CLIM mail archive
[Prev][Next][Index][Thread]

who's designing clim 2.0

    Date: Fri, 30 Aug 91 08:24:45 -0400
    From: robert futrelle <futrell@corwin.ccs.northeastern.edu>

    I have a comment about what I'd like to see in clim 2.0
    and I'm a little concerned that others haven't spoken up.
    The group that is designing clim has been very quiet.
    I don't know who they are and they haven't been actively
    seeking advice from users on this mailing list.

The group that is designing CLIM 2.0 is comprised of representatives from ILA,
Symbolics, Lucid, Franz, Apple, Xerox, DEC, and MCC (the corporate entities
that have invested in CLIM).

    Nevertheless, there has been an enormous amount of
    experience over the last 10 or 20 years with window
    systems and interface building, so I am keeping my
    fingers crossed that they are doing their best to 
    incorporate this experience and are doing great things
    for the rest of us.

We have had to choose our areas of innovation carefully.  We are not trying to
design the ultimate window system, rather a high-level, portable interface to
existing window systems with some carefully selected extensions.

    I can see a spectrum of applications for clim from
    pure text and building editors all the way to 3d
    graphics for cad/cam and for games, for that matter.

Yes, indeed.  Remember, however, that there is only one absolute goal for CLIM:
portability.  Below that, there are various subgoals which involve certain
tensions.  For example, there is a tension between satisfying people who want
specific features and people who want a "small CLIM".  There is a tension
between associating too heavily with a user interface paradigm from one window
system and not providing enough integration with any particular one.

    Our lab is trying to create large systems that present
    scientific and technical documents on the screen.
    This means multi-fonts (e.g., Greek (and Hebrew for math))
    and especially subscripts and superscripts in running
    text and a somewhat richer model for full math equation
    display.  Mike Lesk at Bellcore discovered, in his large
    project with Amer Chem Soc that he needed seven levels
    of vertical text positioning just to do chemical text
    (baseline plus sub and super superscript plus
    sub and super subscript plus normal sub and superscript.

In this case, remember that CLIM does in fact provide arbitrary positioning of
glyphs in device-specific fonts through the DRAW-TEXT primitive.  Equation
formatting is something that you specifically need and which others probably
think would add unnecessary bloat, preferring instead (for example) 3D
graphics, which indicate you don't need.  From CLIM's perspective, this is a
good candidate for an extension library (or an application-specific utility).

    The reason this needs clim is because these document 
    systems have to be intelligent (there's knowledge and
    hyper-links behind all those characters) so that in turn
    needs smart systems built in CL/CLOS (and OODBMS?!!).

Yes, this is the reason we implemented CLIM in CL/CLOS.

    It goes without saying that all these technical texts need
    graphics for all the many diagrams, but these are typically
    2-d diagrams and clim seems to be strong enough in the 
    graphic area for this.

    There is one deficiency which should be addressed:
    There should be the ability to present text rotated 
    90 degrees as appears on the zillions of data graphs
    and on some labeled drawings that are published every
    year. (our field of biology publishes 300,000
    *papers* every year  :-) !!

Yes, we already have some facilities for this.  As you might guess, this is
hard to do uniformly in a portable way.  Maybe if everyone were running Display
PostScript. :-)

    In summary:
      Who's designing clim 2.0?

See above.

      What are their goals - structure, function, features?

CLIM 2.0 is not going to be radically different from CLIM 1.0 in the areas
already covered by CLIM 1.0.  CLIM 2.0 is CLIM 1.0 plus facilities added to
support toolkit-style programming and therefore the integration with the look
and feel of the native platform.

      Many of us need technical text displays.

This is one of many layers that application developers may need.  Not all of
these layers need be specified by the CLIM committee.  They should definitely
not be part of the core module.

      Basic 2-d graphics in clim seems adequate.

Others, however, have requested 3D.  (This is one of those tensions I mentioned
above.)

      We need 90 degree rotated text [we do faces  :-)  ].

We hope to be able to satisfy this for most platforms.

      I hope others will speak up before 2.0 is frozen.

We encourage comments and throughout this entire process we have incorporated
the comments and suggestions from our customer base.

    Thanks for CLIM and keep up the good work --
      and, you vendors,  we need sources! (thanks symbolics).

Dennis Doughty
International Lisp Associates

0,,
References:
Main Index | Thread Index