[Standards-JIG] Re: Summary of roster proposal points

James Bunton james at delx.cjb.net
Wed Sep 8 06:50:10 UTC 2004


On Wed, 8 Sep 2004 02:32 am, Jacek Konieczny <jajcus at bnet.pl> wrote:
> On Tue, Sep 07, 2004 at 05:34:22PM +1000, James Bunton wrote:
> > Jacek Konieczny wrote:
> > > Fixing the latter is just a simple UI change ("Accept All" button below
> > > the contact list).
> >
> > While that would work, it's much better to give the client hints on what
> > to accept and what not too. There's a distinction between subscription
> > requests that are redundant (because the user has already authorised that
> > contact on the legacy service), and those that need to be looked at (new
> > users).
>
> Yes. It is not perfect. But your proposition is not perfect either. If
> we are to introduce new protocol not much better that current one (which
> is only badly implemented or not implemented at all) and which is
> suppost to be obsoleted soon by "the right solution" then it is better
> to fix current protocol implementation and wait for that "right
> solution". Without any hacks.

I've mostly answered this in my reply to Richard, but I'll summarise again for 
you.

I don't understand how JEP0093 can be made to provide nice interaction between 
gateways and clients (including clients that do not support the JEP).

And with regards to fixing the current protocol implementation, that's exactly 
what I'm doing =P
The current protocol (at least for the transports that I use) makes use of 
presence subscribe packets. I'm fixing the existing protocol with the 
addition of a simple XML tag in order to provide the extra information 
clients need to provide a good user experience.
I can see that you're suggesting fixing JEP0093 instead, but I don't see the 
advantage (only the disadvantage of not having absolute client support, as 
presence subscribe does).


---

James



More information about the Standards mailing list