Re: the future of xslt

On Sun, Jun 22, 2008 at 1:34 PM, Liam Quin <liam@xxxxxx> wrote:
> On Sun, Jun 22, 2008 at 11:28:47AM +0200, James Fuller wrote:
>> On Sat, Jun 21, 2008 at 10:21 PM, Dimitre Novatchev
>> <dnovatchev@xxxxxxxxx> wrote:
>> > I am not sure these statistics are useful at all.
>>
>> I agree that presenting anything as 'statistics' is potentially
>> problematic, but these trends must be indicative of 'something' or
>> plain wrong.
>
> Not all "statistics" need to be meaningful.

yes ... and they can be wrong as well.

>> reason why I presented these trends was to understand why google would
>> have declining trend for xslt ... doesn't seem to make any sense.
>
> SQL has also "declined".  So have socks, although not as alarmingly
> as "public nudity".  About the only significant increase I found
> so far was "Google".

> I think it suggests that more and more people are using Google as
> a search engine, and not just a bunch of geeks.  Some terms that
> don't seem to have declined: laundry, handbag, curtains, barbie,
> trousers (although underpants are coming down in all regions),
> "rolling stones", Mozart, ...
>
>> pessimistically, I do not think we will ever see wide spread adoption
>> of XSLT, like lets say java....
>
> I think we already have widespread deplyment of XSLT.  The Java
> statistics include other meanings of the word, more common in
> the World Outside Computers.

deployment is different then developer adoption

and yes, there is problem with spelling (for example; Haskel versus
Haskell) whenever doing this kind of thing.

>> I do have a 'point' ... I am trying to gather adhoc and statistically
>> relevant material on putting some % on the likeliness of any of the
>> following occurring;
>>
>> * will XSLT 2.0 experience significant adoption ? what about xslt 2.0
>> in the browser ?
>> * XSLT on other devices e.g. hardware, mobile platforms
>> * will adoption flow from XSLT 1.0 to XSLT 2.0 or ... XSLT 1.0 to XQuery ?
>> * will we have XSLT 3.0
>
> I suspect that some other approaches to gathering this information
> may be more productive.

<ot>i find it interesting that all of us who use google every day
constantly, once presented with statistical data from same, said
search engine turns us all into 'doubting toms'</ot>

I sent this to the list to get a feel of what other people are
experiencing .. I certainly feel that XSLT usage is expanding and
adoption is 'happening'.

> XSLT 1.0 is in all major Web browsers today.  Well, if a browser
> doesn't have XSLT I don't consider it to be major :-)

I would add EXSLT node-set to that consideration.

> I don't know about XSLT 2 in the browser, nor XQuery, although I
> wouldn't want to rule either of those out.  For what it's worth,
> jquery impements at least some of XPath in JavaScript, which is
> less insane than it sounds when you realise it also has a JSON parser --

I wish we had perl ;)

I think if we had XSLT 2.0 in the browser then we would have XQuery
for nearly free.

> I was reading a Linux magazine on the 'plane today that mentioned
> that the way to turn a JSON stream into objects in the browser was
> to use eval() on it, which as ironic as the cover of the magazine
> had SECURITY on it in huge letters.  No, it needs to be parsed...
> so, people are doing a lot more with JavaScript these days.

yes, I have similar thoughts about JSON ... marshalling and
serializing objects is something else

> Right now, though, I think the "sweet spot" for XQuery is on the
> server, taking advantage of indexes, heavy optimisation, and
> of course things like SQL connectivity.  For XSLT 2 it's less
> clear: pretty much anyone using XSLT will benefit from XSLT 2.

I would further constrain and say I think xquery 'sweet spot' is on
XML database server

cheers, Jim