CVS access probably blocked

[Nicolas Viéville]

> > As a newcomer, I've tried to setup my account on fas.rpmfusion.org, and
> > uploaded my RSA public key. After a few tries to connect to the CVS,
> > making mistakes before getting the right method, it seems that my
> > account is blocked.
> You are not approved yet, so this is not possible for you to access
> cvs even in RO.
> The anonymous access isn't implemented either.
> 
> The best for you is to make a diff between the spec file from the
> latest src.rpm from the f15 updates repository.

Ok! But I think I didn't understood where to post/upload it. Any
information about this would be great.

> About wl-kmod, You might have read this article from the launch of linux 3.1.
> http://linuxfr.org/news/sortie-du-noyau-linux%C2%A031#toc_31

I've also followed the links in this article :

https://lwn.net/Articles/456762/
https://lwn.net/Articles/457674/

It seems that the question is not in a stable state actually. But, if I
understand correctly what is discussed, different Broadcom wireless
drivers will still there, and the user have to choose what is best for
him.

> In short, I wonder it it wouldn't be wise to have this new driver
> enabled from the stating-kmod than the wl-kmod that might become
> unmaintained.

For the moment, as shown in this table
http://wireless.kernel.org/en/users/Drivers/b43#Supported_devices the
bcm4313 wireless chipset isn't supported by the b43 driver (WIP). As
it's my wireless chipset, I've only been able to use it with the wl
driver (hybrid driver: wireless + Bluetooth) provided by Broadcom.
Broadcom just released the 5.100.82.111 version of this driver, and it
seems to be compatible with kernels 3.x as the 5.100.82.38. For my part,
I haven't only tested these two versions with 2.6.40 kernel series on
Fedora. 5.100.82.111 kernel panics, like others maintainers noticed it
(http://aur.archlinux.org/packages.php?ID=19514). 5.100.82.38 works.
I've also tested the 5.100.82.38 driver on Ubuntu Oneiric amd64 with
kernel version 3.0.0-12 (Ubuntu names and versions) and a bcm4313
chipset too. It works too.

> Using staging-kmod with the brcmsmac enabled will requires to
> blacklist b43 and b44. But a usability test has to be done by someone
> who owns the hardware.

Yes, tests have to be done on all supported hardware if possible. But,
if the situation described in the links above persist, with all the
different drivers co-existing, we had to provides blacklist files and
script to maintain/setup them at install time.

> See also: https://bugzilla.rpmfusion.org/show_bug.cgi?id=1997

Thanks for your response!


-- 
Nicolas Viéville