little bit off topic: problem with mm

[Aleksander Morgado]

> 
> Or alternatively how about something like dcbw/06-endpoints?  The only
> reason we care about USBIF #0 on Huawei is that they used to always be
> the modem/PPP interface.  But obviously that's not always true anymore.
> However, we can be reasonably sure that an interface with only two
> endpoints is *not* a modem/PPP interface.
> 
> In any case, I think something like the patch in the branch ported to
> git master would be useful for cases where we don't already have
> explicit modem/PPP port tags for various plugins (especially ZTE), and
> could be useful for the Huawei bits too.
> 

According to the logs usbif#0 in this modem is neither an AT port nor a
QCDM port; maybe GPS data port? Can we use the number of endpoints
reliably to filter out ttys which for sure are *not* AT or QCDM?

Given that the usbif#0 is no longer what we expected; we should probably
change that logic completely... right now it will possibly end up
working (if the previous patch works), but in the worst case we're
really having a long probing time due to not probing the ports in parallel.

-- 
Aleksander