Re: For loop over a collection, multiple

[Hans-Juergen Rennau]

Dana,

maybe there are misunderstandings concerning the purpose of the scripting extension?

I suppose it is emphatically *not* an attempt to make XQuery a general purpose language, but an approach to considerably enlarge its potential for creating focused stand-alone applications, especially powerful tools. It is the answer to an issue of bad proportions:

"Do you REALLY mean to say that if I want to use and increment a counter, or make sure
several REST calls are invoked in order, or that I need to upgrade a global variable I need
to jump OUT of XQuery and INTO another programming paradigm, that is very different then XQuery
--  redundant and sometimes incompatible (two kinds or If-then-elses, two kinds of variables,
two kinds of procedure, two kinds of try-catch, two very different syntaxes, etc, etc, etc) !?"

Frankly, I understand very well that you are upset! I wonder if anybody engaged in tool writing could contradict you.

With kind regards,
Hans-Juergen

 


----- Ursprüngliche Mail ----
Von: Daniela Florescu <http://x-query.com/mailman/listinfo/talk>
An: Michael Kay <http://x-query.com/mailman/listinfo/talk>
CC: Martin Probst <http://x-query.com/mailman/listinfo/talk>; http://x-query.com/mailman/listinfo/talk; Hans-Juergen Rennau <http://x-query.com/mailman/listinfo/talk>
Gesendet: Mittwoch, den 3. März 2010, 19:11:39 Uhr
Betreff: Re:  Re: For loop over a collection, multiple


On Mar 3, 2010, at 1:21 AM, Michael Kay wrote:

>> But you are right, interesting programs have side effects,
>> that is why we are running them. I just don't think the side
>> effects must or should be programmed in XQuery.
> 
> Yes: I have to say I'm inclined to push the model of using XProc as the
> necklace that ties things together, and XQuery/XSLT for the processing
> beads.


Gentlepeople,  this discussion start to seem as surealistic to me as the one we
had on xml-dev 5 years ago about how bad declarativity was for processing XML...
(I had 500 answers in one day at the time...)

Michael, I don't really know what you are trying to say (I understand that you like XProc :-))
, but you are not answering to my comments, and to the need for a GOOD, SOLID scripting extension.

Do you REAALy mean to say that if I want to use and increment a counter, or make sure
several REST calls are invoked in order, or that I need to upgrade a global variable I need
to jump OUT of XQuery and INTO another programming paradigm, that is very different then XQuery
--  redundant and sometimes incompatible (two kinds or If-then-elses, two kinds of variables,
two kinds of procedure, two kinds of try-catch, two very different syntaxes, etc, etc, etc) !?

I hope not. It makes my head hurt just thinking about it.

XProc has usages and advatages, but that's not what I am talking about.
xmlsh has it's own advantages, but that's not what i am talking about.

I am talking about ***simple***,  ***common***  cases of (a) maintaining state and (b) controlling the
order of side effect operations.

And no, for this kind of simple things I do not want to start using angle brackets, and writing maybe
100 lines of "glue" code when I can write simply ":=" or ";" and be done with it.

Again, XProc has usages and advatages, but that's not what I am talking about.
xmlsh has it's own advantages, but that's not what I am talking about.

Just look at the original query that generated this thread, and try to write it with
(a) scripting
(b) XProc and
(c) smlsh

I really don't get this discussion/opposition to scripting extensions, and I really don't understand
this kind of arguments.

Best regards, cheers
Dana


__________________________________________________
Do You Yahoo!?
Sie sind Spam leid? Yahoo! Mail verfügt über einen herausragenden Schutz gegen Massenmails. 
http://mail.yahoo.com