Rough proposal for new standard: Content Mediation.

[Christopher Granade]

I don't know if this is the right place to submit this or not, but I
had a rough idea for a new web standard that would work closely with
XML that I wanted to submit to the community for consideration. I
posted a (very) rough draft to Google Pages for now, if someone wants
to look at it, but the basic idea is to allow non-XML content to be
embedded into XML and then converted to XML at parsetime. This would
allow for a more natural expression of some data, such as allowing
Notation 3 markup to be embedded within an XML document, rather than
working directly with the XML RDF representation.

If anyone is interested, the draft is posted to
http://cgranade.googlepages.com/contentmediation

Thanks for your attention and interest.
--Christopher Granade

 

[Soren Kuula]

I don't know if this is the right place to submit this or not, but I
had a rough idea for a new web standard that would work closely with
XML that I wanted to submit to the community for consideration. I
posted a (very) rough draft to Google Pages for now, if someone wants
to look at it, but the basic idea is to allow non-XML content to be
embedded into XML and then converted to XML at parsetime.

Sure.

1) XML ain't no tool. It is a language
2) Isn't this more specific, more general and invented already?
http://www.brics.dk/xsugar/

Soren

 

[Christopher Granade]

Sure.

1) XML ain't no tool. It is a language
2) Isn't this more specific, more general and invented already?
http://www.brics.dk/xsugar/

Soren

1) I didn't want to call it that, as XML lets you make languages, and
in the context it would be confusing, but yeah... I know what the L
stands for... thanks, though. Seriously, thanks.
2) It looks somewhat different in that it works for languages for which
you can write a nice BNF description, but I don't know how you would
write one for the first language I proposed that would be mediated,
CasualML. If it works, then I'm all for it. The only other thing would
be a question of embedding XSugar in an XML document. I'd still want to
be able to do that, but perhaps that's just a silly idea I've become
attached to.

That is a really cool project, though. I'm not sure if the goals are
exactly the same, but I would love to further the idea of mixing XML
with things that work better in a particular scenario. I guess I hadn't
heard of XSugar, since it doesn't see much use apparantly. It's a
shame, really. I love the BiBTeX embedding idea that BRICS presented.
Surprised I didn't see XSugar, as I loved JWIG (another BRICS project)
before it fell out of active maintenance.

Anyway, thanks again for the pointers!
--Chris

 

[Soren Kuula]

1) I didn't want to call it that, as XML lets you make languages, and
in the context it would be confusing, but yeah... I know what the L
stands for... thanks, though. Seriously, thanks.

Exactly, makes you make languages. M is for Meta, but of course, you know.

2) It looks somewhat different in that it works for languages for which
you can write a nice BNF description, but I don't know how you would
write one for the first language I proposed that would be mediated,
CasualML. If it works, then I'm all for it. The only other thing would
be a question of embedding XSugar in an XML document. I'd still want to
be able to do that, but perhaps that's just a silly idea I've become
attached to.

Well ... I think that the flip side of your idea is: You want to extend
XML with something which is not XML, but might have been described in
XML. That is making things more complicated than necessary (and XML
survived SGML because --- XML was NOT more complicated than necessary.
Some full XML implementations actually exist. Some people actually can
understand all the syntax of any XML document. Can't say that about SGML
)

That is a really cool project, though. I'm not sure if the goals are
exactly the same, but I would love to further the idea of mixing XML
with things that work better in a particular scenario. I guess I hadn't
heard of XSugar, since it doesn't see much use apparantly.

Research project.

Oh, you know BRICS! Aarhus is one COOL place for XML buffs to study CS.
I recently graduated from there (under Møller & Schwartbach!). Made a
static XSLT validator (a thing that can validate XSLT output even before
you know the input).

But keep working on it -- I would just suggest that you leave XML a nice
and simple language as it is (the parser too), and if you want, define
other languages that easily _translate_ to XML. These other languages
could be partially XML themselves...

Soren

 

[Christopher Granade]

Exactly, makes you make languages. M is for Meta, but of course, you know.


Well ... I think that the flip side of your idea is: You want to extend
XML with something which is not XML, but might have been described in
XML. That is making things more complicated than necessary (and XML
survived SGML because --- XML was NOT more complicated than necessary.
Some full XML implementations actually exist. Some people actually can
understand all the syntax of any XML document. Can't say that about SGML
)


Research project.

Oh, you know BRICS! Aarhus is one COOL place for XML buffs to study CS.
I recently graduated from there (under Møller & Schwartbach!). Made a
static XSLT validator (a thing that can validate XSLT output even before
you know the input).

But keep working on it -- I would just suggest that you leave XML a nice
and simple language as it is (the parser too), and if you want, define
other languages that easily _translate_ to XML. These other languages
could be partially XML themselves...

Soren

I suppose that by now some justification for my ideas is in order. I
was hoping to find a way of embedding what I call CasualML
(http://cgranade.googlepages.com/casualml) in an XML file so that a
user can write something and so that the something in question can both
be preserved for later editing and transformed to XML for consumption.
I guess it could all be done server-side, and it probably would be
anyway, but that was just my idea.