Ed,

My experience is with MSXMLDom COM object.  This for simple XSLT is quick
and works very well.

Apart from simple transformations, I really have no significant experience
apart from what I have played around with.

Paul

PS There is always the caveat that it is Microsoft and is not fully
compliant with the W3C standard. Use at your own risk!

> -----Original Message-----
> From: Edward Chowdhury [mailto:[EMAIL PROTECTED]]
> Sent: 07 November 2000 18:31
> To: Fusebox
> Subject: RE: Can fusebox and xslt play nice
>
>
> Yeah...I figured this wouldn't be an easy change to make. I am thinking of
> creating a master document DTD. The "body" of this document would contain
> any kind of other XML document each with their own DTD. So, I have an
> auction engine that will have it's own DTD, an ad engine that
> will have it's
> own DTD, a navigation engine DTD, discussion/review/message board DTD and
> probably a static document DTD. I'm right now working on the various DTD's
> which are based on the tables in my database. Once this is complete I will
> switch to using structures that represent the various DTD's to output HTML
> pages while I work on the XSLT/Cold Fusion integration part.
> Finally I will
> switch the Bodycontent tag to my new xslt implementation. After testing I
> will start targeting other platforms, probably Palm to start off with.
>
> In your experience Paul, how much of a performance hit do you incur when
> using XML/XSLT in a Cold Fusion environment as opposed to
> straight cfoutput
> of HTML? Also are there any scalability issues you've encountered?
>
> Thanks for your help and guidance...
> ed
>
> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Paul Johnston
> Sent: Tuesday, November 07, 2000 7:13 AM
> To: Fusebox
> Subject: RE: Can fusebox and xslt play nice
>
>
> Ed,
>
> Sounds very interesting. There are lots of problems as regards creating
> files for WAP and HTML and everything else too.  However...
>
> 1) I would modularise the navigation to start with (of course using
> fusebox).  This abstracts the navigation and makes it easier to
> update/change/add to a page.  I'm not sure about the benefit of number of
> characters per line or anything like that so I will ignore that.
>
> 2) Well, as we all know Cold Fusion can be used for ouputting any text you
> want.  So therefore, outputting XML shouldn't be a problem.
>
> 3) Certainly outputting a structure as XML isn't hard.  However, the
> processing may be a bit tedious.  If you know the structure keys then it
> should be very fast. Example:
>
> <?xml version="1.0"?>
> <mydoc>
>       <title><cfoutput>#mydoc.title#</cfoutput></title>
>       <body><cfoutput>#mydoc.body#</cfoutput></body>
> </mydoc>
>
> Have a look at the siteobjects (URL in 4 below) XML tags.  These put XML
> documents into CF structures, and I think they might go the other
> way (but I
> can't remember).  Follow the same kind of logic in reverse and
> you should be
> fine.
>
> 4) There are several ways you could do the last part.  Using the MSXMLDOM
> Object you can use the tags from http://www.siteobjects.com which are very
> good.  There is also stuff from http://www.granularity.com which is Java
> based.
>
> Bear in mind that CF_BODYCONTENT is a "process the middle bit", "get the
> header and footer", "put it together" tag.  If all you want to do
> is create
> an XML file and run it through an XSLT engine, then you can do that
> processing within that tag using the custom tags available (I suggest you
> rename it to "CF_XMLBODYCONTENT" though).
>
> As regards your free XML/XSLT stuff, Granularity use Java based XML/XSLT
> stuff and Siteobjects use Microsoft XMLDOM.  If you want correct
> implementation of the XML standard, then use the Java based tags.
>  The tags
> are probably the best way to go, although take a look at
> http://alphaworks.ibm.com for lots of free (mainly Java) stuff.
> I'm pretty
> sure that granularity already uses these though.
>
> Anyway, I hope that helps.  The idea is excellent, but I would suggest
> ensuring that you have your DTD's and Style Sheets down and
> definite before
> you even begin to code.  You will not be able to make large
> changes later on
> because of all of the coding that would be needed.  Also make
> sure that you
> completely understand how the XSLT engine works and how it processes the
> XML.  Knowing this will dramatically increase the effectiveness
> of the stuff
> you're doing (and I suggest you learn this before writing the DTD's and
> Style Sheets).
>
> Paul
>
> PS I hope I haven't put you off!
>
> > -----Original Message-----
> > From: Edward Chowdhury [mailto:[EMAIL PROTECTED]]
> > Sent: 06 November 2000 20:59
> > To: Fusebox
> > Subject: Can fusebox and xslt play nice
> >
> >
> > Hi All.
> >
> > I'm trying to change a web site I'n building from web only to also be
> > accessible from basically any internet capable device. I want
> to use some
> > variation of cf_bodycontent to generate the pages. This is how
> > I've thought
> > about doing it:
> >
> > 1) I set a struct of attributes for such things as number of
> > characters per
> > line and navigation style in the request scope that identify
> the physical
> > characteristics of the device.
> > 2) Use my dsp_ and act_ pages to create XML content rather than HTML.
> > 3) Create a request.bodycontent variable that contains a well formed XML
> > document based on a DTD that I specify. I don't want to use WDDX
> > because it
> > is too focused on handling different data types and I want mine
> to handle
> > different parts of a given page eg title, main heading, body, ad,
> > navigation,auction etc.  and be completely specific to my needs.
> > 4) Create a variation of CF_BODYCONTENT that takes the
> request.bodycontent
> > variable, runs it through an XSLT engine that will transmogrify
> > it into WAP,
> > or HTML or RIM or whatever.
> >
> > I just want to create an abstraction layer for the data.
> >
> > My question is are there any good, preferably free XSLT engines
> that I can
> > use as I'm writing this tag? Is there a better way of doing this
> > within the
> > fusebox environment. I don't want to get into the situation of creating
> > whole directories that do nothing but take existing functionality and
> > repackage it for other devices.
> >
> > ed
> >
> > PS Yes I am aware of the irony of wanting to support these technologies
> > after flaming away at them last week :) I still think HTML will be the
> > primary way of accessing information but little nuggets of VERY
> PERTINENT
> > information might be nice to have when you need them. I would
> also like to
> > think of a way of abstracting data from presentation thus leaving the
> > possibility of opening my web sites to whole universes of alternative
> > devices relatively easily.
> >
> > ------------------------------------------------------------------
> > ------------
> > To Unsubscribe visit
> http://www.houseoffusion.com/index.cfm?sidebar=lists&body=lists/fusebox or
> send a message to [EMAIL PROTECTED] with 'unsubscribe' in
> the body.
>
>
> ------------------------------------------------------------------
> ----------
> --
> To Unsubscribe visit
> http://www.houseoffusion.com/index.cfm?sidebar=lists&body=lists/fusebox or
> send a message to [EMAIL PROTECTED] with 'unsubscribe' in
> the body.
>
> ------------------------------------------------------------------
> ------------
> To Unsubscribe visit
http://www.houseoffusion.com/index.cfm?sidebar=lists&body=lists/fusebox or
send a message to [EMAIL PROTECTED] with 'unsubscribe' in
the body.


------------------------------------------------------------------------------
To Unsubscribe visit 
http://www.houseoffusion.com/index.cfm?sidebar=lists&body=lists/fusebox or send a 
message to [EMAIL PROTECTED] with 'unsubscribe' in the body.

Reply via email to