xml-dev - RE: XSL Translations using Java Servlets / JSP

["Chris Bayes" <Chris@Bayes.co.uk>]

>
>> Checkout www.bayes.co.uk/xml if you have ie5 for a hypoteteical usecase.
>
><ie5_screen>
>"The XML page cannot be displayed
>Cannot view XML input using XSL style sheet.
>Please correct the error and then click the Refresh button, or try
>again later. "
>-------------------------------------------------------------------
>-------------
>Unexpected failure
></ie5_screen>
>

Yeh ie5 isn't that robust. It throws up many different errors on different
systems. I have a couple that are reproducible but aren't the fault of the
xml or xsl and are not reproducible on other systems. I haven't seen that
one though.

>This is from this website. Hard to reproduce. It appears
>( occasionaly ) in 'Announcements' and 'Vote '. It happened
>after I pressed on 5-6 hyperlinks. Don't think such a level of
>robustness  is in any way acceptable for something I call
>website - but probably it is.  ( When I'm saying "there is no
>Mozilla browser"  - I mean that there is no application wich
>is robust enough to be comparable to some other ...
>applications ... )
>
>Anyway. It is cool, I think.

Thanks a lot it is nice to get feedback, I think.

>
>I just don't udnerstand why can't you do the same with
>server-side rendering ?

Because I can't on the server I am using unless i resort to perl (which
wasn't even written when I made the site).

.. Anyway stylesheet is strongly
>CSS based...  And it will be reliable ;-)

Yes I like to use xsl for layout and keep CSS out of the xsl file. It is
just easier that way.

>
>I would like to explain my previuos letters a bit.
>
>No doubt, people can ( and usualy are ) using
>even crazy tools to build some things with those tools.
>They can even build a nice web pages inserting
>1x1.gif's into 'appropriate' place.
>
>Regarding the word 'hypotetical' - you are right
>and I'm wrong:
>
>People do use <?xml-stylesheet to bing XSL to XML.
>People  validate XML documents inserting <SYSTEM
>into well-formed XML document on the fly ( because
>'that's the standard way to invoke the validation if you
>don't want to pollute the document with hardcoded
>path to DTD's), people are suggesting things like:
>"place your variable outside the XSL stylesheet
>and then use document() to convert it into node-set,
>because it is somehow standard way in XSL".
>
>The <?xml-stylesheet provides some way of binding.
>But this way is not scalable. The scalable way should
>allow:

I agree but I like the binding as a default way of rendering.

>
>rendering any document with any stylesheet without
>changing the document itself an any point of processing.
>
>( validating any document with any DTD without changing
>the document itself  - but that's not as important as it is
>with the stylesheets ).
>
>And this 'realy-scalable-way' has nothing to do with
><?xml-stylesheet 'standard'  way of (hardcoded) binding.
>
>Rgds.Paul.
>
>BTW. I have looked at the 'internals' of  www.bayes.co.uk/xml
>and  I think that if you will have many  pairs like 'toc.xsl'  for
>'index.xml' ( but not 'index.xsl'  )

No I only have one toc.xsl for all pages. That is the way I intended it.
That way I only ever change one stylesheet. Same for CSS.

  - the site could become
>unsupportable mess pretty fast.  If you will have 'index.xsl'
>'index.xml' pairs - you don't need  <?xml-stylesheet at all,
>if browser will use 'favicon.ico-alike behavior".
>

Whatever that is?

>That was my point.

Ok.

>
>Not  that "it is impossible to build something
>with MS IE XSLT dialect". It is possible - sure. Your
>stylesheet will just fail with almost any other XSLT
>implementation. If you don't belive me - try rendering
>your xml files with your XML stylesheet with some
>XSLT implementation other than SAXON. ;-)

I know. It won't even render with SAXON I think. I use XT to build the rest
of my site www.bayes.co.uk before I upload it.
The reason I don't do that with the xml is because it is a demo site and a
showcase of how xml/xsl fits together no matter what flavour it is. In fact
it was originaly written to the ie4 xsl spec. As you know ie5 came out
before the spec was ratified. When ie5 supports the full xslt spec or
netscape is available I will change it to that ;-)

>
>I'm sorry - I think I'l not more participate in this thread.
>If something is still not clear - I am ready to explain
>it by mail.
>

No worries.

>
>

Thanks for the comments I appreciate your opinion.

Ciao Chris


***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@xml.org&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/
***************************************************************************