juanrgonzaleza@c... wrote:
> James Fuller said:
> 
>>Melvin Chin wrote:
> 
> There are several advantages. In my previous message i cited
> [http://www.concisexml.org/]
> 
> <blockquote>
> 5. Closing tagname is optional
> Not only does this remove unnecessary clutter, but when ConciseXML is used
> as the syntax for dynamic languages, the tagname may not be known until
> runtime, and therefore the closing tagname must be optional.
> </blockquote>

w/o sounding too argumentative, 'removing unnec clutter' is not solved 
by this...putting in an something like </> this will confuse a number of 
users.

If the goal is to make representation shorter or more readable then 
there are some basic style conventions that can improve things in XML.

IMHO, if we are considering changes to XML spec to achieve something 
more readable there are other things that should happen before tinkering 
with lexical representation changes.

> Hum, liminal presents serious theoretical difficulties with the range
> calculus doing very difficult can be a satisfactory extension of XML.
> 
> [http://www.idealliance.org/papers/extreme/proceedings/html/2004/Piez01/EML2004Piez01.html]
> 
> In fact, liminal is apparently a 'dead horse' (liminal was first presented
> at 2002!).

I would argue that there is a 'long tail' effect with lots of xml 
related technology...people will use when they need to...we dont need to 
wait for mass adoption if you have a particular requirement.

I have considered using lmnl many times, though was always able to 
educate the end user just enough to use an editor/web interface.

> However, i obtained very good ideas from liminal as the generalized
> attribute model and the increase on difference between start and end tags
> for improving readability. However i follow an alternative approach to
> dealing with non-hierarchies because overlap generates many headaches.
> 
> Note that end tags are also optional in liminal _annotations_. End tags
> are required in ranges because of ambiguity in non-hierarchical ranges.
> 
> In short,
> 
> XML 1.0: <textarea>hi</textarea>
> 
> liminal: [textarea}hi{] or [textarea}hi{textarea]
> 
> ConciseXML: <textarea>hi</> or <textarea>hi</textarea>
> 
> SXML: (textarea "hi")
> 
> CanonML (both versions markup and environments): [::textarea hi]

apologies for my selfishness; I am interested in the impact of change of 
lexical representation downstream...not necc in the lexical 
representation itself (we would then have lots of worm cans with 
permathread on them to go through).

cheers, Jim Fuller