[standards-jig] [jepnews] JEP-0020: Client Feature Negotiation

David Waite mass at akuma.org
Thu Feb 28 23:58:58 UTC 2002


Thomas Parslow (PatRat) wrote:

>
>Wouldn't it be better to extend the jabber:iq:browse namespace rather
>than creating a new one? 
>
I  am of the opinion that browse already does too much, in a 
poorly-defined way. :-)

>It seems that there is a bit of overlap
>between jabber:iq:browse and the namespace proposed in JEP-20...
>
browse currently provides a heirarchial structure for representing 
entities within a Jabber system, as well as some degree of introspection 
on those entities. So while browse could be used for feature 
introspection, that means little for feature negotiation. If I support 
four methods for file transfer, I can indicate all of those via browse 
only if all four have different namespaces representing each action. I 
cannot, however, indicate which I prefer for use - unless there is a 
globally understood, predefined choice in order. I believe pgm's case 
here is that we shouldn't have to define this for encryption, then keep 
redefining it for all future things to come down the pipe.

-David Waite




More information about the Standards mailing list