[pkg-wpa-devel] Maybe wpasupplicant just shouldn't roam at all..

[Kel Modderman]

Joachim Breitner wrote:
>> The above paragraph is one of many in a series of thread's just started
>> due to the removal of wpasupplicant's init script that describes the 
>> situation where a wifi roaming daemon is indeed required, but is 
>> out-of-the scope of the wpasupplicant packaging itself.
>>
>> The ifupdown glue for wpa_supplicant should be just the stuff required 
>> for such a daemon to bring up various mappings/network profiles defined 
>> in /e/n/i that require wpa encryption (in the case of a simple mapping 
>> daemon).
>>
>> So, i put it to you, that the removal of the wpasupplicant init script 
>> based daemon has uncovered a void; there is no *single* wifi roaming 
>> daemon available to us (at the time of writing), or we simply have not 
>> found it yet.
>>     
>
> You are right but I'd put it differently: There are at least two wifi
> roaming daemons, but they happen to have names that refer to other
> usage, and they have functionality they should not have. :-)
>   

Well, I didn't know exactly how to express the sentiment with the most 
clarity, it was more of a brain fart.

Agreed, wpa_supplicant is a good roaming daemon, just a shame it prefers 
to focus on the "other" stuff :-)
> Is the wpasupplicant upstream on this list? Maybe someone with a good
> connection to him could kindly ask him if he'd be willing to split the
> roaming code out of wpasupplicant in a single daemon. The code _is_
> there, just in the wrong place. (Of course, anyone with reasonable C
> skills, a bit of WLAN knowledge and some time could easily do that, too,
> I guess. Any volunteers?)
>   

My guess would be that he (Jouni) would not split and maintain it by 
himself, and I'm not suitable for that task either.

Thanks, Kel.