| Re: Coyote's BIG Idea [message #74512 is a reply to message #74510] |
Mon, 24 January 2011 10:30   |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Gray writes:
> Hey David, this may not be a question you can answer, but I'd like to
> know it: what can NG do that CG (coyote graphics) can't, and vice
> versa?
Well, that's a good question. I guess we are still trying
to figure that out. :-)
I'm not extremely familiar with the new function graphics
routines, because I can't keep IDL 8 running on my computer
(Windows 7, 64-bit OS). But I would say one thing Coyote Graphics
can't do yet (but probably because I haven't gotten around to
trying) is save the graphics commands in a display window and
restore them. In fact, the graphics "commands" are objects in
the FSC_Window display window, so I suspect I can do this quite
easily when I get around to it.
One thing I know for sure the new function graphics can't do
that Coyote Graphics can is create a new type of graphics command
in a widget program. My challenge to demonstrate how one would
build a "window-level" program with function graphics goes
unanswered to this day. I strongly suspect this is because it can't
be done. At least not by mortals. Maybe someone at ITTVIS could
figure it out. But, I don't think, frankly, it is possible to
use function graphics in a widget program that requires any
kind of user interaction with the graphics window. So no
XStretch, no WindowImage, no FSC_Zplot, etc, etc.
The reason I think function graphics is limited in this area
is that the graphics functions are very high level routines.
Coyote Graphics routines are much lower-level. You can build
things with Coyote Graphics, while you have to settle for what's
already been built with function graphics.
Function graphics allows you to set graphical "properties"
with property sheets. I have something similar in mind, but
haven't implemented it yet. I think Coyote Graphics at the moment
is closer to how people have traditionally worked in IDL, rather
than introducing a whole new way of working and interacting with
graphics programs. That is to say, they work as well from the
command line as they do in a program. Maybe this is a wash.
I personally think function graphics commands, under the hood,
are so complicated that they will never be bug free. I think there
is a very good chance that every time you try to fix something,
you will break something else. Coyote Graphics are probably more
complicated than I thought they would be, but mostly because I have
tried to preserve as much backward compatibility with traditional
graphics commands as possible. If I had decided to start from scratch,
they would be very, very simple.
I don't know. Maybe other people have some ideas. I suppose it
is no secret that I just consider it a challenge to build IDL 8 like
programs for IDL 6 like people. Maybe no one will use the programs
I write. In truth, that's not something I lose a lot of sleep
thinking about. It's fun, it's challenging, and I am blown
away by what is possible with old, familiar tools. Good enough
for me. :-)
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.idlcoyote.com/
Sepore ma de ni thui. ("Perhaps thou speakest truth.")
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
