Re: QMI support in network-manager-applet

[Aleksander Morgado <aleksander lanedo com>]

On 07/05/13 18:47, Bjørn Mork wrote:
> > > >  and I don't think we should do
> > > > > > > > > that e.g. for ZTE QMI modems (which are not handled with QMI currently,
> > > > > > > > > as they suck).
> > > > > I'm wondering about this...  Except for the ones which appear as routers
> > > > > and run some buggy QMI proxy daemon, I'd expect most of the ZTE modems
> > > > > to have about the same set of bugs as the other Qualcomm firmware based
> > > > > modems.  Modems like e.g. the MF821D look and feel pretty much like any
> > > > > other MDM9200 based device. 
> > >
> > > I have a ZTE MF637 which is QMI based, and it just ends up crashing if
> > > used with QMI. I even have a branch to enable ZTE-based QMI modems
> > > (http://cgit.freedesktop.org/ModemManager/ModemManager/log/?h=zte-qmi)
> > > but didn't spend much time with it after my first unsuccessful attempts...
> Is this a device designed to operate as a NATing router, or is it just a
> "regular" modem? 

It is a plain regular HSPA modem.


> The MF60 I've still got (luckily the owner seems to have forgotten about
> it :) is also crashing really bad on any QMI WDS command, making it
> completely useless with QMI.  But this device is a wifi-3G router which
> it obviously runs some QMI proxy daemon, forwarding most commands but
> intercepting WDS.  And the ZTE WDS implementation sucks bigtime even for
> a modem firmware.
>
> But I was hoping we could detect and avoid the troublesome devices using
> "safe" commands, without making all other devices suffer.  As I said,
> the MF821D works really fine with QMI.

 I'll retest it again tomorrow and gather debug logs to see where
exactly it failed.

-- 
Aleksander