Re: [RFC] Removing sys-kernel/vanilla-kernel
Sam James <[email protected]> Tue, 02 Jun 2026 18:06:43 +0100
| Newsgroups | gmane.linux.gentoo.devel |
|---|---|
| Organization | Gentoo |
| Message-ID | <[email protected]> |
Immolo <[email protected]> writes: > Hi, > > In the more fringe areas I monitor for my work looking for handbook > and doc bugs in the wiki, I have found: > >> 1) Drop vanilla-sources as well. > > I've noticed there is a user base around sys-kernel/vanilla-sources. > Their reasoning is they like pulling kernels directly from kernel.org > and this gives them a package manager solution for their needs. But they're still relying on us to do it promptly *and* not have removed the relevant ebuild. I really think vanilla use is usually fetishisation of an "untainted" kernel without realising: a) we don't change much at all, deliberately, and b) kernel.org is not some sort of perfect entity. > > There should at least be a callout to give those users the chance to > take the maintaining burden themselves before any action is taken IMO. > >> 2) Drop vanilla-sources and add a vanilla USE flag to gentoo-sources. > > This is one of those ideas so good I wonder why we didn't do it before. > > My selfish reasoning is it will allow me to streamline the messy kernel > docs even further, which has been an ongoing task for the last few years. > > > As for vanilla-kernel. I have at best seen one user that uses it in the last > 6 years. The only time I've used it is for bisecting with the 9999s. > > While that sounds bad, it should be noted we don't do a good job of highlighting > it exists and we have only just started to highlight how the dist kernel can be > used in a way that is similar to how we used to explain genkernel. > > I would suggest vanilla-kernel has no purpose now for us to waste time on, so > drop it. However someone should look into if genkernel users use vanilla-sources > as sys-kernel/vanilla-kernel is going to be their migration path keeps > heading on its > current path of issues.There is likely a day we would need to think > about bringing it > back, but in a better way than we currently provide this package. > I'm not sure there's much of a connection there. People may be using vanilla-sources with or without an initramfs, right? Regardless of if genkernel made it or something else. > I hope there are at least some useful insights in here to help further > discussions. > > Kind regards, > > immolo
signature.asc
(application/pgp-signature, 418 B)
-----BEGIN PGP SIGNATURE----- iQEBBAEWCgCpFiEEJaa7iN2bdkxrVUHCc4QJ9SDfkZAFAmofDaQbFIAAAAAABAAO bWFudTIsMi41KzEuMTIsMiwyXxSAAAAAAC4AKGlzc3Vlci1mcHJAbm90YXRpb25z Lm9wZW5wZ3AuZmlmdGhob3JzZW1hbi5uZXQyNUE2QkI4OEREOUI3NjRDNkI1NTQx QzI3Mzg0MDlGNTIwREY5MTkwDxxzYW1AZ2VudG9vLm9yZwAKCRBzhAn1IN+RkBvF AP4uKX2r8hturIOw6DqA6oTdmqVtEtzWYXAn55cGm7KtfQD+MPN33kyST3X1IFhb 80+UKfIf8gsK+XFPwdujDTd76g4= =XUrl -----END PGP SIGNATURE-----