Re: Performance Data for C14N2

Meiko


I have not had a chance to go through the details, but at the high level are the following conclusions correct?

1. Performance on large documents is improved with Canonical XML 2.0, despite the fact that the thesis implementation is not optimized, and is in ruby not C. This suggests a version in C with optimizations could perform better.

2. That said, detailed comparisons with existing libxml2 are not useful due to implementation language differences and concerns with the Ruby parser implementation used. 

How much concern should we have about the ruby parser used?

This is  work very helpful, since we would like to review the performance impact of the WG deliverables. I need to review it in more detail.

 I assume we can put this thesis draft on our public page of articles and you'll share an update when available?

Thanks

regards, Frederick

Frederick Hirsch
Nokia



On Oct 25, 2010, at 11:44 AM, ext Meiko Jensen wrote:

> 
> I uploaded a pre-final version of my student's thesis. The interesting
> matter starts at page 42, providing details on the performance impact of
> each C14N2 parameter. You can download it at
> 
> http://www.nds.rub.de/media/nds/downloads/mjensen/C14N2.pdf
> 
> For clarification: his implementation is done in Ruby, and uses a
> streaming-based parsing approach. Hence, impacts may differ for other
> implementations (see also p. 51+ for a comparison with default libxml2).
> 
> This should close my ACTION-654 .
> 
> best regards
> 
> Meiko
> 
> -- 
> Dipl.-Inf. Meiko Jensen
> Chair for Network and Data Security 
> Horst Görtz Institute for IT-Security 
> Ruhr University Bochum, Germany
> _____________________________
> Universitätsstr. 150, Geb. ID 2/411
> D-44801 Bochum, Germany
> Phone: +49 (0) 234 / 32-26796
> Telefax: +49 (0) 234 / 32-14347
> http:// www.nds.rub.de
> 
>

Received on Tuesday, 26 October 2010 13:50:53 UTC