* sms-split:
smqueue: Add SMSC short code to smqueue.config.example
Checking in forgotten smsc.cpp and smsc.h files.
Ignore User-Data-Header in SMS TPDU when decoding text in TLUserData::encode7bit().
Transparently pass TP-UHDI (User-Data-Header-Indicator) from SMS-SUBMIT to SMS-DELIVER.
Slightly cleaner output in TLUserData::write().
Log contents of a message instead of its memory addresss in SMS::parseTPDU().
Use decoded text from original message in bounce and e-mail messages.
More readable bounce message.
Maximum SMS length is 160 symbols in default alphabet.
Avoid duplication of "IMSI" in SC.Register.Msg.
Set Log.Alarms.* values in smqueue config.
Make "from" address for bounce messages configurable.
Move most of the SMS processing to smqueue (initial check-in).
Better documentation and error reporting for ThreadSemaphore.
Implementation of Semaphore.
Implemented "trans" CLI command to list all active transacions.
Idea is to make OpenBTS as dumb as possible. It should forward all received SMS messages' RPDU to SMSC (smqueue) and vice versa. All actual message decoding and processing is to be done in smqueue.
Certain phones with GPRS capabilities will fail to
initiate calls with the following error. No call output
will be present in the Asterisk log.
1265648543.0255 INFO 3073350512 GSML3Message.cpp:162:parseL3: L3 recv RR GPRS Suspension Request
1265648543.0257 NOTICE 3073350512 DCCHDispatch.cpp:144:DCCHDispatcher: UnexpectedMessage
The issue and patch were discussed on the openbts-discuss
mailing list.
"David is right that openBTS does not emit any GPRS beacon,
so the behaviour is triggered from mobile station which
believe that there has been some GPRS before (i.e. in it's
old location area).
Unfortunately openBTS does not simply ignore this L3 RR
GPRS Suspension Request, instead a expected message flow
sequence (i.e. during call setup) is terminated with an
exception.
However we created a small patch against openbts-2.5.3
Lacassine to really ignore the L3 message. Feel free to
distribute it upstream to the community."
The author of the patch is >> Florian.Wolff@Siemens.com <<
According to his statement, O2 XDA Orbit and Motorola
Milestone are working fine with those changes. They had
the same problems before as with the iPhone.
Reported-by: Michael Folz <michael.folz@fh-kl.de>
Signed-off-by: Thomas Tsou <ttsou@vt.edu>
Originally there was a fixed-size char array, which didn't allow to work with long hex strings.
(cherry picked from commit fdc6fc85f9f431438ef7279975c2f7c230cb4aa5)
Best practice is to not include those in repositories but only
in tar.gz dist tarball.
autoreconf -i will regenerate them
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
Ivan Kluchnikov
Sylvain Munaut