Roger,

I'm coming late to this thread, but I had a few thoughts on this as well.

I think that working with an XML *pipeline* has a lot of advantages, but
realistically, regardless of whether you use XQuery or XSLT, at some point
you'll probably want to bind Java or other external objects to an
appropriate namespace. If I wanted to do Fourier transforms on a set of
data, for instance, while it is (barely) possible to create a native XQuery
library for doing such, it would be hideously slow. On the other hand, if
you created an FFT module in Java that bound the class to an XQuery or XSLT
namespace, then I can see a fairly strong argument for agreeing with you.

I've been following XProc and Norm Walsh's efforts in that regard because I
think that a standardized pipeline language can get you a long way in
building most systems. The key to me seems to be intelligent binding.

On imperative languages - I think that you can in general replace most
imperative "tasks" with declarative stateful diagrams, but its a mistake to
think that there is a one-to-one equivalency there. You have to take
different approaches with a declarative foundation than you do with an
imperative one - even to the level of rethinking architecture. Since most
programmers don't tend to think at the architectural level (a rather
frustrating insight, actually) this means that most programmers who look at
XML technologies try to build imperative solutions using a declarative
hammer, and not surprisingly it just doesn't work.

-- Kurt Cagle
-- Editor, xml.com
-- O'Reilly Media



On Mon, Dec 1, 2008 at 7:38 AM, Costello, Roger L. <costello@m...>wrote:

>
> Hi Folks,
>
> I am exploring the idea of "do all application coding in the XML
> languages."
>
> Here is a response from a colleague:
>
> "... in general XSLT is cool but limited. If your transform requires any
> "higher math" or advanced functionality or external code libraries (such as
> geometry coordinate system libraries), you almost always have to go back to
> a higher level language (such as Java) at some point."
>
> Does my colleague make a TRUE or FALSE statement?
>
> /Roger
>
> _______________________________________________________________________
>
> XML-DEV is a publicly archived, unmoderated list hosted by OASIS
> to support XML implementation and development. To minimize
> spam in the archives, you must subscribe before posting.
>
> [Un]Subscribe/change address: http://www.oasis-open.org/mlmanage/
> Or unsubscribe: xml-dev-unsubscribe@l...
> subscribe: xml-dev-subscribe@l...
> List archive: http://lists.xml.org/archives/xml-dev/
> List Guidelines: http://www.oasis-open.org/maillists/guidelines.php
>
>


-- 
Kurt Cagle
Managing Editor, xml.com
O'Reilly
kurt@o...