Re: XSLT functions for JSON

On Mon, May 19, 2008 at 7:02 AM, Colin Adams
<colinpauladams@xxxxxxxxxxxxxx> wrote:
> 2008/5/19 James Fuller <james.fuller.2007@xxxxxxxxx>:
>> using JSON's serialization format for javascript objects, as a generic
>> approach to marshalling data over the wire is probably ok ...  though
>> I see the decision to use JSON as an exercise in late optimization. I
>> have yet to employ this optimization in any production.
>>
>> I know that javascript is everywhere and is part of the web and people
>> use it and thusly want to use things like JSON ... but comparing (not
>> saying this thread is doing this) it to XML is 'apples' and
>> 'airplanes' ... also there is no reason why the OP can't write their
>> own extension function.
>>
>> I am not against JSON ... I like little domain text based languages
>> (cue perl, sed, awk, ack) but we live in a world where the web is
>> mostly markup; XML is effectively HTML's brother and they share an
>> enormous toolbase ('view source' anyone) and heritage which must be
>> compelling at many levels.
>
> What about YAML?
>
> YAML 1.2 is a pure superset of JSON, I believe. And there is some work
> going on to define an XML <-> YAML mapping (but it may be a subset of
> each).
> If this mapping were to be settled, then you could use that for the
> transformation.

YAML is fine as well and has great perl support

> On the whole, i don't really see any advantage in these things over XML.

I agree; I think most people when they use 'other' things eventually
will come up against some shortcoming. XML has shortcomings, but it
represents to me a very good 'least' amount of features and no
more/less ... though admittedly I am a XML namespace fanboy, I can
imagine there are some people who are not.

cheers, Jim Fuller