| Re: Exceed and IDL [message #16186] |
Thu, 08 July 1999 00:00 |
meron
Messages: 51
Registered: July 1995
|
Member |
|
|
In article <7m142h$474$1@clam.niwa.cri.nz>, "Mark Hadfield" <m.hadfield@niwa.cri.nz> writes:
> Richard G. French <rfrench@wellesley.edu> wrote in message
> news:3783A3A3.233D8535@wellesley.edu...
>> I have been clobbered by using canned routines that set !P. !X, !Y, !Z
>> system variables for their plots. Then, when I am in the same session
>> and try my own interactive stuff, these stale values remain. I should
>> write a routine that I can call that cleans the slate of all set
>> variables and restores them to the defaults -- has someone done this
>> already?
>
> Take a look at the IDL Astronomy Library routine CLEANPLOT:
>
Or at PLVAR_KEEP, in my library.
Mati Meron | "When you argue with a fool,
meron@cars.uchicago.edu | chances are he is doing just the same"
|
|
|
|
| Re: Exceed and IDL [message #16188 is a reply to message #16186] |
Thu, 08 July 1999 00:00  |
Mark Hadfield
Messages: 783
Registered: May 1995
|
Senior Member |
|
|
Richard G. French <rfrench@wellesley.edu> wrote in message
news:3783A3A3.233D8535@wellesley.edu...
> I have been clobbered by using canned routines that set !P. !X, !Y, !Z
> system variables for their plots. Then, when I am in the same session
> and try my own interactive stuff, these stale values remain. I should
> write a routine that I can call that cleans the slate of all set
> variables and restores them to the defaults -- has someone done this
> already?
Take a look at the IDL Astronomy Library routine CLEANPLOT:
http://idlastro.gsfc.nasa.gov/ftp/pro/plot/cleanplot.pro
--
Mark Hadfield m.hadfield@niwa.cri.nz
National Institute for Water and Atmospheric Research
PO Box 14-901, Wellington, New Zealand
|
|
|
|
| Re: Exceed and IDL [message #16189 is a reply to message #16186] |
Thu, 08 July 1999 00:00 |
Mark Hadfield
Messages: 783
Registered: May 1995
|
Senior Member |
|
|
David Fanning <davidf@dfanning.com> wrote in message
news:MPG.11ed28f44513782f989815@news.frii.com...
> Several obtuse reasons for setting RETAIN=0 in a draw widget
> come immediately to mind:
>
> (1) You are using object graphics in the window, in which
> case RETAIN should *always* be set to 0.
I disagree. For object graphics draw widgets on Windows NT, I have found the
best results (fast drawing, fast refresh when a window is re-exposed) are
achieved with:
RENDERER = 0 ; "Hardware" renderer
RETAIN=2 ; IDL provides backing store
EXPOSE_EVENTS=0 ; Not necessary for widget application to redraw its
own windows
I imagine that other platforms have different optimum settings, but you use
NT primarily, don't you David?
--
Mark Hadfield m.hadfield@niwa.cri.nz
National Institute for Water and Atmospheric Research
PO Box 14-901, Wellington, New Zealand
>
> (2) You know how screwy colors can be and you wish to
> protect the colors associated with this application
> by "restoring" your colors every time the widget is
> exposed.
>
> Cheers,
>
> David
>
> --
> David Fanning, Ph.D.
> Fanning Software Consulting
> Phone: 970-221-0438 E-Mail: davidf@dfanning.com
> Coyote's Guide to IDL Programming: http://www.dfanning.com/
> Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
| Re: Exceed and IDL [message #16191 is a reply to message #16186] |
Wed, 07 July 1999 00:00 |
davidf
Messages: 2866
Registered: September 1996
|
Senior Member |
|
|
Mark Hadfield (m.hadfield@niwa.cri.nz) writes:
>> (1) You are using object graphics in the window, in which
>> case RETAIN should *always* be set to 0.
>
> I disagree. For object graphics draw widgets on Windows NT, I have found the
> best results (fast drawing, fast refresh when a window is re-exposed) are
> achieved with:
>
> RENDERER = 0 ; "Hardware" renderer
> RETAIN=2 ; IDL provides backing store
> EXPOSE_EVENTS=0 ; Not necessary for widget application to redraw its
> own windows
>
> I imagine that other platforms have different optimum settings, but you use
> NT primarily, don't you David?
When objects first came out I noticed that my axes
labels would disappear. After about 50 frustrating hours
convincing myself that I was doing everything correctly, I
gave up and contacted RSI. I was told in no uncertain terms
that this was caused by setting the RETAIN keyword to 2
and that RETAIN should *always* be 0 in object graphics.
This made sense, because you already *have* the backing
store to repair the window: the view you want to display.
Why carry around yet more baggage? Sure enough, it
fixed the problem.
I was so scarred by this experience that I have just
been plodding along doing it this way from that moment
forward. Frankly, I'm shocked to hear there is another
way. :-)
With respect to rendering. I spent a pile of dough
getting two really nice 3D graphics cards that had
OpenGL support. *Both* of those cards render slower
in hardware than they do in software, although I have
never tried setting the RENDERER keyword on the
draw widget. I just always set the Hardware Rendering
option in the Option Settings. Of course, this is
one of those situations where *nobody* in technical
support, from *either* company, wants to have anything
to do with you. It's always *their* fault.
Let's just say I *knew* I should have bought the
Dell machine. :-(
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting
Phone: 970-221-0438 E-Mail: davidf@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
| Re: Exceed and IDL [message #16194 is a reply to message #16191] |
Wed, 07 July 1999 00:00 |
Richard G. French
Messages: 65
Registered: June 1997
|
Member |
|
|
Here's a few other values/settings
> that I've learned not to take for granted when an application runs (i.e.
> I don't assume any particular value).
>
> (1) DEVICE keywords:
> RETAIN
> DECOMPOSED
> SET_CHARACTER_SIZE
>
> (2) System variables:
> !ORDER (order for displaying images)
> !D.TABLE_SIZE (size of the color table)
> !D.N_COLORS (number of colors available)
> !D.NAME (name of the current display device)
> !D.WINDOW (index of the current graphics window)
>
> (3) Widget appearance:
> WIDGET_CONTROL keyword DEFAULT_FONT
>
> Anyone else care to add to the list?
>
I have been clobbered by using canned routines that set !P. !X, !Y, !Z
system variables for their plots. Then, when I am in the same session
and try my own interactive stuff, these stale values remain. I should
write a routine that I can call that cleans the slate of all set
variables and restores them to the defaults -- has someone done this
already?
Dick French
|
|
|
|
| Re: Exceed and IDL [message #16197 is a reply to message #16191] |
Wed, 07 July 1999 00:00 |
Liam Gumley
Messages: 473
Registered: November 1994
|
Senior Member |
|
|
David Fanning wrote:
> Several obtuse reasons for setting RETAIN=0 in a draw widget
> come immediately to mind:
>
> (1) You are using object graphics in the window, in which
> case RETAIN should *always* be set to 0.
>
> (2) You know how screwy colors can be and you wish to
> protect the colors associated with this application
> by "restoring" your colors every time the widget is
> exposed.
David points out some good reasons why an IDL programmer should not
assume a particular value for RETAIN. Here's a few other values/settings
that I've learned not to take for granted when an application runs (i.e.
I don't assume any particular value).
(1) DEVICE keywords:
RETAIN
DECOMPOSED
SET_CHARACTER_SIZE
(2) System variables:
!ORDER (order for displaying images)
!D.TABLE_SIZE (size of the color table)
!D.N_COLORS (number of colors available)
!D.NAME (name of the current display device)
!D.WINDOW (index of the current graphics window)
(3) Widget appearance:
WIDGET_CONTROL keyword DEFAULT_FONT
Anyone else care to add to the list?
Cheers,
Liam.
--
Liam E. Gumley
Space Science and Engineering Center, UW-Madison
http://cimss.ssec.wisc.edu/~gumley
|
|
|
|
| Re: Exceed and IDL [message #16198 is a reply to message #16191] |
Wed, 07 July 1999 00:00 |
davidf
Messages: 2866
Registered: September 1996
|
Senior Member |
|
|
bkrrrrr (beaker@llama.pilz.cack) writes:
> I copied your script to idl_startup.pro, and saw it
> run at startup. However, I then found the problem is actually
> caused by the person who wrote the program setting retain=0 when
> calling WIDGET_DRAW for some obtuse reason. grrrrr...
Several obtuse reasons for setting RETAIN=0 in a draw widget
come immediately to mind:
(1) You are using object graphics in the window, in which
case RETAIN should *always* be set to 0.
(2) You know how screwy colors can be and you wish to
protect the colors associated with this application
by "restoring" your colors every time the widget is
exposed.
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting
Phone: 970-221-0438 E-Mail: davidf@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
| Re: Exceed and IDL [message #16201 is a reply to message #16191] |
Wed, 07 July 1999 00:00 |
bkrrrrr
Messages: 3
Registered: July 1999
|
Junior Member |
|
|
Liam Gumley wrote:
>
> bkrrrrr wrote:
>>> Issuing the following command at the start of your IDL session should do
>>> the trick:
>>>
>>> DEVICE, DECOMPOSED=0, RETAIN=2
>>>
>> FWIW, this has never worked for me, in that
>> curves drawn in the work area are not retained
>> with retain=2 (or anything else).
>
> Strange. Did you issue the command
>
> DEVICE, RETAIN=2
[snip]
I copied your script to idl_startup.pro, and saw it
run at startup. However, I then found the problem is actually
caused by the person who wrote the program setting retain=0 when
calling WIDGET_DRAW for some obtuse reason. grrrrr...
thx,
bkr
|
|
|
|
| Re: Exceed and IDL [message #16202 is a reply to message #16191] |
Wed, 07 July 1999 00:00 |
Liam Gumley
Messages: 473
Registered: November 1994
|
Senior Member |
|
|
bkrrrrr wrote:
>> Issuing the following command at the start of your IDL session should do
>> the trick:
>>
>> DEVICE, DECOMPOSED=0, RETAIN=2
>>
> FWIW, this has never worked for me, in that
> curves drawn in the work area are not retained
> with retain=2 (or anything else).
Strange. Did you issue the command
DEVICE, RETAIN=2
*before* creating any graphics windows, just after starting IDL? This
command *must* be executed prior to creating a graphics window,
otherwise it will have no effect (at least it has no effect on my SGI
box running IDL 5.2). I recommend using this command in a startup file.
For example, all our local IDL users have the following startup file
(named idl_startup.pro):
;---cut here---
;- Set up 8 bit display and grab colors
if !version.os_family eq 'unix' then device, pseudo = 8
device, retain = 2, decomposed = 0
window, /free, /pixmap, colors = -5
plot, [ 0 ]
wdelete, !d.window
print, 'Color table size is ', !d.table_size
;- Set graphics and widget font sizes
device, set_character_size = [ 6, 9 ]
widget_control, default_font = '7x13'
;- ensure IDL 5.0 XMANAGER behavior is consistent with IDL 4.0
xmanager, catch = 0
;---cut here---
Then use the system variable IDL_STARTUP to point to the full path and
name of the startup file e.g.
setenv IDL_STARTUP $HOME/idl_startup.pro (C shell)
export IDL_STARTUP=$HOME/idl_startup.pro (Korn shell)
This code will be executed every time you start a new IDL session, and
it should give you a consistent well-behaved 8 bit IDL display on all
Unix platforms (as long as your Unix desktop is set to 8 bit or 24 bit
mode: if you use a 16 bit or 32 bit desktop, all bets are off).
Cheers,
Liam.
PS: Let me know if this does not work.
--
Liam E. Gumley
Space Science and Engineering Center, UW-Madison
http://cimss.ssec.wisc.edu/~gumley
|
|
|
|
| Re: Exceed and IDL [message #16209 is a reply to message #16191] |
Tue, 06 July 1999 00:00 |
bkrrrrr
Messages: 3
Registered: July 1999
|
Junior Member |
|
|
Liam Gumley wrote:
>
> Barry Latter wrote:
>> I've had problems with window buffering using Exceed on a PC for
>> displaying IDL plots. Basically, if a window is opened in front of an
>> IDL graphics window, the IDL window is blanked out when it's brought
>> back to the front. Is this an IDL or Exceed problem? Has anyone found a
>> way around it?
>
> Issuing the following command at the start of your IDL session should do
> the trick:
>
> DEVICE, DECOMPOSED=0, RETAIN=2
>
FWIW, this has never worked for me, in that
curves drawn in the work area are not retained
with retain=2 (or anything else).
Here is the device info, if it helps:
IDL> help, /device
Available Graphics Devices: CGM HP LJ NULL PCL PRINTER PS REGIS TEK X Z
Current graphics device: X
Server: X11.0, Sun Microsystems, Inc., Release 3600
Display Depth, Size: 8 bits, (1152,900)
Visual Class: PseudoColor (3)
Bits Per RGB: 8
Physical Color Map Entries (Used / Total): 256 / 256
Colormap: Private, 256 colors. Translation table: Enabled
Graphics pixels: Combined, Dither Method: Ordered
Write Mask: 255 (decimal) ff (hex)
Graphics Function: 3 (copy)
Current Font: <default>, Current TrueType Font: <default>
Default Backing Store: Pixmap.
thx,
bkr
|
|
|
|
| Re: Exceed and IDL [message #16218 is a reply to message #16209] |
Tue, 06 July 1999 00:00 |
Liam Gumley
Messages: 473
Registered: November 1994
|
Senior Member |
|
|
Grady Daub wrote:
>
> Sometimes I use Exceed to run a Common Desktop Environment on an IBM
> type computer. (Pentiums and all that.) After a while, I discovered that
> I had to type
>
> DEVICE, DECOMPOSE=0
>
> for the colors to work. (Thanks to RSI for letting me discover that
> while searching their site for something else and for RSI tech. support
> for not realizing that that was the solution to my problem.)
>
> Now, I've a new problem. I wrote a program that cycles the palette (and
> a few other things). It would create plasma if you have a plasma-like
> picture showing. (A color mandelbrot would work.)
>
> The problem is that,while using exceed, the colors don't cycle
> on-screen. The displayed picture (or, say, XLOADCT) colors stay the
> same. You have to reload the picture to see the final outcome of the
> palette change.
(1) Close Exceed.
(2) Reset your Windows desktop to 8 bit mode (Right click on the
Desktop, select Properties; Settings; Color Palette 256 colors).
(3) Restart Exceed.
(4) Restart IDL on the remote server.
Now changes to the color table should be reflected instantly onscreen.
Note that this is *only* possible in 8 bit mode.
Cheers,
Liam.
--
Liam E. Gumley
Space Science and Engineering Center, UW-Madison
http://cimss.ssec.wisc.edu/~gumley
|
|
|
|
| Re: Exceed and IDL [message #16219 is a reply to message #16218] |
Tue, 06 July 1999 00:00 |
Liam Gumley
Messages: 473
Registered: November 1994
|
Senior Member |
|
|
Barry Latter wrote:
> I've had problems with window buffering using Exceed on a PC for
> displaying IDL plots. Basically, if a window is opened in front of an
> IDL graphics window, the IDL window is blanked out when it's brought
> back to the front. Is this an IDL or Exceed problem? Has anyone found a
> way around it?
Issuing the following command at the start of your IDL session should do
the trick:
DEVICE, DECOMPOSED=0, RETAIN=2
Cheers,
Liam.
--
Liam E. Gumley
Space Science and Engineering Center, UW-Madison
http://cimss.ssec.wisc.edu/~gumley
|
|
|
|
| Re: Exceed and IDL [message #16221 is a reply to message #16219] |
Tue, 06 July 1999 00:00 |
Barry Latter
Messages: 1
Registered: July 1999
|
Junior Member |
|
|
Hi,
Different question regarding Exceed and IDL
I've had problems with window buffering using Exceed on a PC for
displaying IDL plots. Basically, if a window is opened in front of an
IDL graphics window, the IDL window is blanked out when it's brought
back to the front. Is this an IDL or Exceed problem? Has anyone found a
way around it?
Any help appreciated!
Baz
Department of Meteorology
Reading University
|
|
|
|
| Re: Exceed and IDL [message #16224 is a reply to message #16221] |
Mon, 05 July 1999 00:00 |
davidf
Messages: 2866
Registered: September 1996
|
Senior Member |
|
|
Grady Daub (gadZOOKS8371@garnet.acns.fsuMMER.edu) writes:
> Sometimes I use Exceed to run a Common Desktop Environment on an IBM
> type computer. (Pentiums and all that.) After a while, I discovered that
> I had to type
>
> DEVICE, DECOMPOSE=0
>
> for the colors to work. (Thanks to RSI for letting me discover that
> while searching their site for something else and for RSI tech. support
> for not realizing that that was the solution to my problem.)
What!? Don't those folks read this newsgroup? That is the
answer to 95% of the problems that have the word "color"
in their descriptions! :-)
> Now, I've a new problem.
> The problem is that,while using exceed, the colors don't cycle
> on-screen. The displayed picture (or, say, XLOADCT) colors stay the
> same. You have to reload the picture to see the final outcome of the
> palette change.
The answer to the other 5% of the color questions is "that
is how 24-bit color works". Here are some articles you
might want to read:
http://www.dfanning.com/documents/tips.html#UsingColors
Specifically, be sure you read these:
http://www.dfanning.com/tips/colors24.html
http://www.dfanning.com/tips/noxloadct.html
http://www.dfanning.com/tips/image24_colors.html
The kind of information you have been asking about lately is
covered in *extensive* detail in my book. If you are a
student, you might be interested in one of my "student
discounted books". I got some books back from Amazon
this weekend, whose covers had been damaged in shipment.
(More likely in *unpacking* the shipment, but they are
selling so many books for me, I'm not inclined to complain.)
In any case, these books don't look so good on the outside,
but they save you *months* of aggravation on the inside. :-)
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting
Phone: 970-221-0438 E-Mail: davidf@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
