Re: Review of EXI

David Orchard (31 oct. 2007 - 08:30) :
> The EXI WG did a huge amount of measurement work, referenced in my
> original message as
> http://www.w3.org/TR/2007/WD-exi-measurements-20070725/.
> Section 9.1.3 shows the Processing Efficiency Analysis Details
> http://www.w3.org/TR/2007/WD-exi-measurements-20070725/#Ax-details-pe
>
> Out of curiosity Karl and John, did you not see this section or  
> does the
> section not address your needs?  I made an assertion that the document
> is too detailed for any but the most diligent reader so you can offer
> your own data points on that.

Ah excellent.
Hmm that would be cool to have id on images :)
     <p><img src="summary-processing-encoding-sax-document.png"
         width="907"
         height="621" alt="Encoding summary: Document class"
     title="Encoding summary: Document class" /> </p>

I see in this graph XML, XML.gz and different techniques Xebu, FXDI,  
etc.
http://www.w3.org/TR/2007/WD-exi-measurements-20070725/summary- 
processing-encoding-sax-document.png

It seems to show in the graph, that there is no benefits over XML.gz.  
Please stand me corrected, if I misinterpret the graph. I was also  
wondering if XML+gzip performances compared to other were due to  
compression algorithm.

Using the same algorithm for EXI and gzip, are there differences of  
performances.

Interesting article too
http://www.ibm.com/developerworks/xml/library/x-matters13.html



--
Karl Dubost - W3C
http://www.w3.org/QA/
Be Strict To Be Cool