Remarks and Ideas (fwd)

[mbp]

Andrae,
    Here are some responses to the questions you sent Daeron last
week.  Many of the problems you found are particular to the NeXT
version of Geomview, which is not yet as complete as the SGI version.
Also, I wrote the manual specifically for the SGI version, so there
are some things that it says which don't apply to the NeXT version.
We'll write a more complete manual sometime, including NeXT-specific
details, but it will probably be a while before we're able to get to
it.  Sorry for the confusion, and thanks for your interest.  If
you have more questions let us know (mail to software@geom.umn.edu).

--Mark Phillips

> 1. ap-override arguments differs to the manual.
>    c: (ap-override on)
>    o: ap-override: id expected in arg position 1 (got "on")
>    ---
>    (ap-override off)
>    ap-override: id expected in arg position 1 (got "off")
>    ---
>    c: (ap-override g1)
>    o: 

Actually the manual is correct; this is a bug in the program.  I've
fixed it in our copy so it will be fixed in the next release.  The
"override" button on the appearance panel does the same thing as this
command and should work in your version.  If you need the GCL version
of this, let me know and we'll either get a patch to you or find a
temporary workaround.

> 2. (car LIST) we do not understand. We thought (all geometry) returns a list.
>    c: (car (all geometry))
>    o: lisp error: call to unknown function "square.quad"
>       car: list expected in arg position 1 (got nil)
>    c: (echo (car (all geometry)))
>    o: lisp error: call to unknown function "square.quad"
>       car: list expected in arg position 1 (got nil)
> 
> 3. (cdr LIST) same problem like in 2.

I think this is a bug.  It's rather embarassing that such simple
functions as car and cdr don't seem to work, but in fact no one here
has ever used them.  (Actually I think they worked at one time but
broke when we redid the parser a year ago).  If you need to access
particular elements of a list, I think the only way to do it right now
is to parse it manually.  If you need this functionality please let me
know and I will try to fix it for the next release.


> 4. (cull-backface on|off) meaning and reactions ?

This is described in the manual, but it only works in the
SGI version.  It does nothing on the NeXT.


> 5. (cursor ...) function ?
>    c: (cursor c0 off)
>    o: Reading "[1]/tmp/geomview/OOGL": call to unknown function ""cursor"":
>       > ... (cursor c0 off)
>       -------------^
>       lisp error: call to unknown function "cursor"

This function only exists in the SGI verison.

> 6. (cursor-still N) function for cursor quiet indication.
>    No different reaction for N=10, N=100, N=1000 ?

This function exists in the NeXT version but does nothing
because we haven't gotten around to writing that part of the code
for the NeXT yet.  Sorry.

> 7. (interest (command [arg masks])). Undependent load from panel or load from
>    togeomview did not produce an echo to togeomview.
>    c: (interest (load *))
>    c: (load br4fd.grp)
>    o:

Right.  "togeomview" is not able to receive data back from geomview;
it is for one-way communication only.  For bi-directional
communication you should write an external module as described in the
manual.  For playing around with GCL commands interactively you can
invoke geomview with the "-c -" option, i.e.  "geomview -c -" which
causes it to read GCL commands from stdin; output will go to stdout.
I tried the above example in this way and it works fine.

> 8. (lines-closer c0 N) meaning and reaction ? No differences in output for the
>    follwing command sequence.
>    c: (lines-closer c0 1)
>    c: (lines-closer c0 10)
>    c: (lines-closer c0 100)
>    c: (lines-closer c0 1000)

This one is also only implemented on the SGI.  On the NeXT it
does nothing.

> 9. (ND-axes) (ND-color) (ND-xform) meaning and reaction ?
>    c: (echo (ND-xform g1))
>    o: no output because all geomview windows are hide now and geomview.app is
>       no longer active as task (see process monitor).
>
> 10. (time-interests deltatime initial prefix suffix) meaning and reaction ?

I'll defer to Stuart Levy on these questions because he knows a lot
more about them than I do.  He'll send a separate message about these.

> 11. (update N) meaning and reaction. For N=0.1/1.0/10.0/100.0 no differences
>     detected (two square.quad instance are rotating with different transform
>     parameters in the world).

This function causes Geomview to leap ahead N seconds in time (also
works for N<0, in which case Geomview jumps backwards in time).  It
does work on both SGI and NeXT.  This only affects the motion of
objects.

> 12. (update-draw c0 N) meaning and reaction. Same problems detected like shown
>     in 11.

This is similar to "update" except that it forces the image
in camera c0 to be redrawn.

> But the implementation of idea one (at ...),(after ...) is the more interested
> wish. Are there any possiblities to take such ideas into consideration ? We 
> would be very pleased about it.

This looks it's worth considering, but there are some complications
related to the way we implemented the parser.  Unfortunately we are
very busy with another project right now and I can't say when we might
be able to get to it.  It could be several months.

For now I would suggest letting your external module control the
timing --- i.e. it should emit all commands to geomview at the time
that they should be interpreted.  This doesn't guarantee completely
accurate timing, because their might be some lag time after your
module emits a command before Geomview acts on it (if there is a lot
of data going back and forth, or if Geomview is running slowly), but
it usually works pretty well.