OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help



   performant templating language for XML

[ Lists Home | Date Index | Thread Index ]


I'm going to ask the following question:  Is there a good templating  
language that can be used to build a high demand website from a collection  
of XML documents?

My assertions are going to be that XSL is not performant, that XQuery is  
not a templating language, and that no other well established language  
exists for creating web pages out of XML documents. I would very much like  
to see any or all of those assertions contradicted.

The influence of recent threads on this post is going to be obvious, at  
the same time previous threads have left some fundamental questions  

Let me give a little more background about the problem we're trying to  

We have a publishing system holding a great number of (sometimes on the  
large side for web use) XML documents.  The system is capable of  
generating additional meta data in XML format.

We are pretty well committed to serving pages from a java servlet  
container.  Each request requires one main transform where the XML is up  
to 80k (very occasionally very much larger, but those larger cases might  
be able to be pre-rendered), and up to two additional transforms on <10k  
docs.  We're looking at peaks of 10 users per second.

Assume that we have some ability to distribute this load among multiple  
servers but assume that this ability is limited in arbitrary but often  
constraining ways.

Now on to my assertions:

*** XSL is not performant
When answering requests, a templating language that forces the document in  
its entirety to be held in memory quickly hits load limits.  My  
understanding is that XSLT either cannot be streamed or current  
implementations are not optimized to stream when appropriate

Initial tests seem to show that XSL is a no-go in the load environment  
outlined above. Does anyone have production experience that would argue  
that XSL can be viable at these rates and that we should be looking harder  
for optimization strategies?

*** XQuery is not a templating language
I have no idea if I'm going to encounter disagreement with this idea or  
not.  Everything I've seen of XQuery suggests that the results of a query  
could be sent to a formatting engine that would apply a template to them,  
but not that XQuery would in itself be used to apply formating to a  
document.  I know that a previous thread asserted that XQuery can  
technically do anything XSL can do, but using it as a templating language  
looked to me like lining up camels and needles in an un-fun way.

*** There is no other well established templating language for formatting  
XML into web output
This isn't entirely true, we have looked at some other alternatives.  Jsp  
with DOM methods, sucks in comparison to XSL.  Anakia, yet another step  
down.  DVSL... don't get me started.

The most likely contender has ended up being: pre-render the main contents  
using XSLT on the docs within the content management system and cache and  
serve up the results.  In conjunction have a lightweight run time  
templating language to do navigation, authentication, and other aspects of  
UE that need run time processing.

Other options gratefully considered!

Thanks in advance!




News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS