[XSL-LIST Mailing List Archive Home] [By Thread] [By Date]

RE: [xsl] :o) (Re: qualitative decline of xsl-list questions)

---

---

>XSLT has now
>been in existence for six years, formally, which means that it is now
being
>used by people who have absolutely no clue what they're working with,
have
>no real grounding in the technology or its imperatives, and are looking
for
>turnkey solutions from vendors.


I'd think a lot of adapters probably start off as generalists. Or to put
it another way, some generalists have it in them to become adapters;
maybe even higher. I mean that I would hope that any generalist is a
generalist by necessity but has some skills in which they could well be
described as a specialist, expert, adapter, what-you-will. This is in
fact what applies to me, I am in many ways a generalist over a wide
array of disciplines, but every now and then I find a particular
discipline that appeals to me and I begin to learn it in the depth. 


>"Oh, you must be one of the five people on the planet who know how to
work
>with XSL."

This would seem to be a weird comment coming from Microsoft; I can
understand you getting floored.



>In order to set XSLT
>parameters, you have to instantiate a secondary XPathNavigator object
to
>retrieve an object that lets you assign parameters, that then needs to
be
>passed as an argument into the Transformer object -- it's
understandable
>from a class perspective, but is so friggin complex that you REALLY
need to
>understand what you're doing just to do what should be a simple action
(and
>IS in the corresponding Transformer class in Java.

I don't much care for the use of XPathNavigator in .Net to assign
parameters, but not for the reasons you name; actually for what I've
done so far which admittedly has not been especially difficult (I tend
to add my parameters as a nodeset of parameters), I can't see much
difference between creating a Navigator and passing it an
XsltArgumentList and creating an object "msxml2.xsltemplate.4.0",
loading a DOM into that object, creating a processor and then passing
that processor my parameters.
The thing which really bugs me .Net vis-à-vis passing of parameters is
that when one passes a nodeset - so far as I have experience - you
cannot access that nodeset with node-set(), instead you query it
straight on with $paramsNode/params/param and so forth, which irritates
me because I see it as creating a debugging problem, in that if you
create an inline nodeset, again in my experience using .Net, you have to
access it via node-set(); it's those kinds of things that I think
demonstrate an imbalanced logic on the implementation side, and that for
me at least makes it harder to design a good solution in that
environment where xslt is an integral part of the solution.





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

---