Anyhow, I would like to see alsa-drivers removed, but only with the condition that the in-kernel ALSA drivers work as expected. The bug I mentioned, if you reread it, you will notice that if there was no external alsa-driver, hda-intel would be marked broken as well, for some users.
Just to make it short, in-kernel hda-intel works for some people, but it doesn't for others. The logical bug to be fixed is that the kernel guys (not our kernel guys) check what's going wrong in there. If that did happen, then your point of view wouldn't need any argumentation. What really has annoyed me is the fact that the external driver works, and I hope to see that not happen again. Unfortunately, that's not in my hands right now. If we are able to convince upstream maintainers about this, to report the bugs properly and to get them fixed as required, we could go ahead and remove alsa-drivers. Until that happens, seeing the current proofs, the best choice is "if it works don't fix it". On 3/28/07, Daniel Drake <[EMAIL PROTECTED]> wrote:
Ioannis Aslanidis wrote: > Well, to be honest, I am neither supporter nor detractor. I think that > it's upstream that should go and fix themselves. It's them who have > caused all this. The bug you linked to is a natural effect of maintaining kernel drivers outside of the kernel source tree. It is not the fault of the alsa-driver maintainers (either Gentoo or upstream people who push tarballs), since they don't control the in-kernel API's. It's not the kernels fault because the kernel does not claim to maintain a stable internal API, and it actually readily claims the opposite. This is not going to change. With reference to the problems an unstable internal API causes, kernel people will give you one answer and one answer only: get your code in the kernel. Problems like these simply do not happen there. So, in this particular case, if the alsa-driver portage package did not exist, the bug in question (plus a whole series of others in the same class) simply will not ever occur. This is a good thing. > ALSA guys do not support in-kernel stuff. ALSA guys being upstream alsa-project.org people? That's not correct, they are the people who put the code in the kernel. ALSA guys being downstream gentoo maintainers? That's true, since there is a separate team who maintain the kernel. > Kernel guys do not support alsa stuff. Upstream kernel guys? No, they support alsa, actually most of the bugs are handled by the alsa-project.org people who generally handle them very well. Downstream kernel guys (i.e. Gentoo kernel herd)? No, we support the sound subsystem and the ALSA drivers just like all other subsystems and drivers in the kernel. Thanks, Daniel -- gentoo-dev@gentoo.org mailing list
-- Ioannis Aslanidis <deathwing00[at]gentoo.org> 0xB9B11F4E -- gentoo-dev@gentoo.org mailing list
