| Re: Another VM conundrum [message #37277] |
Fri, 12 December 2003 09:21 |
btt
Messages: 345
Registered: December 2000
|
Senior Member |
|
|
David Fanning wrote:
> Ben Tupper writes:
>
snip
>>
>> I noticed that the terminal session I used changed from tcsh to idl and
>> the the x11 terminal shows two windows: an xTerm (which I have to have
>> up and ruu=ning but don't use for the idl command prompt) and viola!
>> Test_VM. So, unlike Windows, you can track down the the widget even
>> when it's hiding.
>
>
> Well, this is not exactly testing the functionality that
> I was talking about. Here is a modification that does.
> Here is what you want to do.
>
> 1. Maximize your IDLDE window.
>
> 2. Type "TEST_VM". The application appears on the display,
> and, yes, it has it's own application icon on the task
> bar at the bottom.
>
> 3. Click the "Test It" button. A TextBox dialog appears.
> Before you type text into this dialog, click in your
> IDLDE window. Whoops! The dialog disappears behind the
> IDLDE window and it is going to be *hard* to find!
> You should be able to get the Test_VM application
> back in front of the IDLDE, but you won't be able
> to *do* anything with it until you find that damn
> modal dialog!
>
Ok! I'm with you now. The textbox window was buried as you describe.
On a different note, I have to mention that I seriously doubt that
anyone using IDL on MacOSX is using IDLDE. Even cutting and pasting
is badly disabled in the editor window.
Ben
|
|
|
|
| Re: Another VM conundrum [message #37282 is a reply to message #37277] |
Fri, 12 December 2003 08:24  |
JD Smith
Messages: 850
Registered: December 1999
|
Senior Member |
|
|
On Fri, 12 Dec 2003 07:18:02 -0700, Ben Tupper wrote:
> JD Smith wrote:
>> On Thu, 11 Dec 2003 12:46:11 -0700, David Fanning wrote:
>>
>>
> SNIP
>
>> I'm thinking I can probably get away
>> with:
>>
>> device,DECOMPOSED=0,RETAIN=2
>>
>> which will always run since you need IDL>=6.0 to run the VM. What I
>> worry about is how this will interact with various other systems (like
>> Windows, Solaris, or MacOSX). Can anyone offer any feedback as to
>> whether this call succeeds in doing what I asked without creating other
>> problems?
>>
>>
> Hello,
>
> I tested the routine appended below as a VM on MacOSX X11. It seems to
> work fine when other windows cover/uncover and the window is
> minimized/restored.
>
> I noticed that the terminal session I used changed from tcsh to idl and
> the the x11 terminal shows two windows: an xTerm (which I have to have
> up and ruu=ning but don't use for the idl command prompt) and viola!
> Test_VM. So, unlike Windows, you can track down the the widget even
> when it's hiding.
>
> I wonder if you might have in mind a different kind of test - something
> we code build as VM-ware and test out. I'd be happy to give it a whirl
> if needed.
>
> Is this potential weakness/problem limited to direct graphics?
>
Thanks Ben. Glad to hear it works well enough. The test I wanted was
just the "device,DECOMPOSED=0,RETAIN=2" you use, and could have been
done in the VM or normal version. The reasons the VM even comes into
the question at all is that, with command-line IDL, you can just
instruct the user to use a DEVICE setting which works in their IDL
startup file, whereas in the IDLVM, no startup file is ever consulted,
so you must compile in a best-guess DEVICE call.
JD
|
|
|
|
| Re: Another VM conundrum [message #37283 is a reply to message #37282] |
Fri, 12 December 2003 07:25 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Ben Tupper writes:
> I tested the routine appended below as a VM on MacOSX X11. It seems to
> work fine when other windows cover/uncover and the window is
> minimized/restored.
>
> I noticed that the terminal session I used changed from tcsh to idl and
> the the x11 terminal shows two windows: an xTerm (which I have to have
> up and ruu=ning but don't use for the idl command prompt) and viola!
> Test_VM. So, unlike Windows, you can track down the the widget even
> when it's hiding.
Well, this is not exactly testing the functionality that
I was talking about. Here is a modification that does.
Here is what you want to do.
1. Maximize your IDLDE window.
2. Type "TEST_VM". The application appears on the display,
and, yes, it has it's own application icon on the task
bar at the bottom.
3. Click the "Test It" button. A TextBox dialog appears.
Before you type text into this dialog, click in your
IDLDE window. Whoops! The dialog disappears behind the
IDLDE window and it is going to be *hard* to find!
You should be able to get the Test_VM application
back in front of the IDLDE, but you won't be able
to *do* anything with it until you find that damn
modal dialog!
You will need the TextBox program from my web page
to run the following program:
http://www.dfanning.com/programs/textbox.pro
Cheers,
David
;*********************************************************** ***
Pro Test_VM_Event, event
Widget_Control, event.top, Get_UValue=labelID
theText = TextBox(Title='Can you see me now?', Label="Type Something")
Widget_Control, labelID, Set_Value=theText
END
PRO Test_VM
device,DECOMPOSED=0,RETAIN=2
file = FILEPATH('rose.jpg', SUBDIRECTORY=['examples','data'])
trueImage = read_image(file)
dim = SIZE(trueImage,/Dim)
byteImage = bytscl(hanning(50,50))
Base = Widget_Base(title = 'Test_VM', column=1)
Draw = Widget_Draw(base, xsize = dim[1], ysize = dim[2])
labelID = Widget_Label(base, Value='Text Here', /Dynamic)
button = Widget_Button(base, Value='Test It')
Widget_Control, base, /realize, Set_UValue=labelID
Widget_Control, draw, get_value = drawID
Wset, drawID
TV, trueImage, true = 1
TV, byteImage
XMANAGER, 'test_vm', base
END
;*********************************************************** ***
--
David W. Fanning, Ph.D.
Fanning Software Consulting, Inc.
Phone: 970-221-0438, E-mail: david@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
| Re: Another VM conundrum [message #37286 is a reply to message #37283] |
Fri, 12 December 2003 06:18 |
btt
Messages: 345
Registered: December 2000
|
Senior Member |
|
|
JD Smith wrote:
> On Thu, 11 Dec 2003 12:46:11 -0700, David Fanning wrote:
>
>
SNIP
> I'm thinking I can probably get away
> with:
>
> device,DECOMPOSED=0,RETAIN=2
>
> which will always run since you need IDL>=6.0 to run the VM. What I worry
> about is how this will interact with various other systems (like Windows,
> Solaris, or MacOSX). Can anyone offer any feedback as to whether this
> call succeeds in doing what I asked without creating other problems?
>
Hello,
I tested the routine appended below as a VM on MacOSX X11. It seems to
work fine when other windows cover/uncover and the window is
minimized/restored.
I noticed that the terminal session I used changed from tcsh to idl and
the the x11 terminal shows two windows: an xTerm (which I have to have
up and ruu=ning but don't use for the idl command prompt) and viola!
Test_VM. So, unlike Windows, you can track down the the widget even
when it's hiding.
I wonder if you might have in mind a different kind of test - something
we code build as VM-ware and test out. I'd be happy to give it a whirl
if needed.
Is this potential weakness/problem limited to direct graphics?
Ben
;;;;;;BEGIN CODE
PRO Test_VM
device,DECOMPOSED=0,RETAIN=2
file = FILEPATH('rose.jpg', SUBDIRECTORY=['examples','data'])
trueImage = read_image(file)
dim = SIZE(trueImage,/Dim)
byteImage = bytscl(hanning(50,50))
Base = Widget_Base(title = 'Test_VM')
Draw = Widget_Draw(base, xsize = dim[1], ysize = dim[2])
Widget_Control, base, /realize
Widget_Control, draw, get_value = drawID
Wset, drawID
TV, trueImage, true = 1
TV, byteImage
XMANAGER, 'test_vm', base
END
;;;;;;END CODE
|
|
|
|
| Re: Another VM conundrum [message #37290 is a reply to message #37286] |
Thu, 11 December 2003 15:05 |
Rick Towler
Messages: 821
Registered: August 1998
|
Senior Member |
|
|
"JD Smith" wrote in message...
> On Thu, 11 Dec 2003 12:46:11 -0700, David Fanning wrote:
>
>> JD Smith writes:
>>
>>> Having fixed my main-level XManager issues (thankfully fairly trivial),
>>> I now find myself with another problem. Typically, I recommend setting
>>> something like:
>>>
>>> device,DECOMPOSED=0,TRUE=16,RETAIN=2
>>>
<snip>
> Thanks David. What's different about this problem is I'm not interested
> in restoring the decomposed state, etc. since only the single program will
> be running under the VM; in this sense I can be very selfish and just set
> it to what I need -- if I knew what that was. What I want is a setup such
> that, for any combination of visual (TrueColor, PseudoColor, etc.), depth
> (8,16,24bit), and backing behavior (stored by server, stored by OS, stored
> by IDL), will give "correct" colors, and images which won't be clobbered
> when windows are drug across them. I'm thinking I can probably get away
> with:
>
> device,DECOMPOSED=0,RETAIN=2
>
> which will always run since you need IDL>=6.0 to run the VM. What I worry
> about is how this will interact with various other systems (like Windows,
> Solaris, or MacOSX). Can anyone offer any feedback as to whether this
> call succeeds in doing what I asked without creating other problems?
FWIW, I have been throwing in the same line with the direct graphics
applications I have been compiling for the VM and haven't run into any
issues. Tested on 8 and 32bit windows systems, 24bit X (solaris), and 8bit
X (win32 X server).
-Rick
|
|
|
|
| Re: Another VM conundrum [message #37291 is a reply to message #37290] |
Thu, 11 December 2003 14:14 |
JD Smith
Messages: 850
Registered: December 1999
|
Senior Member |
|
|
On Thu, 11 Dec 2003 12:46:11 -0700, David Fanning wrote:
> JD Smith writes:
>
>> Having fixed my main-level XManager issues (thankfully fairly trivial),
>> I now find myself with another problem. Typically, I recommend setting
>> something like:
>>
>> device,DECOMPOSED=0,TRUE=16,RETAIN=2
>>
>> in your IDL startup file, where RETAIN=2 is good for most Linux
>> machines (whose window managers don't provide decent backing store),
>> DECOMPOSED=0 is immortalized on David's site, and TRUE=16 is just for
>> good measure, for those X servers which *claim* to offer DirectColor
>> but don't really (sadly, not few in the Linux world). However, I leave
>> it up to the end user to find some combination of device incantations
>> which work for them.
>>
>> That's fine for interactive usage, but with a runtime/VM application,
>> the startup file is never consulted, so you need to perform these
>> device incantations yourself in the code. Has anyone solved this
>> problem in a platform-independent way? Would something like:
>>
>> device,DECOMPOSED=0,RETAIN=2
>>
>> always be safe, even for older 8-bit PSEUDOCOLOR machines? Or is there
>> some method of interrogating DEVICE for info before making a final
>> decision on RETAIN and color-model settings? What guidance from the
>> IDL color gurus?
>
> Oh, I *hate* to put those DEVICE commands into the code! A pure
> nightmare when you are trying to write code that works *everywhere*:
> Z-buffer, PostScript, on various and sundry displays, etc. Here are a
> couple of tips (taken from the TVIMAGE code, of course, which *does* run
> everywhere).
>
>
Thanks David. What's different about this problem is I'm not interested
in restoring the decomposed state, etc. since only the single program will
be running under the VM; in this sense I can be very selfish and just set
it to what I need -- if I knew what that was. What I want is a setup such
that, for any combination of visual (TrueColor, PseudoColor, etc.), depth
(8,16,24bit), and backing behavior (stored by server, stored by OS, stored
by IDL), will give "correct" colors, and images which won't be clobbered
when windows are drug across them. I'm thinking I can probably get away
with:
device,DECOMPOSED=0,RETAIN=2
which will always run since you need IDL>=6.0 to run the VM. What I worry
about is how this will interact with various other systems (like Windows,
Solaris, or MacOSX). Can anyone offer any feedback as to whether this
call succeeds in doing what I asked without creating other problems?
Thanks,
JD
|
|
|
|
| Re: Another VM conundrum [message #37294 is a reply to message #37291] |
Thu, 11 December 2003 11:46 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
JD Smith writes:
> Having fixed my main-level XManager issues (thankfully fairly
> trivial), I now find myself with another problem. Typically, I
> recommend setting something like:
>
> device,DECOMPOSED=0,TRUE=16,RETAIN=2
>
> in your IDL startup file, where RETAIN=2 is good for most Linux
> machines (whose window managers don't provide decent backing store),
> DECOMPOSED=0 is immortalized on David's site, and TRUE=16 is just for
> good measure, for those X servers which *claim* to offer DirectColor
> but don't really (sadly, not few in the Linux world). However, I
> leave it up to the end user to find some combination of device
> incantations which work for them.
>
> That's fine for interactive usage, but with a runtime/VM application,
> the startup file is never consulted, so you need to perform these
> device incantations yourself in the code. Has anyone solved this
> problem in a platform-independent way? Would something like:
>
> device,DECOMPOSED=0,RETAIN=2
>
> always be safe, even for older 8-bit PSEUDOCOLOR machines? Or is
> there some method of interrogating DEVICE for info before making a
> final decision on RETAIN and color-model settings? What guidance from
> the IDL color gurus?
Oh, I *hate* to put those DEVICE commands into the code!
A pure nightmare when you are trying to write code that
works *everywhere*: Z-buffer, PostScript, on various and
sundry displays, etc. Here are a couple of tips (taken
from the TVIMAGE code, of course, which *does* run
everywhere).
; Which release of IDL is this?
thisRelease = Float(!Version.Release)
; Decomposed color off if device supports it.
CASE StrUpCase(!D.NAME) OF
'X': BEGIN
Device, Get_Visual_Depth=thisDepth
IF thisRelease GE 5.2 THEN $
Device, Get_Decomposed=thisDecomposed
Device, Decomposed=0
ENDCASE
'WIN': BEGIN
Device, Get_Visual_Depth=thisDepth
IF thisRelease GE 5.2 THEN $
Device, Get_Decomposed=thisDecomposed
Device, Decomposed=0
ENDCASE
'MAC': BEGIN
Device, Get_Visual_Depth=thisDepth
IF thisRelease GE 5.2 THEN $
Device, Get_Decomposed=thisDecomposed
Device, Decomposed=0
ENDCASE
ELSE: thisDepth = 8
ENDCASE
; Restore Decomposed state if necessary.
CASE StrUpCase(!D.NAME) OF
'X': BEGIN
IF thisRelease GE 5.2 THEN Device, Decomposed=thisDecomposed
ENDCASE
'WIN': BEGIN
IF thisRelease GE 5.2 THEN Device, Decomposed=thisDecomposed
ENDCASE
'MAC': BEGIN
IF thisRelease GE 5.2 THEN BEGIN
Device, Decomposed=thisDecomposed
Cheers,
David
--
David W. Fanning, Ph.D.
Fanning Software Consulting, Inc.
Phone: 970-221-0438, E-mail: david@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
