[JDEV] Protocal Specifier on jabber addresses; creation of JUPLs
Sanjay Ratnalikar
sanjay at pinnacleinfo-sys.com
Fri Nov 23 22:24:39 CST 2001
-----Original Message-----
From: jdev-admin at jabber.org [mailto:jdev-admin at jabber.org]On Behalf Of
Julian Missig
Sent: Saturday, November 24, 2001 12:57 AM
To: jdev at jabber.org
Subject: Re: [JDEV] Protocal Specifier on jabber addresses; creation of
JUPLs
The problem I have with treating communication with other instant
messaging systems as other protocols is that, well, they're not. We're
still speaking Jabber. We're just speaking jabber to a server module
which then translates it. We'd have to have an infinite number of
protocol specifiers for the infinite possibility of server modules.
What happens when you're already in a multi-protocol environment? In
mozilla (with proper hacks), if I want to do something over jabber, then
it's jabber:user at server/resource. What happens when I want to send
something over jabber to msn? jabber:msn:user at server/resource is not a
valid URI. That URI means username msn with password user over protocol
jabber. Gabber already interprets jabber: as separate from mailto: for
jabberids vs email addresses... and I believe jabber: is the proper
protocol specifier for Jabber.
It would be nice to be able to specify that something is using a certain
server module in the jabberid itself, but using a protocol specifier is
not the solution.
Julian
--
email: julian at jabber.org
jabber:julian at jabber.org
Richard Dobson wrote:
> As a little project of mine I have written my own instant messaging system
> which uses ideas from the different IM systems that I researched, after
> seeing jabber I based my protocol on XML and a lot of the XML packet
> structures are similar, but I do some things different, i.e. the
addressing.
>
> e.g.
>
> I represent native users of my system in a similar way to email and jabber
> id's like this.
> me at domain.com
>
> But addresses of users on different systems work as follows.
>
> MSN Messenger
> msn:user at hotmail.com
>
> AIM
> aim:screenname
>
> E-mail
> smtp:user at domain.com
>
> I got the idea for this from web and ftp addresses where addresses are
> prefixed with the protocol identifier such as http: or ftp:
>
> All of the actual communication with those users is tranparently handled
at
> the server end by transport servers which work like the jabber transports
> establishing connections and translating the protocols between the
systems.
> Also they can be duplicated as many times as needed to handle the load and
> the routing scheme distributes new sessions to the least loaded servers as
> well as messages for existing sessions to the correct server.
>
> Anyway after that ramble thats my thoughts on IM addressing.
>
> Rich
>
>
> ----- Original Message -----
> From: "Matthew Bell" <matthew at bells.mcb.net>
> To: <jdev at jabber.org>
> Sent: Thursday, November 22, 2001 1:35 PM
> Subject: [JDEV] Protocal Specifier on jabber addresses; creation of JUPLs
>
>
>
>>I'm sure this has probably been suggested before but perhaps it would be a
>>nice idea to add protocol specifiers to addresses (JabberIDs), to create
>>Jabber Uniform Personal Locator. This means we could access different
>>instant messaging architectures without pretending (i.e.. setting up
>>pseudo-servers to sit between jabber and the other protocols) that they
>>
> work
>
>>in the same way that jabber does (an e-mail like set-up), which strikes me
>>as an ugly way of doing things. This means that other protocols will be
>>treated as other protocols, and seems a much nicer way of dealing with
>>things. Of course this probably will mean a major rewrite of some things,
>>
> so
>
>>is probably one for the really far into the future list, or the to hard to
>>do list (which is situated in the bin).
>>
>>Matthew Belll
>>
>>ps. that was a bit of a ramble wasn't it. I don't think I really got my
>>point across.
_______________________________________________
jdev mailing list
jdev at jabber.org
http://mailman.jabber.org/listinfo/jdev
More information about the JDev
mailing list