just a user, but I argue that just because the Chaotic-AUR maintainers want to avoid epochs at all cost doesn't mean it's the right thing to do. Yes, they should be used sparingly, I agree with that. But I feel that the end-user experience is more important than wanting a "clean" version number. If the Chaotic-AUR maintainer wants to avoid epochs that badly, it seems like they could maintain their own version of the PKGBUILD.
If the Chaotic-AUR maintainer wants to avoid epochs that badly, it seems like they could maintain their own version of the PKGBUILD.
They're already patching AUR packages that are missing some stuff (so they could use that mechanism to remove the epochs): https://github.com/chaotic-aur/interfere
Ahoj, On Wed, 24 Jul 2024 22:29:20 +0200 (CEST), Brian Allred <brian@allred.io> wrote:
Jul 24, 2024, 15:03 by bert@bertptrs.nl:
I don't think there's anything to say that hasn't been said on the forums already.
I'm just a user, but I argue that just because the Chaotic-AUR maintainers want to avoid epochs at all cost doesn't mean it's the right thing to do. Yes, they should be used sparingly, I agree with that. But I feel that the end-user experience is more important than wanting a "clean" version number.
given the answers that were given so far, and given that there is no authorative answer since 6 days, can I conclude that there is no authorative rule regarding the use of `$epoch` as questioned (only suggestions, but nothing that must be followed for an AUR package)?, and I can mentally "close" this issue? Or is a decision still to me made? Regards!
Hi
I don't think there's anything to say that hasn't been said on the forums already.
I'm just a user, but I argue that just because the Chaotic-AUR maintainers want to avoid epochs at all cost doesn't mean it's the right thing to do. Yes, they should be used sparingly, I agree with that. But I feel that the end-user experience is more important than wanting a "clean" version number. given the answers that were given so far, and given that there is no authorative answer since 6 days, can I conclude that there is no authorative rule regarding the use of `$epoch` as questioned (only suggestions, but nothing that must be followed for an AUR package)?, and I can mentally "close" this issue?
Some answers in the forum were already given, the wiki itself for PKGBUILD [1] states:
Used to force the package to be seen as newer than any previous version with a lower epoch.
This is clear for anyone, its use covers exactly the cases where the upstream version numbering changes or when the downstream version number change would make the newer package with a version number previous than the previous. For git packages the pkgver is set by downstream in the PKGBUILD so it applies also to the previous condition. There's not any exception for VCS packages (including -git of course), so there's no reason for not applying the change (setting the epoch) when the version numbering change, regardless being a VCS package or not. Regards [1] https://wiki.archlinux.org/title/PKGBUILD#epoch -- Fabio Castelli aka Muflone
Ahoj, On Sun, 4 Aug 2024 16:24:19 +0200, Muflone <muflone@archlinux.org> wrote:
I don't think there's anything to say that hasn't been said on the forums already.
I'm just a user, but I argue that just because the Chaotic-AUR maintainers want to avoid epochs at all cost doesn't mean it's the right thing to do. Yes, they should be used sparingly, I agree with that. But I feel that the end-user experience is more important than wanting a "clean" version number. given the answers that were given so far, and given that there is no authorative answer since 6 days, can I conclude that there is no authorative rule regarding the use of `$epoch` as questioned (only suggestions, but nothing that must be followed for an AUR package)?, and I can mentally "close" this issue?
Some answers in the forum were already given, the wiki itself for PKGBUILD [1] states:
Used to force the package to be seen as newer than any previous version with a lower epoch.
This is clear for anyone, its use covers exactly the cases where the upstream version numbering changes or when the downstream version number change would make the newer package with a version number previous than the previous. For git packages the pkgver is set by downstream in the PKGBUILD so it applies also to the previous condition.
There's not any exception for VCS packages (including -git of course), so there's no reason for not applying the change (setting the epoch) when the version numbering change, regardless being a VCS package or not.
OK. Can then you or anyone else with "authority" reach out to xiota if that really is true what you write? I do not dare & I am just a user. Regards!
participants (6)
-
Bert Peters
-
Brian Allred
-
DodoGTA GT
-
dreieck
-
Muflone
-
oid-maps＠seznam.cz