RE: [xsl] Wikipedia on XSLT

Subject: RE: [xsl] Wikipedia on XSLT
From: "Michael Kay" <mike@xxxxxxxxxxxx>
Date: Fri, 7 Aug 2009 11:36:07 +0100
I think there's a tendency to structural decay in Wikipedia articles. This
one is certainly long due for some refactoring. It just needs someone with
sufficient expertise and time to do it, and then diligence to stop the rot
reappearing. Generally the problem in articles like this is not vandalism,
but (a) well-intentioned contributions from the ignorant, (b) blatant or
subtle promotion of commercial products or off-beat ideas, and (c) a poor
sense of perspective about what is important and what isn't (articles tend
to accumulate irrelevant detail). So you have to be prepared to put some
initial effort into the restructuring, and then be firm in preserving the
quality thereafter.

Obviously a balanced approach to 1.0/2.0 differences is part of the picture.
The XPath article got split into two, one for each version, because it just
seemed too difficult to cover both in a single article.

I think that it's best in an article like this to say very little about
specific products: at most, a list of links to implementations. But any such
list is a magnet for linkspam. A better approach might be a "category" for
XSLT implementations, and a link to the category.

I think an article about a programming language should tell people what kind
of language it is, why it is interesting, and where to find out more; and it
should be targeted at people who know about other programming languages but
not this one. It should not be a tutorial: the aim is to inform, not to
teach skills.


Michael Kay 


> -----Original Message-----
> From: Andrew Welch [mailto:andrew.j.welch@xxxxxxxxx] 
> Sent: 07 August 2009 11:17
> To: xsl-list@xxxxxxxxxxxxxxxxxxxxxx
> Subject: [xsl] Wikipedia on XSLT
> There's a push on xml-dev at the moment to sort out the XML 
> page on Wikipedia which is going well apparently, so maybe we 
> should have a look at the XSLT page :)
> There doesn't appear to be much in the way of XSLT 2.0 on 
> there - should that have its own page?  And Saxon doesn't get 
> mentioned apart from in the title of an article and in the 
> link to Kernow...  which can't be right!
> Also the top "See Also" link points to an "XSLT elements" page:
> which is pretty poor in general, but contains this example:
> <xsl:stylesheet>
> <xsl:template match="//input">
> <xsl:variable name="type" select="@type"/> <xsl:variable 
> name="name" select="@name"/> <xsl:if test="$type='text' or 
> $type='password' or $type='radio' or $type='checkbox'">
> 	<xsl:choose>
> 		<xsl:when test="$type='radio'">
>                     <xsl:if 
> test="not(preceding-sibling::input[@type='radio'])">
>                         <select name="{@name}">
>                         <xsl:for-each select="../input[@name=$name]">
>                                 <option value="{@value}">
>                                         <xsl:apply-templates/>
>                                 </option>
>                         </xsl:for-each>
>                         </select>
>                     </xsl:if>
> ...and that's loaded with faults.
> If nothing else I think we should remove that link, and 
> ensure XSLT 2.0 gets mentioned right at the top, with an 
> obvious link to Saxon (and other 2.0 processors)... and then 
> add examples for the most common use cases - copying all but 
> 1 element, grouping etc  (and the many others that I cant 
> think of now)
> Anyone available/interested to do this?
> --
> Andrew Welch
> Kernow:

Current Thread