Hi!

CCing NetworkManager mailing list. I'm sorry that the thread below is not 
formatted the way you like it but I'm keeping it below for reference.

The discussion is about sms support in ModemManager. I've started implementing 
it a bit but ran into issues.

I think that a SetStorageLocation method as suggested by Eric would be sane to 
add. And have that mirror the CPMS command as Jason suggested.

Should I still include the mem-type in SmsReceived signal? Or should we assume 
that the caller takes care of setting the storage correctly? The same question 
is valid for the SmsGet method (and delete).

About the SmsReceived vs Completed signals, I interpreted it as the SmsReceived 
signal is emitted for all sms and for the final sms _both_ the SmsReceived and 
Completed signals are emitted. I don't have the code available right now 
though, so I'll have to look at it again later.

Another issue is how to handle the completed flag in the SmsReceived signal and 
the Completed signals. I suppose one have to parse the pdu to be able to tell? 

Regards
Torgny

________________________________________
Från: Jason Glasgow [[email protected]]
Skickat: den 21 mars 2011 15:32
Till: Eric Shienbrood
Kopia: Torgny Johansson; Jonas Sjöquist; Malin Gustafsson; Trond Wuellner; 
Nathan Williams
Ämne: Re: Ericsson Support for SMS receive

Hi Torgny,

I'd like to move this discussion to the nm list, but before I do so I want to 
make sure that you are okay with that.  If you agree, please CC the list on the 
next reply.  If not, let me know.

I'm somewhat partial to having SetStorageLocation and GetStorageLocation either 
take three arguments to mimic the CPMS command, or just set the location of the 
received messages and allow the List,Get (etc) commands to call CPMS (if 
necessary) on each invocation.

-Jason


On Mon, Mar 21, 2011 at 9:55 AM, Eric Shienbrood 
<[email protected]<mailto:[email protected]>> wrote:
Hi Torgny,

I don't think one would typically want to change the storage location on a 
call-by-call basis. So it probably makes the most sense to have a separate call 
as you suggest for setting the type of memory to be used. Maybe we could pick a 
name like SetStorageLocation. It could set the storage to be used for Get, 
Save, List, and SendFromStorage methods. I think we should also have a 
GetStorageLocation method.

Also, the spec says that a "Completed" signal is supposed to be send when all 
parts of a message have been received. From the description, it sounds like 
SmsReceived is supposed to be sent when the first part is received, Completed 
is supposed to be sent when the last part is received, and no signals are sent 
for any parts between the first and the last.

Eric

On Sun, Mar 20, 2011 at 1:20 PM, Torgny Johansson 
<[email protected]<mailto:[email protected]>> wrote:
Hi!

I started writing a more proper implementation but ran into some questions...

I extended the SmsReceived signal to also include the memory where the sms is 
stored (but not the actual sms). Then I started implemeting the GetSms method 
(also extending it to include the memory type as an argument). The issue I ran 
into is that the at command to read messages doesn't use the memory type, only 
the index. The memory type should be set with the CPMS command. So should the 
GetSms method set that every time or should we instead add a SetMemory method 
to do that and remove the memory type from the GetSms method again?

Please let me know what you think.

Thanks!

Regards
Torgny


_______________________________________________
networkmanager-list mailing list
[email protected]
http://mail.gnome.org/mailman/listinfo/networkmanager-list

Reply via email to