btw., I forgot to mention earlier that the logs were only from 3g mode, as I'v recently moved out of Oslo which is the only place with 4g coverage. So since I've been to Oslo today, I finally got around to test 4G as well, and here's a dump of the traffic, initializing, connecting and disconnecting:...

Well, from a more complete (capturing bigger data packets), I discovered that some of the first traffic on the control port is actually a dhcp request: http://proyvind.net/~peroyvind/kalmia-init-connect-disconnect.txt (kalmia is btw. the chip name) Also by researching on LTE, I found several referen...

hum, after looking a bit more and testing a bit, I now see why the bulk input I was expecting to see doesn't appear after initializing and attempting to connect with the modem, it's obviously being sent to the other interface (control port) in stead. I guess in stead of ppp, binding it to an etherne...

yeah, from looking at the sniffing log, it doesn't seem to go online using ppp, but there are other things expected to happen which don't before getting that far.. Several of the commands sent to the modem first are also for initialization *only*, sending the same commands again once initialized, it...

Most of these commands are crucial for making the modem initialize properly, there's a couple of them that I from reading the 3GPP AT reference I'm uncertain about whether's actually required or not, but leaving them in for now won't hurt at least.. The dongle doesn't even have the ATDT command and ...

I did some more testing again and I now got the adapter working in windows under vmware which allowed me to sniff a bit further. First of all, adding a 'NeedResponse=1' to the config should be done. From the sniffing, I've put together what should be a correctly working chat script with the right ti...

ah, k, got it working now: http://svn.mandriva.com/cgi-bin/viewvc.cgi/packages/cooker/usb_modeswitch-data/current/SOURCES/usb-modeswitch-data-samsung-gtb3730.patch?view=log thx :) For the modem, I'm still having difficulties getting online with it, one issue I've figured out is when communicating wi...

hm, I'm uncertain how to do this in a proper manner though, trying to add a rule for the new id in udev rules, it won't kick in and act on it after the switch..

I've been succesfully able to communicate with it on /dev/ttyUSB1 now, but this requires sending messages to the new device after the switch has happened, which usb_modeswitch currently does not support. To achieve it currently, usb_modeswitch has to be run a second time with the following configura...

no, it won't unload the usb-storage module (I doubt that would cause the switch anyways, also as you may have more than device using the module, that wouldn't be very reliable, safe or anything... :p).

it will detach the device only (similar to eject)

hm, for my DWM-156 it's sufficient to just detach storage for it to switch mode..

Here's a simple patch with the config:
http://svn.mandriva.com/cgi-bin/viewvc. ... iew=markup

Here's a trivial patch that fixes a warning revealed when using higher optimization levels, it will also pass '-O2' by default if there's no CFLAGS defined so that such warnings will be catched right away in the future: http://svn.mandriva.com/cgi-bin/viewvc.cgi/packages/cooker/usb_modeswitch/curren...

Josh, unfortunately not, I haven't really had much chance to look closer into it either due to lack of time, hardware etc. to test with as I'm kinda in between places moving etc... IIRC the log I provided earlier should contain activity with the modem getting online as well, for anyone who wants to ...

Well, since the linking is mandatory to make it work, it shouldn't be overridable, and it's also a linker flag, not a compiler flag, so even if you would want for it to be overridable, it should've then be passed with LDFLAGS variable ;p This prevents anyone accidentally overriding it with ie. 'make...

yeah, I know the log is much larger than necessary, but since I had some issues with usbsnoop and getting it to log at the appropriate moment, I just started logging long before when I started the install since it didn't seem to pick up everything when I plugged it in after installing the driver in ...