ublox SARA-U260 not showing up in mmcli

[Dan Williams]

On Mon, 2017-05-15 at 21:31 +0200, Aleksander Morgado wrote:
> On Mon, May 15, 2017 at 5:48 PM, Dan Williams <dcbw at redhat.com>
> wrote:
> > > > ...So, does mean I need to modify the code somewhere to disable
> > > > the
> > > > send
> > > > delay?  Could you point me to where that may be in the
> > > > code?  Thanks!
> > > 
> > > We have some plugins already setting send_delay=0 during a custom
> > > setup_ports() method (see e.g. mm-broadband-modem-samsung.c) but
> > > your
> > > case is a bit special because you also need the send delay even
> > > during
> > > port probing, and for that we'd need to create the MMPlugin
> > > object
> > > with MM_PLUGIN_SEND_DELAY property set.
> > > 
> > > Dan: the Altair, Samsung and HSO plugins use send_delay=0 in
> > > setup_ports() (i.e. once the modem has been created), but they
> > > don't
> > > use it for the port probing phase (i.e. no MM_PLUGIN_SEND_DELAY
> > > setting when creating the plugin). Is this an error?
> > 
> > I think just 2006-era modems and some old phones would get
> > confused,
> > and we C&P-ed that everywhere.  If we want to speed up probing a
> > bit,
> > we could remove that for sane plugins.  We don't want to remove
> > that
> > for the more esoteric ones (anydata, iridium(?), motorola, pantech)
> > and
> > also for Huawei/ZTE which still support really old stuff.
> > 
> > For Samsung and HSO, we can definitely set to 0 and test it.  I
> > have
> > both Samsung modems and a bunch of HSOs, so no problem there.
> 
> How about we set MM_PLUGIN_SEND_DELAY to 0 in all Samsung, HSO and
> Altair, to make it consistent with what we have in the modem object;
> and then add the same setting for the u-blox plugin? I can test the
> u-blox changes with a Toby-L2 I have here.

Sure.

Dan