EXI and XML Core

I had put EXI on our f2f agenda [1], and yesterday I
received email about trying to have some kind of joint
session at the TP on this topic.

In my reply, I mentioned:

> I have to say that many (perhaps most, though that might
> just be my impression) of the XML Core members are not
> in favor of what we still think of as "binary XML"

Daniel Vogelheim, of EXI, replied (in part):

> A really quick answer of what I suspect is a core issue here:
> The EXI WG positions EXI as an alternative (more efficient) 
> _encoding_, on pretty much the same level that gzip is sometimes 
> used as an encoding for transmission or storage of XML data. 
> In particular, feeding EXI to an XML processor is considered 
> wrong in the same way that feeding gzip-ed XML to the same 
> processor is. So in e.g. an HTTP transmission, EXI would
> show up as a Content-Encoding:-option, thus allowing the 
> default HTTP content negotiation to make sure that clients 
> or servers that do not explicitly agree to exchange using 
> EXI would not be bothered with it. 

which I thought a useful (if only partial) explanation.

Then just today, there was email on the tag list at
http://lists.w3.org/Archives/Public/www-tag/2007Oct/0151
about EXI that seemed to raise many concerns.

I forward all this info for the background information
of XML Core WG members.  Of course, any discussion on
our email list is more than welcome as well as any input
during whatever discussion we may have at the f2f.

paul

[1] http://www.w3.org/XML/Group/2007/09/xml-f2f-20071105-agenda.htm

Received on Thursday, 25 October 2007 14:39:26 UTC