RE: Ericsson T68 as SMSC

-----Original Message-----
From: Bruno David Rodrigues [mailto:[email protected]]

>B. Now, with +CNMI=3,1,0,0
>2002-02-26 01:59:06 [6] DEBUG: AT2[/dev/ttyS9]: <-- +CMTI: "ME",27
>2002-02-26 01:59:08 [6] ERROR: AT2[/dev/ttyS9]: got +CMT but waiting
for next
>line timed out
>
>Uhmm.. right. the "1" says that the message goes to Mobile Memory and
we need 
>Matt's patch.
>But kannel shouldn't interpret CMTI as a CMT anyway.

That is correct - like I said in my previous email, both fixes are
needed.

>Could you please tell me the state of Matt's patch (or the new one from
Odded)
>so I could help them ?

The at2 patch is 100% safe as far as I'm concerned, but - thinking about
it again - I could clean it up by removing unnecessary #ifdefs (ones
that do not affect logic directly, for example - the change to the
modems' data structure), so it'll be prettier. 
Or better yet - change the patch to use a configuration parameter
instead of #defines - will that be ok ? it will require major surgery in
at2 code - I'm now confident in my ability to do so w/o breaking too
many things <wicked grin>, but before trying, I'd like to hear the
developer's opinion on that (and should I use an existing configuration
param or invent one of my own, and if the latter - how to call it).

I can also resubmit Matt's AT(3) patch - I've cleaned it up a bit and
changed it to use  a #define (so the changes will be turned off by
default) - I did that as preperation for porting the changes to AT2.

Oded Arbel
m-Wise Inc.
[email protected]

--
Contrary to popular belief, the world is not ASCII. 
	-- from the xemacs 19.12 documentation on define-key