Re: Obstacles (?) to XSLT 2.0 in C++

i heard oracle going to rewrite his databases to java sometime

so why we should go against?

about XRX - yep once you do it you never go back

but i drop MarkLogic and start use eXist database

with new Java7 and G1 garbage collector i guess we not need so much C/C++

using perl/php/ruby for web-applications - can you tell at least one
reason except "we always did this so"?

2009/12/11 Andrew Welch <andrew.j.welch@xxxxxxxxx>:
>> Perhaps the world has also moved on from Apache HTTP server and PHP? B Is
>> that what you are saying?
>
> For what it's worth I'm currently working for Mark Logic writing
> applications where the XML Server is the back end, XQuery (plus
> extensions) is the application code, generating html/xhtml etc. B This
> works really well: you can quickly and easily create a web app that
> would previously require knowledge of Java, Spring, Hibernate, app
> servers and so on to create the equivalent.
>
> XSLT will be added to mix in the next version of Mark Logic, but is
> already supported in eXist, so the combination becomes XML Server back
> end, XQuery and XSLT as the application code (where XQuery supplies
> XML to the XSLT, although XSLT working directly against the database
> is hopefully in the pipeline), generating XHMTL/XHTML/XForms.
>
> The LAMP style equivalent acronym "XRX" kind of covers the above, if
> we can massage it to be "XQuery/XSLT, Rest, XHTML/XForms" :)
> (although I'm yet to be convinced about XForms)
>
> XSLT is absolutely essential to the mix, as XQuery's limited
> typeswitch struggles to emulate XSLT's recursive descent, or trying to
> generate any output in a namespace (such as XHTML or XForms) when your
> input is no namespace is a real pain... (the way xquery handles the
> default namespace is flawed).
>
> If you data is XML, then the benenfits of end-to-end XML from using an
> XML Server are massive.
>
>
> --
> Andrew Welch
> http://andrewjwelch.com
> Kernow: http://kernowforsaxon.sf.net/