[SIPForum-discussion] [Sip] SIPit 20 survey summary

[Henning Schulzrinne]

For the record, I will note that I have made proposals for simple non- 
multipart location conveyance (using the data: URL), but various  
process-related arguments were made as to why we weren't allowed to  
look at that. (I'd prefer even simpler solutions, such as the one  
that XMPP uses, but that's beyond the political correctness limit in  
GEOPRIV.)

I tend to agree that exhortations to developers generally achieve  
little. On the other hand, I'm not sure that belly-aching about  
multipart is all that helpful. After all, most email clients support  
it and there are libraries in various languages to help with  
implementation. Generating multipart bodies is pretty trivial (as  
opposed to parsing them), and that's all embedded devices will  
generally have to do for location conveyance.

On Apr 29, 2007, at 9:36 AM, Jeroen van Bemmel wrote:

> Hi Hannes,
>
> I was responding to Brian's message. He basically says: SIPit  
> results show that the mechanisms needed to implement location  
> conveyance are not widely implemented yet, we need to tell  
> implementors to hurry it up
>
> I question whether that approach works, and turn it around by  
> asking: why are these mechanisms not implemented widely?
>
> Personally I believe a significant part of the answer lies in the  
> complexity of the proposed mechanisms, architecure, etc. So instead  
> of trying to push the market to adopt the solution that is now on  
> the table, perhaps we should look into what we can do to lower the  
> barriers for adoption
>
> Regards,
> Jeroen
>