<snip>
...
</snip>
I think XSLT is quite good for non-programmers, it's programmers with
some experience of other things that find it really difficult I've
found.

I always find it very surprising that programmers with experience only of 
imperative languages find taking up a declarative language difficult - 
after all we were all taught declarative semantics at school - in our 
mathmatics lessons.



But it is a very common experience - can anyone suggestion why this might 
be?



I teach XSLT and also find this to be true.



Why?



I think it is a case of ... "when you are a hammer - all problems are 
nails".



A programmer "writes code" ... so the solution to all problems must be 
"writing code".



They can't seem to get it that an XSLT processor executes and has default 
behavior - which can be modified by "templates" you supply for each node in 
the input tree of nodes.. and that they are called templates since you 
provide a template of what you want as output.... you don't write procedural 
code to generate it.



Not a new problem of course... In the late 60's in Bell Northern Research - 
we were using something called "Mark IV" that was a generic solution to the 
classical "posting problem".  it had default behavior that you could modify 
by supplying the equivalent of templates that would be invoked at specific 
points in the processing.



The "programmers" could not get it through their heads that this default 
processing was there and already provided a solution to the posting problem. 
Instead they found that you could exit MarkIV via some assembler level 
call - and invoke an external procedure.  So then they proceeded in some 
external language to solve the posting problem all over again... but always 
forgetting some condition or other.



So I think it is a compelling need to "write code" to solve any problem that 
gets in the way.



Cheers....Hugh

CyberSpace Industries 2000 Inc.

Multimedia Promotions/ XML Training and Consulting