[Gta04-owner] Modem crashing?

[Neil Jerram]

Radek Polak <psonek2 at seznam.cz> writes:

> Anyways i have cherry-picked your commit for next qtmoko version and i will 
> now comment out AT_OPSYS=0,2 to enable 3g and we will see if the reenumeration 
> will be gone. And i could finally get rid of the ugly patch that deals with 
> reenumeration.

That reminds me of something I was meaning to ask.  Why is that patch
kept as a Debian patch, and not already integrated into the QtMoko code?
I think it means one gets different results if one builds as a Debian
package, or per the README instructions.  Is that the intention?

For added confusion, I think that patch _was_ in the main QtMoko code
for a while when it got accidentally committed together with my first
(wrong) DTMF change.  It disappeared again when that change was
reverted.

Following that, I included the patch for a while in my own
work-in-progress branch, and saw reenumerations and QtMoko restarts
quite regularly, especially when travelling on a train with 3G data
connection.

Then a few weeks ago - I think - I did some rebase and decided not to
keep that patch in my branch anymore.  Since then I haven't seen any
QtMoko restarts - which I presume is because the code to do that is not
there - but I do still see occasional evidence of modem breakage,
e.g. SMS's not being sent out, or data not working anymore.

In summary, in other words,

- if you (or anyone) are trying 3G but not seeing QtMoko restarts, could
  it be because you've built QtMoko in a way that doesn't include that
  patch?

- I'm curious about whether it was intended for the patch to be only a
  Debian patch, or if that was an oversight.

Regards,
        Neil