Re: Using HyTime Arcform stuff (was Re: Ephemeral XML?)

At 9:37 AM 1/14/97, Ralph Ferris wrote:
>At 09:46 AM 1/14/97 -0900, Eliot Kimber wrote:
>>Can we stipulate that it is assumed that nobody in this discussion will
>>knowingly suggest that any syntax be required if it can be reasonably
>>inferred?  I believe that's a basic design principle behind XML.

At least not as you have taken the point. Eliot was scolding me for
assuming that his example of how HyTime does things was a suggestion about
how XML should do things. It is still our task to decide what syntax will
be required, and what it is "reasonable" to infer.

> Said another way, the set of archforms defined in the XML spec make up
>the API for an XML browser (what Len has suggested we consider defining).

I still disagree that we should be saying anything about APIs, except in
the context of style sheets. Markup is for declarative, structural
information. Stylesheets are for declarations of formatting or other
processing behavior.

> To
>override the default behavior, you don't necessarily even have to declare
>another architecture. In parsing a FORM element, for example, the parser can
>determine if it complies with the architectural form for FORM-type elements.

This means that syntax errors in form elements can silently turn off form
processing. This is even worse than getting form-processing errors when
using a doucment without its DTD.

>If so, the browser processes it accord to the behavior defined for that
>architectural form (which would conform to the behavior defined in HTML for
>FORM elements).

We are going to need the stylesheet anyway, why not go there in every case?

> If not, it looks in the stylesheet. If the desired behavior
>is too complex for a stylesheet definition, possibly the stylesheet itself
>could point off to another architecture or alternate architectural form.
>Something to consider when we get to that point in the discussion.

There is nothing that is too complex for a stylesheet definition.
Stylesheets are for attaching operational semantics to the structural
relations given in a document instance.

   I still think there is no reason to pollute the namespace. If users want
form semantics, and we decide that these function should _not_ be a part of
general XML stylesheets, then they are free to use HTML, or the HTML-in-XML
DTD and stylesheet (which simply says "use HTML").

   I don't see any problems with putting forms into DSSSL style sheets
though. The hard part may be putting forms into CSS -- but I think we can
burn that bridge when we come to it.

   -- David

I am not a number. I am an undefined character.
David Durand              dgd@cs.bu.edu  \  david@dynamicDiagrams.com
Boston University Computer Science        \  Sr. Analyst
http://www.cs.bu.edu/students/grads/dgd/   \  Dynamic Diagrams
--------------------------------------------\  http://dynamicDiagrams.com/
MAPA: mapping for the WWW                    \__________________________