faucamp
commented
11 years ago Reposting most of my response from the previous thread :)
Thanks for the (very) thorough investigation! Your modem's behaviour in this regard is very strange; I've seen slow SIM cards and slow modems before, but it shouldn't give a 330 error if it is waiting for the SIM (or something else) - that's what the "_writeWait" attribute and automatic handling of CME 515 (device busy) and CME 14 (SIM busy) errors are for (in write()).
I'll try to come up with something that handles this in a similar fashion as the 515 and 14 errors, but for that I need more output from you (so that I can modify the modem profile I'm using for unit testing). Could you please modify that script you posted to print out the exact response from the modem instead of 'retry get smsc'? You'll need to change your while loop's condition as well. Try something along the lines of:
starttime = time.time()
print("start time %i"%(starttime))
while True:
response = self.write('AT+CSCA?', parseError=False)
print('response:',response)
if response and response[-1] == 'OK':
break
print('got smsc')
print("took seconds: %i"%(time.time() - starttime))
currentSmscNumber = self.smscThanks (and thanks for your patience!)
dauheeIRL
Apologies if I'm causing confusion I've opened up a fresh issue for what I've experienced. The code for stopping the SMSC from being wiped was perfect!
I was getting confused due to unplugging the modem sometimes, and also because self.smsc caching, changing versions of code, different timings because of me stepping through code in debug mode etc . . and most of all for me rushing!!
On first loading of this application, I still get a 330 no smsc error on the following line (175 of modem.py) of the latest code: currentSmscNumber = self.smsc
however on the second execution (while modem still connected) the error is not there. I believe this issue is due to an inital slowness of connection to the network, causing the SMSC not to be available for a few seconds. And because the AT+CSMP command needs the smsc somehow (and wipes it on me), the error was generating at that point for me.
So based on a possible timing issue, I tried the following grotesque test code after freshly plugging in the modem:
and got the following output: Initializing modem... start time 1367503949 retry get smsc retry get smsc retry get smsc retry get smsc retry get smsc retry get smsc retry get smsc retry get smsc retry get smsc got smsc took seconds: 9 Waiting for SMS and/or calls...
So it appears for my modem + network combination it is taking about 9-10 seconds for the SMSC to become available. By leaving in the modem, the next time I start the application, it takes 0 seconds - this means there will only be a delay for slower modem/networks and shouldn't impact other users of this code. I used the following at different points in the connect:
print("step 1 %s"%(self.write('AT+CSCA?', True, 5, False)))
and it proved that the SMSC just takes time to become available after the first execution, and is available from the start on the second execution.
So getting the currentSmscNumber before AT+CSMP and setting it afterwards was perfect. This has now moved onto another issue for me. You were totally right about the shotgun :)
Apologies I will leave you alone now and stop interfering :)