Created
November 18, 2024 08:04
-
-
Save douglarek/5f796de1563b817f96632f634e64d42a to your computer and use it in GitHub Desktop.
这个是在 gentoo IRC 群维护者回复 Chromium 为什么会给 ~arch 用户默认升级到 beta
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
[2024年11月17日星期日] [GMT+8 19:51:10] <douglarek> Why does the recent update of the chromium package include beta and dev without setting them as masked? Do ~amd users have to use dev by default? https://packages.gentoo.org/packages/www-client/chromium | |
[2024年11月17日星期日] [GMT+8 19:52:02] <douglarek> s/dev by default/beta by default | |
[2024年11月17日星期日] [GMT+8 20:03:25] <sam_> ~amd64 users have always had beta exposed to them | |
[2024年11月17日星期日] [GMT+8 20:03:32] <sam_> dev is unkeyworded (so functionally the same as masked) | |
[2024年11月17日星期日] [GMT+8 20:03:33] <sam_> cc kangie | |
[2024年11月17日星期日] [GMT+8 20:03:43] <sam_> i.e. nothing has functionally changed | |
[2024年11月18日星期一] [GMT+8 00:49:40] <Kangie> douglarek_: what sam_ said. You can select the 0/stable subslot if you don't want beta | |
[2024年11月18日星期一] [GMT+8 10:10:29] <douglarek_> Kangie: I know I have a choice, I just don't understand why ~amd users are expected to upgrade to beta by default. | |
[2024年11月18日星期一] [GMT+8 10:11:13] <Kangie> because you're EXPLICITLY OPTING INTO IT | |
[2024年11月18日星期一] [GMT+8 10:11:25] <Kangie> sorry that was a little OTT | |
[2024年11月18日星期一] [GMT+8 10:11:31] <Kangie> 1 sec though, I'll dig out the reference | |
[2024年11月18日星期一] [GMT+8 10:13:00] <Kangie> Chromium project policy before my time: https://wiki.gentoo.org/wiki/Project:Chromium#Where_can_I_find_more_info_about_the_releases.3F | |
[2024年11月18日星期一] [GMT+8 10:13:08] <Kangie> Gentoo devmanual entry on keywording: https://devmanual.gentoo.org/keywording/ | |
[2024年11月18日星期一] [GMT+8 10:13:23] <douglarek_> Kangie: "because you're EXPLICITLY OPTING INTO IT", To be honest, I don't understand why I clearly chose him. It was always the stable version before, but after a sudden update, it switched me to the beta version. If this is supposed to be my choice, then it's really overboard. | |
[2024年11月18日星期一] [GMT+8 10:13:36] <Kangie> the tl;dr here is that, by selecting ~arch, as a user you are _explicitly_ opting in to using packages that are not marked "stable" | |
[2024年11月18日星期一] [GMT+8 10:13:53] <Kangie> note that they're not "unstable", they're just "insufficiently tested to have the stable keyword" | |
[2024年11月18日星期一] [GMT+8 10:14:04] <Kangie> > arch (~x86, ~ppc-macos) ("testing") | |
[2024年11月18日星期一] [GMT+8 10:14:06] <Kangie> > The package version and the ebuild are believed to work and do not have any known serious bugs, but more testing is required before the package version is considered suitable for arch. | |
[2024年11月18日星期一] [GMT+8 10:14:36] <Kangie> also noting the chrommium release cycle, beta always contains the security updates for stable as they're both picking from dev | |
[2024年11月18日星期一] [GMT+8 10:14:39] <tdr> https://wiki.gentoo.org/wiki/Handbook:X86/Full/Portage is also nice | |
[2024年11月18日星期一] [GMT+8 10:15:17] <Kangie> anyway douglarek_, you can restrict it to only picking stable on ~arch | |
[2024年11月18日星期一] [GMT+8 10:15:28] <sam_> Kangie: well, I don't know if I fully agree with that | |
[2024年11月18日星期一] [GMT+8 10:15:43] <Kangie> select the '0/stable' slot | |
[2024年11月18日星期一] [GMT+8 10:15:47] <sam_> in general, we try to avoid RCs and betas out of ~arch, not least because the idea is "this thing will eventually become stable", otherwise it doesn't ever get tested | |
[2024年11月18日星期一] [GMT+8 10:15:47] <Kangie> https://packages.gentoo.org/packages/www-client/chromium | |
[2024年11月18日星期一] [GMT+8 10:15:48] <sam_> now, the difference is | |
[2024年11月18日星期一] [GMT+8 10:15:53] <sam_> for chromium, it works differently | |
[2024年11月18日星期一] [GMT+8 10:16:05] <sam_> chromium rebrands a stream, and often, the beta tag literally becomes the stable one | |
[2024年11月18日星期一] [GMT+8 10:16:09] <sam_> same code, just rebranded | |
[2024年11月18日星期一] [GMT+8 10:16:16] <sam_> so it _is_ valuable to be testing it and have stuff smoked out | |
[2024年11月18日星期一] [GMT+8 10:16:22] <Kangie> early stable is 99.9% of the time 'beta' | |
[2024年11月18日星期一] [GMT+8 10:16:27] <sam_> exactly | |
[2024年11月18日星期一] [GMT+8 10:16:37] <sam_> but make that argument, not the one that we regularly put betas on ~arch | |
[2024年11月18日星期一] [GMT+8 10:16:38] <sam_> *in | |
[2024年11月18日星期一] [GMT+8 10:16:49] <Kangie> but also this behaviour hasn't changed in _many years_ | |
[2024年11月18日星期一] [GMT+8 10:16:54] <sam_> yes, that too | |
[2024年11月18日星期一] [GMT+8 10:16:59] <Kangie> there's no new surprise here | |
[2024年11月18日星期一] [GMT+8 10:17:14] <sam_> also, frankly, Chromium is so hard to maintain, that such a policy is IMO reasonable anyway because otherwise it would become unsustainable to package it (because of huge pressure to have stable always working) | |
[2024年11月18日星期一] [GMT+8 10:17:40] <Kangie> no joke I would move to immediately mask and treeclean if it had to go stable-only | |
[2024年11月18日星期一] [GMT+8 10:17:50] <sam_> but yes, nothing's changed here, and I think it's very reasonable given ^ and the fact that beta often becomes literally stable (same exact code gets relabelled) | |
[2024年11月18日星期一] [GMT+8 10:17:53] <Kangie> there have been 6+ week periods where I've had to wark on dev or beta | |
[2024年11月18日星期一] [GMT+8 10:18:07] <Kangie> beside the point theugh | |
[2024年11月18日星期一] [GMT+8 10:18:16] <Kangie> the other part is that the chromium beta channel is misnamed | |
[2024年11月18日星期一] [GMT+8 10:18:24] <Kangie> it's the 'next stable' channel | |
[2024年11月18日星期一] [GMT+8 10:18:25] <Kangie> if that helps | |
[2024年11月18日星期一] [GMT+8 10:18:28] <sam_> yes | |
[2024年11月18日星期一] [GMT+8 10:18:31] <sam_> that's the key bit really | |
[2024年11月18日星期一] [GMT+8 10:18:35] <Kangie> the dev channel is explicitly 'unstable' | |
[2024年11月18日星期一] [GMT+8 10:18:48] <Kangie> but upstream calls it beta and it'd me more confusing for us to differ | |
[2024年11月18日星期一] [GMT+8 10:19:29] <Kangie> so douglarek_ sorry for jumping down your throat there - it was uncalled for | |
[2024年11月18日星期一] [GMT+8 10:20:04] <douglarek_> Kangie: I am just confused, thank you for your further clarification. | |
[2024年11月18日星期一] [GMT+8 10:20:16] <Kangie> but this is not new, nor is it really surprising. Honestly I'd be comfortable running the 'dev channel' (and often do) | |
[2024年11月18日星期一] [GMT+8 10:20:32] <sam_> we should add an explanation to that wiki section you linked | |
[2024年11月18日星期一] [GMT+8 10:20:35] <sam_> I'll write it up in a bit | |
[2024年11月18日星期一] [GMT+8 10:21:02] <Kangie> hate google's release cycle/process sometimes, but their 'dev' channel is more thoroughly tested than a lot of released software out there | |
[2024年11月18日星期一] [GMT+8 10:21:12] <Kangie> and when it's not, well a new version is out next week! | |
[2024年11月18日星期一] [GMT+8 10:21:47] <Kangie> thanks sam_, we could always stand to improve docs and make that easier to find | |
[2024年11月18日星期一] [GMT+8 10:22:08] <Kangie> and the chromium stuff really does need a top-to-bottom review by someone that isn't me. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment