Shay wrote:
My point about AntiX was it properly identifies the card and installs the right driver is all.
Shay,
That is a perfectly reasonable assumption.
The long story is that MX contains two different sets of broadcom drivers. wl (a proprietary driver supplied by broadcom through the package broadcom-sta-dkms) works with MORE parts, especially newer parts. The open source set (b43, b44, tg3, etc...) works with many parts, especially older parts, but not as many of the newer parts.
The downside is that the two sets of drivers conflict, so both can't be active at the same time. MX tries to sort this out with a preemptive script that tries to determine which driver to use and then blacklist the conflicting one. This is done by pid numbers. Its a neat idea, but its also the first time we've tried it and there are apparently some issues to work out. We only had a limited set of broadcom parts to test with.
To my knowledge, the parts that you and conn.sk are the first to encounter issues, although for different reasons. Your part is actually listed as supported by wl (and only partially by b43), but your ethernet is apparently using tg3, which also conflicts with wl. In your case, I would blacklist wl and unblacklist ALL of the open source drivers (or just uninstall broadcom-sta-dkms in an installed or live w/persistence session).
The reason that antiX works OOTB for your part is that the conflicting wl driver isn't present, so no conflicts. The open source drivers do not conflict with one another.
In conn.sk's case, it looks like the detection algorithm failed, hence our desire for his device output.
***edit*** I might be wrong about tg3 conflicting with wl. Full disclosure.