notes on conference call of July 1

Notes on HTML-Math ERB Conference Call
1 July 96
---------------------------------------------------------------------

In attendance:

Stan Devitt     Maplesoft
Patrick Ion     Mathematical Reviews
Robert Miner    Geometry Center
Dave Raggett    W3C
Neil Soiffer    Wolfram Research
Ron Whitney     American Math Society
Ralph Youngen   American Math Society

[Notes prepared by RW.  Corrections welcome.]
---------------------------------------------------------------------


Ron asked if there were any questions or misgivings about the schedule
which Dave posted last week.  Ron's question was specifically
concerned with the targets of producing a W3C Working Draft this year
and holding a meeting of this ERB in September.  Neil felt that we
were on target with the schedule since the renderer implementations
under way (Bruce's in Mathematica, Robert and Dave's in Java) should
be available soon and we will also have a draft of the Draft ready
soon.  Developing the W3C Working Draft will then proceed with the
meeting in September and subsequent work.

Patrick voiced that he felt we had a lot to do between now and
September.  Ron concurred and mentioned that he felt more work was
needed in specifying where we stand in relation to, say, the efforts
of the OpenMath Consortium and the ISO 12083 Update Committee, as well
as further discussion of some of the problems (e.g. 'macros' or
'template matching') we haven't addressed in detail.

We tentatively scheduled an ERB Meeting for September 26 and 27, to be
held at or near W3C in the Boston area.  [Board members: please let
me know whether you want to attend a group meeting and whether you
will be able to attend on these dates. -RW]

Ron asked about Robert's work and whether it would be publicly
accessible.  Robert said he's creating various Java classes for the
display list schemata of Bruce's proposal.  Robert's work will be
tentatively finished very soon (perhaps within the week), but it does
require the display list format of Bruce's proposal as input.  (The
code will be publicly available.)  Dave will complement this work with
a parser to go from data entry to display list.  Dave feels his work
on this parser will be done a couple of weeks after he's able to start
work in France, thus by the end of the month (July).

Ron returned to asking about generalities regarding where we should be
in order to hold the meeting in September, whether anyone felt there
was more than just the details on which we're now working that should
be in place for the meeting.

Stan asked about the degree to which we're interacting with the
OpenMath Consortium.  Neil had seen the work of OpenMath a couple
years ago, and had recent indications that matters hadn't developed a
great deal further [Is that fair, Neil? -RW].  He felt that the
emphasis of the OpenMath group was entirely different than ours
(OpenMath's being much more semantically oriented, whereas we are
looking to routes to enable easy passage from data entry to some form
suitable for different rendering media, as well as more
semantically-oriented operation).  Pike's is perhaps the purist
position beyond that of OpenMath.  Dave hopes to continue some
dialogue with his OpenMath connections while he is in France.  [I do
feel a need to become much more informed about the Consortium's work
and its relation, or possible relation, to ours.  I aim to post some
notes on my understanding, if only to place our project in a wider
context.  This seems a logical point of inquiry by W3C members as we
propose our Working Draft. -RW]

Stan said he'd be posting some detailed reactions and questions
concerning the Wolfram proposal this week.  In particular, he asked
how will we enable semantic context changes?  Neil responded that
semantical mappings (renderings) will be handled with macro/pattern
matching on the expression tree.

Stan asked how stylistic attributes can be carried forward as they go
through the semantical wash.  (One may wish to render 'x' as 'green'
in some context.  Plain 'x' may contain all semantical content, but
how can we ensure that when a semantical engine returns 'x' in another
context that it is rendered 'green' again so that style is preserved?)
Neil mentioned that Mathematica uses a 'SpecialBox' [this is the word
I have jotted in my notes, but I don't see it in the Mathematica Book
for version 3; there are various style boxes and I suspect I got the
vocabulary wrong -RW].  This is a topic we haven't discussed in
detail, although Neil felt it could be addressed.

Stan asked how the local context (both semantic and stylistic) will be
specified.  Neil and Dave spoke of specifying contexts either by
standard name or URL.  Dave felt a solution similar to that of 'style
sheets' could be used.

[I believe that further discussion on each of the topics Stan raised
in the telephone conversation will be possible as Stan posts more detail
this week and next. -RW]

Patrick asked for general reaction to Richard Fateman's recent posting
[available on request -RW] about the Academic Press CDROM of the
Gradshetyn and Rhyzik table of integrals.  The general reaction was
one of surprise that Fateman was so het up about it.  He is perhaps
goading us all into more vigorous discussion.

The conversation concluded with Neil asking Ron to post some AMS math
papers which show some difficult TeX examples.  Ron promised (again)
to do this.