|> This means a restructuring of the logic of smsc_at wrt to reading and
|> writing
|> to the modem. My concern is that if I take this on, I could
|easily introduce
|> as many problems as I fix. I have already experienced this when I "fixed"
|> the
|> way that read timeouts occur, and found that I had created
|another problem.
|
|Well, Alex has already embarked on that journey, and I was hoping to be
|able to apply his latest and greatest patch against the smsc_at code which
|will hopefully make things much more comprehensible. Alex?
Yes, I want for exactly that structure in order to add functionality that
gave more complex responses than simply OK or add PIN. Basically I added a
new routine called raw_modem_command which allows you to pass an octet
string through as a parameter to recieve the response back from the modem.
This way you can check it for whatever you want.
This is currently in production on our SMS modem systems in London and New
York and seems to behave quite happily (caveat as happily as Kannel ever
did!). The diff between the CVS release and this modifications is available
on our site at http://195.188.176.104/diff.patch.Z and the modified files in
completeness are at http://195.188.176.104/kannelfiles.tar.Z
What would be good is if you take these files and modify from here then it
should be much quicker for you and it would be good to get a few more eyes
wondering over all the code changes. Then Nick/I/You can merge back into the
Kannel parent tree for the next stable release.
Any questions you have feel free to give me a shout.
Alex
Alex Judd
CTO
Skywire
Mob. + 44 (0) 7971 052 642
