Message-Id: <9302081956.AA02303@sc.zib-berlin.dbp.de>
Reply-To: Mailing list for the LaTeX3 project <LATEX-L@vm.urz.Uni-Heidelberg.de>
In-Reply-To:  <01GUHD64KHKIG4LPG6@MATH.AMS.ORG>
Date:         Mon, 8 Feb 93 20:48:07 CET
From: Michael Downes <MJD@MATH.AMS.ORG>
Sender: Mailing list for the LaTeX3 project <LATEX-L@vm.urz.Uni-Heidelberg.de>
To: Multiple Recipients of <LATEX-L@vm.urz.Uni-Heidelberg.de>
Subject:      RE: task list, diagrams in particular
Status: R


Paul Taylor wrote:

> George Greenwade has recently mailed to a sub-list:
> > One of the enumerated responsibilities outlined for task coordinators in
> > the ``Volunteer work for the LaTeX3 project'' document is:
> > > The second job of a task coordinator is to write a `task
> > > specification' containing more details than were in the
> > > volunteer task list. It should describe the goals and any
> > > restrictions that apply. This task specification will need to be
> > > reviewed by someone on the LaTeX3 kernel team, and afterwards it
> > > can be sent out to any additional volunteers for the same task,
> > > as a guide for the work.
> but it seems more appropriate to reply to the general latex 3 list.
>
> I agreed in October to be "co-ordinator" for commutative diagrams
> (in category theory (pure mathematics) and theoretical computer science)
> and expected to receive some comments on this subject from other who are
> interested (Kris Rose, for example), but got nothing.
>
> The "task list" says that what I am supposed to do should take "2-3 days".
> Now the programming I have done has taken two orders of magnitude more time
> than that, so I'm a bit puzzled as to what job does take that long.

It is hard to make an accurate time estimate when the scope of the
task is ill-defined. Thus getting a "task specification" from the task
coordinators is a necessary first step towards refining the time
estimates. Also, probably the time estimate for the commutative
diagrams task was not intended to cover programming time to produce a
full set of diagram macros, but only analysis of

  (a) what a good commutative diagram package should be able to
      handle.

  (b) what would be good user syntax for various features (for
      example, compare LAMSTeX's [x,y] coordinate approach for
      specifying the destination of arrows versus the ul ur dll
      notation used by XYpic.)

  (c) outline of a proposed plan for creating a pilot implementation
      that works with LaTeX 2.09.

In the case of commutative diagrams, the answer to (c) may be
something relatively simple such as "use Paul Taylor's stuff, but
borrow two or three nice features from XYpic and one from LAMS-TeX".

For LaTeX3, it is envisioned that special applications such as
commutative diagrams would be accessible as "modules" through a
well-defined interface (something like the idea of LaTeX's current
documentstyle option interface, but probably rather different in the
details). It would seem preferable in the LaTeX3 documentation to be
able to describe a single commutative diagram module rather than three
different alternatives, *but* it would be quite appropriate for a task
coordinator such as yourself to say, "For this particular application
I think LaTeX3 ought to support the three alternatives rather than
choosing one, or making a composite out of their various features" if
that is indeed your judgment. In any case, even if LaTeX3 settles on
a single "official" commutative diagram package, as determined by the
what winds up in the official documentation, there is no reason to
discourage the authors of independent packages from making them usable
with LaTeX3 and continuing to maintain them.

> Now turning to my "task" in particular, I have in mind to write to those,
> including Kris, who have worked in this area. Maybe there are other readers
> of this list who have something to contribute, and I would like to hear
> (by private email) from them. I haven't drafted the letter yet.

That's a good example of the kind of investigation that, it is hoped,
task coordinators will do.

> There was a correspondence on the category theory list some months ago
> in which some people got very upset when I rejected suggestions which
> were unspecified (in the technical sense), of minority interest and far
> from what I could implement

Winnowing and synthesizing of suggestions and opinions is the essence
of the task. Your own best judgment is in fact the most valuable thing
you contribute to the job of task coordinator, so by all means feel
free to exercise it. Think of yourself as a benevolent dictator or
something like that.

Michael Downes                              mjd@math.ams.org (Internet)