Re[3]: patch for empy message and UDH > 140

Hello Oleg,

Monday, May 05, 2003, 7:12:37 PM, you wrote:

OT> Monday, May 5, 2003, 10:59:15 AM, you wrote:
AF>> On Montag, Mai 5, 2003, at 05:47  Uhr, Oleg Titov wrote:
>>> hi, please check out this patch used to solve the "empty" message and
>>> kannel smsbox crash on UDH > 140.
>>>
>>> First part of the patch checks for the presence of the text message
>>> and UDH header, now it should be possible to send pure UDH messages
>>> (without "<Empty ..." string as text).
>>>
>>> Second part, checks for UDH length, if it more than 140 - the reject
>>> message will be sent to http interface.
>>>
>>> -- 
>>> Best regards,
>>>  Oleg<smsbox.patch>

AF>> I think there was some other place in the code where an empty text
AF>> message is discarded or similar.
AF>> At least "should we allow empty text or not" was a discussion a while
AF>> back. Proper testing with every SMSC type would be advisable.
OT> Sorry, I am new on this list and I missed that discussion. Let me check
OT> in archives. I was looking for solution to send EMS melodies without
OT> any text message attached.

AF>> For the UDH, what happens if UDH is 139 and text is lets say 20 chars?
AF>> The text would  be split into two SMS but there's not enough space left
AF>> to add the segmentation header to the UDH.
AF>> Have you checked that case?
OT> That was out of my scope. I deal with a security hole for smsbox. If you send
OT> UDH grater than 140 octets than smsbox dies.

OT> In general, I feel that's responsibility of sender to prepare well
OT> formed UDH which takes into account split details.

Yep. You right.
I offer another solution to solve this problem. Get summary after
encoding of all data and if it is greater than allowable message
reject it.

---------------------------------------------------------
Jury Cherepanov
System Developer (Technical Director) , HIGH TECHNOLOGIES
Turaidas 2a, Riga, Latvia
Phone +371-9633398
mailto:[email protected]