Re: Loading an external file of index

Subject: Re: Loading an external file of index
From: Paul Tchistopolskii <paul@xxxxxxx>
Date: Tue, 27 Jun 2000 12:38:29 -0700
>  I would prefer this requirement to be got rid of in the XSLT spec.
>  Instead, I would put a special PI (or something similar) telling the
>  XSLT processor that the document retrieved is ever the same.
 
>  What do you think about this? Am I brain-damaged or what?

I think you are not brain-damaged at all. 

I hardly belive that all off XSLT engines implement the required 
caching.

I mean if I try document('http://localhost/cgi-bin/randomizer.cgi') twice
in the same stylesheet - I don't think most of  XSLT engines will 
return the same result.  ( SAXON???  ;-) ).

I think that the behavior required by XSLT spec (  which requires XSLT 
processor to  cache document() ) is kinda consistent with current 
behavior that HTML browsers have  on images.

That also means it could have the same workaround : when you want 
image to be reloaded for sure - use ?<generated-timestamp-here>
trick.

Ah, this all makes some sense only for client side and even on 
client side, because XSLT specs  it is not reglamenting 
caching to happen between different  stylesheets - this
makes this 'caching' requirement almost useless.  

I wish you agree that even XSLT requirement  looks mad
and is not implemented in some ( if not all ) engines - 
this madness is consistent with the previous madness 
we got with images.

Rgds.Paul.

PS.

Please - don't think you are brain-damaged when you hit another 
one and nobody will answer some of your questions.  That's 
the way it usually happens with W3C.

For example, XML single-root limitation almost  kills redirects, 
concatenations, and  some  other  basic things which people 
do with documents for decades ( every time you are writing 
>> on your unix / DOS command-line - you are doing something 
which is not supposed to work in XML world ever ). 

But nobody will answer you what was the reason of single-root 
limitation in XML ( because there was no reason.  / in XPath 
points to a 'mythical' root anyway ).



 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list


Current Thread