W3C home > Mailing lists > Public > www-forms@w3.org > June 2002

RE: XSLT and XForms

From: Schulze, Matthias <schulze@dresden-informatik.de>
Date: Mon, 10 Jun 2002 10:38:32 +0200
Message-ID: <7140E412EB5ED411B8720050BF1123B019ADDE@magnum2.dresden-informatik.de>
To: "'www-forms@w3.org'" <www-forms@w3.org>

These are my experiences with GUIs and XSLT:

I've started creating client side GUI components for IE6 a few 
months ago.
First I used XSLT for the HTML generation. The transformation 
alone was really fast but the IE parsing and rendering took too
long, especially for large grids. 

Later on I found that XSLT was unpractical, especially for small
pieces of HTML. And when it comes to user interactivity changes 
to the HTML are really small!

At this time I had already started coding GUI changes as DOM 
manipulations. Suddenly I had two implementations for one 
problem: one as XSLT and another as JScript DOM manipulation. 
Thus I decided to abandon XSLT.
Now the 100% DOM version is not really faster, but more flexible.
IE parsing and rendering still takes about 70% of the response 
time. Perhaps M$ can improve that...

But anyway, IMHO DOM manipulation is the best way for 
implementing rich client side GUIs in a browser. 



> -----Ursprüngliche Nachricht-----
> Von: Dan Dennedy [mailto:DDennedy@digitalbang.com]
> Gesendet: Freitag, 7. Juni 2002 17:31
> An: Stefano Debenedetti
> Cc: Lambert Charlie; www-forms@w3.org
> Betreff: RE: XSLT and XForms
> I think server-side processors will represent just the first 
> generation of xforms implementations due to timing. It just 
> makes too much sense to start leveraging it to formalize some 
> mundane aspects of web app development while still being 
> compatible with today's browser technology. I am very 
> optimistic that XForms will someday appear in mozilla. I 
> would love to start working on it, but my independent 
> programming is totally consumed by multimedia on GNU/Linux 
> where there is much work to do! I'm searching for a way to 
> get my employer or a client to invest some effort into 
> Mozilla. At least my server-side processor would serve as practice.
> > -----Original Message-----
> > From: Stefano Debenedetti [mailto:sdebenedetti@e-tree.com]
> > Sent: Friday, June 07, 2002 11:01 AM
> > To: Dan Dennedy
> > Cc: Lambert Charlie; www-forms@w3.org
> > Subject: Re: XSLT and XForms
> > 
> > 
> > This is the very only reason that makes me find it hard to 
> believe a 
> > server-side XForms processor would spare me some debugging time.
> > Despite the nice, declarative approach, applications 
> > generated this way 
> > would be far more difficult to mantain and test than 
> > applications that 
> > are run on the client. Not to mention that serious (consistent) 
> > client-side interactivity (as allowed by the wonderful 
> XForms WD for 
> > example) is something the web is sorely missing, not only where the 
> > connection speed is bad.
> > Anyway good luck to all the server-side XForms processor 
> that are on 
> > their way, I really hope they will save me some debugging time!
> > ciao
> > ste
> > 
> > 
Received on Monday, 10 June 2002 04:41:50 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:36:06 UTC