UMW190 not supported - confusion over CDMA vs GSM

[Noel J. Bergman]

Binary package hint: modemmanager

The UMW190 is Verizon's current CDMA *PLUS* GSM offering. Modem Manager / Network Manager fails to use it because it is being detected as a GSM modem, even though here in the USA, it needs to be used as a CDMA modem on Verizon's network.

See http://ubuntuforums.org/showpost.php?p=8996651&postcount=2 for details and a workaround using Gnome PPP.

ProblemType: Bug
DistroRelease: Ubuntu 10.10
Package: modemmanager 0.4+git.20100809t153145.be28089-0ubuntu1
Uname: Linux 2.6.36-020636rc7-generic x86_64
NonfreeKernelModules: nvidia
Architecture: amd64
Date: Fri Oct 8 19:59:05 2010
InstallationMedia: Ubuntu 10.10 "Maverick Meerkat" - Alpha amd64 (20100601)
ProcEnviron:
 LANG=en_US.utf8
 SHELL=/bin/bash
SourcePackage: modemmanager

[Dan Williams (dcbw)]

Upstream modem-manager fix is:

http://cgit.freedesktop.org/ModemManager/ModemManager/commit/?id=49097b6de5daa63de84f16386825de5b17590fcc

but confirmation that it works (preferably in the referenced gnome bugzilla bug) would be nice. This fix will treat the device as a CDMA device instead of GSM, since the device reports that it can do both command sets.

[Noel J. Bergman (noeljb)]

The fix is in the Network Manager Daily PPA. After I test it, I'll mark it FIX COMMITTED here if it works.

[Noel J. Bergman (noeljb)]

The upstream fix committed by Dan Williams works (I'm using it now from the Network Manager Daily PPA - https://launchpad.net/~network-manager/+archive/trunk).

This should be updated in Lucid and Maverick. The specific path is clear, concise, and specific. Only dual mode modems are effected, and it selects CDMA over GSM by reversing the order of the conditional checks in the code, which previously preferred GSM over CDMA.

[Mathieu Trudel-Lapierre (cyphermox)]

Well, this is fix committed in upstream, but really not in progress for the Ubuntu packaging side.

It could be possible to ship that as a SRU fix in Maverick or Lucid. Let's keep it Triaged, then I (or someone else) can look into packaging this fix...

I'd really like to see some thorough testing on this though; since I'm pretty much convinced it would have an adverse effect on anybody using such a device outside North America or Japan, even if the chance of it happening seems pretty slim.

[Mathieu Trudel-Lapierre (cyphermox)]

This was uploaded in Natty in revision 0.4+git.20110124t203624.00b6cce-0ubuntu1; so marking as Fix Released.