Re: [office] Formula subcommittee - olive branch?

[robert_weir@us.ibm.com]

"David A. Wheeler" <dwheeler@dwheeler.com>
wrote on 02/01/2006 05:33:42 PM:

> FYI, I've had a private email asking if in the future there might
be
> interest in our working on spreadsheet formulas somehow
> with ECMA and/or the folks involved with EOOX.

Interesting thought.  Why do you think they suggested
formulas in particular?  With a combined size of almost 3 kilo-pages
of specification, there are numerous other points of interest when comparing
the two.  

This just isn't a question of spreadsheet formulas.
 It is a difference of philosophy.  ODF was designed to reuse
existing XML vocabularies wherever possible.  If inventing something
from scratch could be avoided by using an established standard, it was
used.  This is a great benefit to implementers and developers, since
you can use the tools and standards you already know.  On the other
hand,  EOOX seems to avoid obvious choices of existing standards.
  For example, ODF uses MathML for equations, while EOOX uses a nonstandard
called OOML.  Anyone ever here of OOML?  ODF uses SVG for vector
graphics.  EOOX uses VML and DrawML, etc.

So, I'd suggest the first step for making EOOX more
in harmony with ODF would be for it to reuse some of the same existing
standards which ODF uses.  We're talking about uncontroversial W3C
standards like SVG and MathML.  But unfortunately the point is moot.
 My understanding of the ECMA EOOX submission was under Terms of Reference
which required that ECMA standardize an XML format which is 100% backwards
compatible with the existing Office 12 XML formats.  So, there doesn't
seem to be any way in which ECMA TC45 could change their formula language
to be closer to ODF's. 
  
And on the other side, I'm not sure there is anything
this TC can do to be closer to ECMA's formula language.  Why?  Because
it simply doesn't exist.  The posted draft EOOX does not specify formulas
as being anything more than xsd:string.


> 
> I can't speak for anyone else.  I can say that for myself, I
would
> be happy to work with ANY party, ESPECIALLY any implementor,
> that is interested in creating an open standard.  On the other
hand,
> it seems unwise to sit in a "wait forever" mode for
> something that may never occur. ECMA has not committed
> to doing this work, to my knowledge, and the OpenDocument TC
> members see a need to start NOW.
> 

OASIS does it work openly and publically.  They
are free to observe what we do, and participate if they wish to join.  Some
ECMA TC45 member companies (including the submittor of EOOX) are already
OASIS members, so they can join the TC instantly with a few mouse clicks,
if they wish.


> But how about this - as part of the establishment of the
> formula subcommittee, make a point to invite all implementers
> to join the TC subcommittee, and that you'd like all to join.
> You'd mean that anyway, but it'd be a nice gesture; is there
> any objection to making an invitation as a formal announcement?
> 

We should certainly publicize the creation of the
subcommittee as we have previously with the Accessibility subcommittee.
In that case, an announcement was made on the TC homepage as well as the
OASIS homepage.  If there is someone of particular expertise in a
special area (formal grammars, financial functions, etc.),then we should
be sure to solicit them.  But keep in mind that subcommittee participation
requires OASIS membership.

> If there are other ways to extend an olive branch, or make it
> easier to bring those people into a process, that'd be great too.
> If anyone has any ideas on that score, please voice them.
> 

Where do formula people hang out? <g>  Honestly,
I'd expect that the existing OpenFormula contributors would be the best
pool of recruits, since they have already expressed interest in the topic.

Also, perhaps a representative from Corel could join?
 Their Quattro Pro perspective could be useful.

> This way OASIS can continue at top speed, but if others want
> to join the standardization process later on, they can do so.
> I don't know if we can make the standardization stars align,
> but I would be DELIGHTED if it could happen somehow.
> If everyone could agree on ONE spec that'd be great.

We do what we can, right?  Since OASIS operates
under a transparent, open and public standards development process, it
will produce standards that reflect the needs and interests of those who
come to the table, and this includes a diversity of government, commercial,
academic, non-profit and open source interests.  There is no single
captain at the wheel who can, by his own will, steer it in one direction
or another.  But if on the EOOX side there is no give-and-take of
open standards development, and the only condition is that their specification
remain compatible with Office 12, then I'm not optimist that these ships
will ever meet.

-Rob