Package Details: b43-firmware 6.30.163.46-1

Git Clone URL: https://aur.archlinux.org/b43-firmware.git (read-only)
Package Base: b43-firmware
Description: Firmware for Broadcom B43 wireless networking chips - latest release
Upstream URL: https://wireless.wiki.kernel.org/en/users/Drivers/b43
Licenses: unknown
Conflicts: b43-firmware-classic
Submitter: Xavion
Maintainer: Xavion
Last Packager: Xavion
Votes: 234
Popularity: 1.708775
First Submitted: 2008-11-22 01:47
Last Updated: 2015-12-31 21:30

Latest Comments

Xavion commented on 2015-12-31 22:05

Okay, I've changed the license name and downgraded the warning to a post-installation message. I'm a bit surprised I didn't do the latter that way the first time to be honest.

Anyway, this should bring about a peaceful end to our discussion. You might recall that the last one we had didn't conclude quite so smoothly. You ended up saying: "Don't bother emailing me again. You're now blocked." Am I still blocked, BTW? :-)

Scimmia commented on 2015-12-31 15:44

The license is what gets put on the end package. It is most definitely not GPL. Change it to custom, proprietary, unknown, whatever. Just not GPL.

I read the discussion, you're the only one who mentioned a warning in the PKGBUILD. That's not where it belongs. Even if you want it there, making the user acknowledge it is the real problem.

Xavion commented on 2015-12-31 08:46

Thanks for letting me know about the 'conflicts' line; I've just put brackets around it. Fortunately, it was correct in all of my other PKGBUILDs.

The license is listed as 'GPL' because that's the way it is for the "b43-fwcutter" package in [core]. Both packages point to the same upstream URL. The license for the tarball isn't listed on the download page, nor is it mentioned in the tarball itself.

Regarding the warning message, I know it's not an ideal solution. There's a lengthy discussion about it in the comments below, starting on "2014-02-09" and ending on "2015-05-22" :-). After reading through all of it, let us know if you have a better solution.

Scimmia commented on 2015-12-31 04:32

With pacman from git HEAD:
==> ERROR: conflicts should be an array

Edit: The prompt in the build function is a bad idea. You're not supposed to need input for a pkgbuild. The license is also wrong; if it was GPL, we could redistribute it.

Xavion commented on 2015-10-22 07:48

I'm certainly not doing it intentionally. The "b43-fwcutter" tool creates the 'b43' directory with 750 permissions. I'm guessing those guys did it for a reason, so my inclination is to leave it alone.

fightcookie commented on 2015-10-22 00:44

Is it intenional and right that after the installation the b43 directory in /lib/firmware/ has 750 permissions as opposed to all the other directories there, that all have 755?

Xavion commented on 2015-05-22 22:18

Okay, I've gone ahead and split the packages as noted below. Each PKGBUILD contains a warning message to inform the user of the other's existence.

If the v6.30 device array ever expands significantly, I will notify "b43-firmware-classic" users accordingly via a comment to that package's webpage.

Kuci commented on 2015-05-22 16:17

There is http://www.lwfinger.com/b43-firmware/broadcom-wl-6.30.163.46.tar.bz2 available.

ewtoombs commented on 2015-05-12 21:36

I'm having problems with version 6 as well. Sometimes, the interface gets disconnected, requiring manual intervention to bring the interface back online. When this happens, no programs can access the network, but NetworkManager still says the link is still fine. When I force NetworkManager to disassociate and reassociate, the link is fixed until the next episode. I'm trying firmware v5 now to see if there's a difference. I'll keep you posted.

skydrome commented on 2015-04-15 15:45

i would ask for v5 to stay its own pkg also, for my aging Acer laptop v6 firmware kills all connectivity for some reason, reverting back to v5 fixes it. using the -classic tag sounds good for it.

All comments