[standards-jig] JEP-0067

Peter Ronez prnz404 at yahoo.com
Mon Jul 28 13:46:59 UTC 2003


--- "Ulrich B. Staudinger" <chicago5 at gmx.de> wrote:
> Hi Peter Ronez,
> 
> i hope you had a great weekend.

Not exactly :)

> >My largest objection to the JEP is that it's not Jabber.org job's to
> >standardize on XML formats for data not even remotely related to instant
> >messaging. There are plenty of other standards bodies with domain specific
> >knowledge that are coming up with representations for the data. So aside
> from
> >the technical problems with multicasting financial data over XMPP, it's
> still
> >not a good idea for Jabber.org to produce its own data format for everything
> >under the sun.
> >
> Name me one sufficient xml standard for stocks and i will look into it. 
> Neither the microsoft xml standard from 1997 nor most of the other 
> examples out there do half a good job for stock quotes as this jep. This 
> jep is very fundamental, it is afaik the only one of it's sort on/in the 
> internet.
> 
> I agree it's not j.o's job to produce data formats and j.o doesn't 
> anyway, but it's the the best place to put the jep for public discussion 
> and to secure the document. XMPP fits well for financial data delivery.

Well, to be fair I didn't know any XML standard for market data either so I had
to go look for it. It doesn't seem that people are too interested on
standardizing market data. Anyway, take a look at 

http://www.mddl.org/mddl/2003/2.2-beta/ 

Other things of interest:

http://www.fpml.org (Mostly for Fixed Income products)
http://www.rixml.org
http://www.fixprotocol.org

Anyway, regardless of the fact that there may or may not be a suitable XML
standard for market data out there, it does not imply that Jabber.org ought to
standardize on a format for market data. If anything you can form your own
standard on your own website. XMPP is a transport layer and should try to be
agnostic as possible about that the data that goes across it. If Jabber.org
were to standardize on data representation which it doesn't have any domain
specific knowledge about, it will more than likely detract others from using it
if they happen not to agree with the current specifications for how to
represent their data. Imagine if HTTP worked in this way.


> I think about moving the jep to an informational track. However, i am 
> quite sure it is gone from the jep list in a year, if i do so.
> 
> A word about multicasting/Rendezvous: I agree, multicasting is good and 
> has it's advances - be it for video, audio or groupchat data 
> transmission, but multicasting is not as supported as it should be. I.e. 
> over here in germany we have no non-university DSL provider who supports 
> multicasting. Internet projects, as the MBone, did not succeed because 
> of this reason. Traffic would have exploded many times earlier, 
> therefore i see no real use for multicasting except in a multicasting 
> environment - be it installed manually with software routers or hardware 
> installation in intranet environments.
> 

Multicasting is only practical inside a corporation's LAN or WAN. 



__________________________________
Do you Yahoo!?
The New Yahoo! Search - Faster. Easier. Bingo.
http://search.yahoo.com



More information about the Standards mailing list