Re: MMS trials, getting insane trying to decode pdu's!

Hi Stipe,

On Monday, March 31, 2003, at 11:42 PM, Stipe Tolj wrote:

>> I'm looking at WAP-209-MMSEncapsulation-20020105-a.pdf and my hex dump.
>> As an example, I'm looking for the 'Message-Type' field (0x0c according
>> to section 7.3) which should be equal to 'm-send-req' (0x80, according
>> to section 7.2.14.). I would expect to find a sequence of 0x0c 0x80 but
>> it's not there.
>
> NOP!! that's the first gap all of us fall in. The Tabe 8 (sectin 7.3)
> tells the offset values to 0x80 as defined in WAP-WSP tables.

Hmm. How on earth would I have known that? :) I figured with about 3 
documents in front of me I'd have enough but nooooo! I also need the WSP 
spec (WAP-203-WSP-20000504-a.pdf in WAP 1.2.1 which is 
WAP-230-WSP-20010705-a.pdf in WAP 2.0)... Luckily that one has some 
examples on decoding.

Those specs are really hard to work with!


>> What am I missing? Is there (yet) another encoding going on that I'm
>> forgetting about?
>>
>> Here's the start of my hex dump:
>>
>> [MMS]0x8c       140     ?
>> [MMS]0x80       128     ?
>
> yep here you go: 0x8c is 0x80 (dec 128) as a base + 0x0c for the
> 'Message-Type' field. So these two tokens represent the Header
>
>   Message-Type: m-send-req

Right, thanks, at least I have a little head start now.


> and so on. These are the "easy" things. It gets harder when you don't
> have direct two pairs that represent the field name and field value.
> But as I said, you can do this allmost all with WSP code of Kannel.

I'm not sure what's better: trying to understand it with the specs or by 
reading the WSP code in Kannel :)

Thanks a bunch, gonna do some more decoding today

BTW, can anyone recommend a good resource on this (besides the specs 
themselves)? A book maybe or a web page? Better examples sure would help 
out here!

Bas.