RE: First working draft of XSL

Subject: RE: First working draft of XSL
From: Mark_Overton@xxxxxxxxx
Date: Tue, 18 Aug 1998 14:04:23 -0400
Yes, it is more concise.  But it has alot of problems which I don't know if
they thought about.
1. No way to enforce XSL structure using a DTD.
2. Need for a new parser.
3. Learning curve (now you need to know XML and this new syntax)
4. No way to access the components through XML.  For example, how could I
get the target element type for a rule using standard XML parser methods?

The thing that really bothers me is this, if every user of XML starts
coming up with grammer like this then the whole purpose of XML is
destroyed.  We finally are getting a standard data format and even in the
W3C stylesheet standard for XML we are going to come up with another
syntax.  If other XML application look like this then we should just forget
the whole thing and go back to proprietary formats.

Imagine the usefullness of xml documents that look like this
<Person Name="Mark|Overton" Attitude "NeedsAdjusting('a little')"
Hobbies="*\Fishing | Coming Up With Silly Examples"/>

The most telling example of why we shouldn't be building XML documents like
How would you ever create an XSL stylesheet for and XSL stylesheet?  The
reason it would be nearly impossible is because this is not really good

-Mark Overton

David Schach <davidsch@xxxxxxxxxxxxx> on 08/18/98 01:51:59 PM

Please respond to xsl-list@xxxxxxxxxxxxxxxx

To:   "'xsl-list@xxxxxxxxxxxxxxxx'" <xsl-list@xxxxxxxxxxxxxxxx>
cc:    (bcc: Mark Overton/PTSLS)
Subject:  RE: First working draft of XSL

The original XSL submission used XML for the patterns.  However, the new
pattern syntax is much more concise and easier to read than an XML based
pattern syntax.

> -----Original Message-----
> From:   Mark_Overton@xxxxxxxxx [SMTP:Mark_Overton@xxxxxxxxx]
> Sent:   Tuesday, August 18, 1998 10:30 AM
> To:     xsl-list@xxxxxxxxxxxxxxxx
> Subject:     Re: First working draft of XSL
> My first thought is this:
> Why did they not use XML for the structure of the patterns, etc.
> For example,
> Here is a rule example from the new spec
> <xsl:template match="book[excerpt]/author[attribute(degree)]">
> ...
> </xsl:template>
> This could have been something like:
> <xsl:template>
>      <match>
>           <element type="book">
>                <element type="excerpt"/>
>                <target type="author>
>                     <attribute name="degree"/>
>                </target>
>           </type>
>      </match>
> <action>
>   ...
> </action>
> </xsl:template>
> This way the xsl processor could read the stylesheet without having to
> parse all of this new syntax.  We have a great tool in XML for
> representing
> structured data so why did we have to come up with another?  Now, to read
> an XSL stylesheet I need to parse all of these new delimiters and more
> ('/'
> | '//' | '(' | ')' | '|' | '[' | ']' | ',' | '=' | '.' | '..' | '*' | '{'
> |
> '(' |, etc.......).  All of the built in functionality of my XML parser
> of no use.  What a shame.
> -Mark Overton
>  XSL-List info and archive:

 XSL-List info and archive:

 XSL-List info and archive:

Current Thread