New version of Tweak-OpenGL package

> A new version of the Tweak-OpenGL package is on the Hedgehog
> PublicContribs repository.  The required package OpenGL-OpenGL needs
> an "OGLExtManager initialize" to be done after it is loaded.  From the
> package comments:
>
> -----------------------------
> Name: Tweak-OpenGL-daf.6
> Author: daf
> Time: 4 July 2007, 11:17:06 am
> Ancestors: Tweak-OpenGL-daf.5
>
> A few fragment shaders to choose from.
>
> Requires the OpenGL-OpenGL package from the PublicContribs repository
> to be installed first.
>
> Install both packages, then in Squeak open a Tweak Project, enter it,
> get a Workspace from the Tools menu, enter "Wrinkle1 open" and doIt.
> Don't complain if your video card doesn't support GLSL shader
> programs.
>
> A Tweak Project Window can be used instead for better debugging, but
> the offsets are screwed up a bit.
> -----------------------------
>
> Try it out and let me know what you think.
>

> Hi David, followed your instructions on 1.0.18 image (updated, on WinXP SP2, AMD 64 3000+, see
> below for gfx) in a Tweak project window and it worked, so +1 for that :) However on creating a
> tweak project I first entered the project then went immediately back to morphic to save the image.
> Shortly after the save completed Croquet hung or went into an endless loop or something. I can
> interrupt it but cannot gain any further control, ie, it is now unresponsive to mouse clicks.
> Unfortunately I hadn't done much in this image except apply updates and run a few tests so I had
> not got around to backing it up. Same situation even if I restart the vm, not even time to take
> remedial action. I let it run for a minute or two and while it does not appear to be eating memory
> it does hog the CPU at around 99%.
>
> If you know of any way to gain back control and figure out what has gone wrong then I'm happy to
> help track down the bug, otherwise I'll start with a fresh image sometime.
>
> > GPUInfo new chkGLSLSupport
>
> OpenGL drivers
> version: 2.0.6590 WinXP Release
> vendor: ATI Technologies Inc.
> renderer: ATI MOBILITY RADEON X700 x86/MMX/3DNow!/SSE2
> GLSL is supported, hurray!
>
>
>
>
>
>      _____________________________________________________
> Do not compromise. Get unlimited storage and first rate spam protection with Yahoo! Mail.
> http://uk.mail.yahoo.com
>

> Another new version.  Let me know what you think.  I crave  
> feedback, +1 or
> -100.  From the package comments:
> -----
> Some error handling that actually works and a few fragment shaders  
> to choose
> from.
>
> Requires the OpenGL-OpenGL package from the PublicContribs  
> repository to be
> installed first.  It is necessary to manually do "OGLExtManager  
> initialize"
> after loading the OpenGL-OpenGL package.
>
> This package also depends on a fix for Mantis #6553, which can  
> temporarily be a
> pretty trivial do-it-yourself kind of thing. Just change  
> VectorColor#asVector4
> so that it returns a Vector4 instance instead of a Vector3.
>
> Install both packages, do the couple of manual items above, then in  
> Squeak open
> a Tweak Project, enter it, get a Workspace from the Tools menu,  
> enter "GPUInfo
> new chkGLSLSupport" and doIt.  If this says that you have GLSL  
> support, then
> enter "Wrinkle1 open" and doIt.
>
> A Tweak Project Window can be used instead of a Tweak Project for  
> somewhat
> better debugging, but the offsets are screwed up a bit.
> -----
> It takes a couple of minutes to do the GLSL compiles on my machine.

> Looks great David. Any idea why it takes so long to set up?
>
> On Jul 14, 2007, at 10:10 AM, David Faught wrote:
>
> > Another new version.  Let me know what you think.  I crave
> > feedback, +1 or
> > -100.  From the package comments:
> > -----
> > Some error handling that actually works and a few fragment shaders
> > to choose
> > from.
> >
> > Requires the OpenGL-OpenGL package from the PublicContribs
> > repository to be
> > installed first.  It is necessary to manually do "OGLExtManager
> > initialize"
> > after loading the OpenGL-OpenGL package.
> >
> > This package also depends on a fix for Mantis #6553, which can
> > temporarily be a
> > pretty trivial do-it-yourself kind of thing. Just change
> > VectorColor#asVector4
> > so that it returns a Vector4 instance instead of a Vector3.
> >
> > Install both packages, do the couple of manual items above, then in
> > Squeak open
> > a Tweak Project, enter it, get a Workspace from the Tools menu,
> > enter "GPUInfo
> > new chkGLSLSupport" and doIt.  If this says that you have GLSL
> > support, then
> > enter "Wrinkle1 open" and doIt.
> >
> > A Tweak Project Window can be used instead of a Tweak Project for
> > somewhat
> > better debugging, but the offsets are screwed up a bit.
> > -----
> > It takes a couple of minutes to do the GLSL compiles on my machine.
>
>

> David Faught wrote:
>
>> Note that there is an environmental difference in opening a "Tweak
>> Project" and opening a "Tweak Project Window" from the open menu.  I
>> have not had the same problems with a Tweak Project Window, and in
>> fact it generally seems to cooperate with Morphic more cleanly.  
>> It is
>> actually much easier to develop and test code using a Tweak Project
>> Window because of this cleaner interaction with Morphic.  The only
>> problem with this approach is that I can't seem to figure out the
>> proper screen offsets for the OpenGL player in this environment.
>
> This is actually easy to fix. All you need to do is to use a  
> subclass of CPanel which will automatically get its "window" iVar  
> to be populated with the CProjectMorph you are using. So, for example:
>
> CPanel subclass: #COpenGLPanel
>   instanceVariableNames: ''
>   classVariableNames: ''
>   poolDictionaries: ''
>   category: 'Tweak-OpenGL'
>
> COpenGLPanel>>oglBounds
>   "Answer the bounds for OpenGL rendering"
>   ^self globalBounds translateBy: (window boundsInWorld origin +  
> window borderWidth).
>
> This is really all you need.

>
> On 7/16/07, Peter Moore <[hidden email]> wrote:
>> Looks great David. Any idea why it takes so long to set up?
>>
>> On Jul 14, 2007, at 10:10 AM, David Faught wrote:
>>
>> > Another new version.  Let me know what you think.  I crave
>> > feedback, +1 or
>> > -100.  From the package comments:
>> > -----
>> > Some error handling that actually works and a few fragment shaders
>> > to choose
>> > from.
>> >
>> > Requires the OpenGL-OpenGL package from the PublicContribs
>> > repository to be
>> > installed first.  It is necessary to manually do "OGLExtManager
>> > initialize"
>> > after loading the OpenGL-OpenGL package.
>> >
>> > This package also depends on a fix for Mantis #6553, which can
>> > temporarily be a
>> > pretty trivial do-it-yourself kind of thing. Just change
>> > VectorColor#asVector4
>> > so that it returns a Vector4 instance instead of a Vector3.
>> >
>> > Install both packages, do the couple of manual items above, then in
>> > Squeak open
>> > a Tweak Project, enter it, get a Workspace from the Tools menu,
>> > enter "GPUInfo
>> > new chkGLSLSupport" and doIt.  If this says that you have GLSL
>> > support, then
>> > enter "Wrinkle1 open" and doIt.
>> >
>> > A Tweak Project Window can be used instead of a Tweak Project for
>> > somewhat
>> > better debugging, but the offsets are screwed up a bit.
>> > -----
>> > It takes a couple of minutes to do the GLSL compiles on my machine.
>>
>>

> On Jul 16, 2007, at 20:33 , Andreas Raab wrote:
> > David Faught wrote:
> >
> >> Note that there is an environmental difference in opening a "Tweak
> >> Project" and opening a "Tweak Project Window" from the open menu.  I
> >> have not had the same problems with a Tweak Project Window, and in
> >> fact it generally seems to cooperate with Morphic more cleanly.
> >> It is
> >> actually much easier to develop and test code using a Tweak Project
> >> Window because of this cleaner interaction with Morphic.  The only
> >> problem with this approach is that I can't seem to figure out the
> >> proper screen offsets for the OpenGL player in this environment.
> >
> > This is actually easy to fix. All you need to do is to use a
> > subclass of CPanel which will automatically get its "window" iVar
> > to be populated with the CProjectMorph you are using. So, for example:
> >
> > CPanel subclass: #COpenGLPanel
> >   instanceVariableNames: ''
> >   classVariableNames: ''
> >   poolDictionaries: ''
> >   category: 'Tweak-OpenGL'
> >
> > COpenGLPanel>>oglBounds
> >   "Answer the bounds for OpenGL rendering"
> >   ^self globalBounds translateBy: (window boundsInWorld origin +
> > window borderWidth).
> >
> > This is really all you need.
>
> I wonder why my Sensor/hand hack doesn't work anymore. "self hand"
> appears to be nil in the drawOn: method. Why?

> > I have not done any profiling to find out, but my guess is that the
> > GLSL compiler just isn't very swift.  I have seen some comments to
> > that effect on some of the OGL/GLSL support lists.  It seems to get
> > noticably longer with longer shader programs, which is a bad thing
> > since you can't really save the compiled programs.
>
> Hey David,
>
> In the C++ programs I've written they load and compile almost
> instantly (at least on the Mac) for even some decently long shaders.
> When working with shaders in Croquet I notice performance is roughly
> equal for everything but creating the first Program Object (eg the
> first piece of code to be created when setting up shaders).  I assume
> this has to do with deeper elements of Croquet preparing/initializing
> the GL layer in some way before the GL driver comes into play.
>
> So my question is, what's going on there.
>
> Cheers,
> Derek
>
> >
> > On 7/16/07, Peter Moore <[hidden email]> wrote:
> >> Looks great David. Any idea why it takes so long to set up?
> >>
> >> On Jul 14, 2007, at 10:10 AM, David Faught wrote:
> >>
> >> > Another new version.  Let me know what you think.  I crave
> >> > feedback, +1 or
> >> > -100.  From the package comments:
> >> > -----
> >> > Some error handling that actually works and a few fragment shaders
> >> > to choose
> >> > from.
> >> >
> >> > Requires the OpenGL-OpenGL package from the PublicContribs
> >> > repository to be
> >> > installed first.  It is necessary to manually do "OGLExtManager
> >> > initialize"
> >> > after loading the OpenGL-OpenGL package.
> >> >
> >> > This package also depends on a fix for Mantis #6553, which can
> >> > temporarily be a
> >> > pretty trivial do-it-yourself kind of thing. Just change
> >> > VectorColor#asVector4
> >> > so that it returns a Vector4 instance instead of a Vector3.
> >> >
> >> > Install both packages, do the couple of manual items above, then in
> >> > Squeak open
> >> > a Tweak Project, enter it, get a Workspace from the Tools menu,
> >> > enter "GPUInfo
> >> > new chkGLSLSupport" and doIt.  If this says that you have GLSL
> >> > support, then
> >> > enter "Wrinkle1 open" and doIt.
> >> >
> >> > A Tweak Project Window can be used instead of a Tweak Project for
> >> > somewhat
> >> > better debugging, but the offsets are screwed up a bit.
> >> > -----
> >> > It takes a couple of minutes to do the GLSL compiles on my machine.
> >>
> >>
>
>

> David Faught wrote:
> > Your hack works fine in a Tweak Project, but not in a Tweak Project
> > Window.  I have not looked too deeply into the differences between the
> > two, but I can tell that there are differences.  Andreas' approach is
> > a little cleaner, but I haven't yet gotten it to work in a Tweak
> > Project Window either.
>
> So what's the problem? You should be able to calculate the offset simply
> via "window boundsInWorld origin". Perhaps you're not launching your app
> correctly? The panel needs to be fired up via
>
>        CProjectMorph open: COpenGLPanel.
>
> In any case, let me know what the problem is.

> - 77160 tallies, 80837 msec.
>
> **Tree**
> 99.3% {80271ms} OGLWin32(OpenGL)>>glCreateShaderObjectARB:
>  99.3% {80271ms} *OGLExtManager(OGLExtManager)>>doesNotUnderstand:
>    99.3% {80271ms} *OGLExtManager(OGLExtManager)>>loadExtension:
>      99.2% {80190ms} *OGLExtManager
> class(Behavior)>>addSelector:withMethod:
>        99.2% {80190ms} *OGLExtManager
> class(ClassDescription)>>addSelector:withMethod:notifying:
>          99.2% {80190ms}
> SystemChangeNotifier>>methodAdded:selector:inClass:requestor:
>            99.2% {80190ms} SystemChangeNotifier>>trigger:
>              99.2% {80190ms} AddedEvent(AbstractEvent)>>trigger:
>                99.2% {80190ms}
> SystemEventManager(Object)>>triggerEvent:with:
>                  99.2% {80190ms}
> SystemEventManager(Object)>>triggerEvent:withArguments:
>                    99.2% {80190ms}
> WeakActionSequenceTrappingErrors>>valueWithArguments:
>                      99.2% {80190ms}
> WeakActionSequenceTrappingErrors>>valueWithArgum...startingFrom:
>                        99.2% {80190ms} WeakMessageSend>>valueWithArguments:
>                          99.2% {80190ms} MCWorkingCopy
> class(MCPackageManager class)>>methodModified:
>                            99.2% {80190ms} MCWorkingCopy
> class(MCPackageManager class)>>managersForClass:selector:do:
>                              99.2% {80190ms} MCWorkingCopy
> class(MCPackageManager class)>>managersForClass:category:do:
>                                99.2% {80190ms} MCWorkingCopy
> class(MCPackageManager class)>>managersForClass:do:
> [98.8% {79867ms} PackageInfo>>includesClass:
> [  98.7% {79786ms} *OGLExtManager class(Class)>>category
> [    89.5% {72349ms} SystemOrganizer(Categorizer)>>categoryOfElement:
> [      |89.5% {72349ms}
> SystemOrganizer(Categorizer)>>numberOfCategoryOfElement:
> [      |  51.8% {41874ms} ByteString(String)>>=
> [      |    |36.4% {29425ms} primitives
> [      |    |15.4% {12449ms} ByteString(String)>>compare:caseSensitive:
> [      |  37.8% {30556ms} primitives
> [    9.0% {7275ms} SystemOrganizer(Categorizer)>>listAtCategoryNamed:
> [      9.0% {7275ms} Array(SequenceableCollection)>>indexOf:ifAbsent:
> [        9.0% {7275ms}
> Array(SequenceableCollection)>>indexOf:startingAt:ifAbsent:
> [          6.0% {4850ms} ByteSymbol(Symbol)>>=
> [            |3.3% {2668ms} primitives
> [            |2.7% {2183ms} ByteSymbol(String)>>=
> [          3.0% {2425ms} primitives
> 
> **Leaves**
> 39.2% {31688ms} ByteSymbol(String)>>=
> 37.8% {30556ms} SystemOrganizer(Categorizer)>>numberOfCategoryOfElement:
> 15.4% {12449ms} ByteString(String)>>compare:caseSensitive:
> 3.3% {2668ms} ByteSymbol(Symbol)>>=
> 3.0% {2425ms} Array(SequenceableCollection)>>indexOf:startingAt:ifAbsent:
>
> **Memory**
>     old            +0 bytes
>     young        +298,768 bytes
>     used        +298,768 bytes
>     free        -298,768 bytes
>
> **GCs**
>     full            0 totalling 0ms (0.0% uptime)
>     incr        1948 totalling 3,992ms (5.0% uptime), avg 2.0ms
>     tenures        0
>     root table    0 overflows
>
>
>
> On 7/16/07, Derek Arndt <[hidden email]> wrote:
>> > I have not done any profiling to find out, but my guess is that the
>> > GLSL compiler just isn't very swift.  I have seen some comments to
>> > that effect on some of the OGL/GLSL support lists.  It seems to get
>> > noticably longer with longer shader programs, which is a bad thing
>> > since you can't really save the compiled programs.
>>
>> Hey David,
>>
>> In the C++ programs I've written they load and compile almost
>> instantly (at least on the Mac) for even some decently long shaders.
>> When working with shaders in Croquet I notice performance is roughly
>> equal for everything but creating the first Program Object (eg the
>> first piece of code to be created when setting up shaders).  I assume
>> this has to do with deeper elements of Croquet preparing/initializing
>> the GL layer in some way before the GL driver comes into play.
>>
>> So my question is, what's going on there.
>>
>> Cheers,
>> Derek
>>
>> >
>> > On 7/16/07, Peter Moore <[hidden email]> wrote:
>> >> Looks great David. Any idea why it takes so long to set up?
>> >>
>> >> On Jul 14, 2007, at 10:10 AM, David Faught wrote:
>> >>
>> >> > Another new version.  Let me know what you think.  I crave
>> >> > feedback, +1 or
>> >> > -100.  From the package comments:
>> >> > -----
>> >> > Some error handling that actually works and a few fragment shaders
>> >> > to choose
>> >> > from.
>> >> >
>> >> > Requires the OpenGL-OpenGL package from the PublicContribs
>> >> > repository to be
>> >> > installed first.  It is necessary to manually do "OGLExtManager
>> >> > initialize"
>> >> > after loading the OpenGL-OpenGL package.
>> >> >
>> >> > This package also depends on a fix for Mantis #6553, which can
>> >> > temporarily be a
>> >> > pretty trivial do-it-yourself kind of thing. Just change
>> >> > VectorColor#asVector4
>> >> > so that it returns a Vector4 instance instead of a Vector3.
>> >> >
>> >> > Install both packages, do the couple of manual items above, then in
>> >> > Squeak open
>> >> > a Tweak Project, enter it, get a Workspace from the Tools menu,
>> >> > enter "GPUInfo
>> >> > new chkGLSLSupport" and doIt.  If this says that you have GLSL
>> >> > support, then
>> >> > enter "Wrinkle1 open" and doIt.
>> >> >
>> >> > A Tweak Project Window can be used instead of a Tweak Project for
>> >> > somewhat
>> >> > better debugging, but the offsets are screwed up a bit.
>> >> > -----
>> >> > It takes a couple of minutes to do the GLSL compiles on my machine.
>> >>
>> >>
>>
>>
>

> David Faught wrote:
>> Okay, here is a simple profile of my loadshader method, which should
>> happen just once for the first render.  This looks pretty hinkey to
>> me, but I'm not sure exactly what to make of it.  Someone more
>> knowledgable care to comment?
>
> Ouch. This is a bug that I had long fixed and just forgotten. I  
> just posted it again to the Hedgehog repository. If you load the  
> updated OpenGL package this problem will be gone.
>
> Cheers,
>   - Andreas
>
>> - 77160 tallies, 80837 msec.
>> **Tree**
>> 99.3% {80271ms} OGLWin32(OpenGL)>>glCreateShaderObjectARB:
>>  99.3% {80271ms} *OGLExtManager(OGLExtManager)>>doesNotUnderstand:
>>    99.3% {80271ms} *OGLExtManager(OGLExtManager)>>loadExtension:
>>      99.2% {80190ms} *OGLExtManager class(Behavior)
>> >>addSelector:withMethod:
>>        99.2% {80190ms} *OGLExtManager
>> class(ClassDescription)>>addSelector:withMethod:notifying:
>>          99.2% {80190ms}
>> SystemChangeNotifier>>methodAdded:selector:inClass:requestor:
>>            99.2% {80190ms} SystemChangeNotifier>>trigger:
>>              99.2% {80190ms} AddedEvent(AbstractEvent)>>trigger:
>>                99.2% {80190ms} SystemEventManager(Object)
>> >>triggerEvent:with:
>>                  99.2% {80190ms}
>> SystemEventManager(Object)>>triggerEvent:withArguments:
>>                    99.2% {80190ms}
>> WeakActionSequenceTrappingErrors>>valueWithArguments:
>>                      99.2% {80190ms}
>> WeakActionSequenceTrappingErrors>>valueWithArgum...startingFrom:
>>                        99.2% {80190ms}  
>> WeakMessageSend>>valueWithArguments:
>>                          99.2% {80190ms} MCWorkingCopy
>> class(MCPackageManager class)>>methodModified:
>>                            99.2% {80190ms} MCWorkingCopy
>> class(MCPackageManager class)>>managersForClass:selector:do:
>>                              99.2% {80190ms} MCWorkingCopy
>> class(MCPackageManager class)>>managersForClass:category:do:
>>                                99.2% {80190ms} MCWorkingCopy
>> class(MCPackageManager class)>>managersForClass:do:
>> [98.8% {79867ms} PackageInfo>>includesClass:
>> [  98.7% {79786ms} *OGLExtManager class(Class)>>category
>> [    89.5% {72349ms} SystemOrganizer(Categorizer)>>categoryOfElement:
>> [      |89.5% {72349ms} SystemOrganizer(Categorizer)
>> >>numberOfCategoryOfElement:
>> [      |  51.8% {41874ms} ByteString(String)>>=
>> [      |    |36.4% {29425ms} primitives
>> [      |    |15.4% {12449ms} ByteString(String)
>> >>compare:caseSensitive:
>> [      |  37.8% {30556ms} primitives
>> [    9.0% {7275ms} SystemOrganizer(Categorizer)>>listAtCategoryNamed:
>> [      9.0% {7275ms} Array(SequenceableCollection)>>indexOf:ifAbsent:
>> [        9.0% {7275ms}
>> Array(SequenceableCollection)>>indexOf:startingAt:ifAbsent:
>> [          6.0% {4850ms} ByteSymbol(Symbol)>>=
>> [            |3.3% {2668ms} primitives
>> [            |2.7% {2183ms} ByteSymbol(String)>>=
>> [          3.0% {2425ms} primitives
>> **Leaves**
>> 39.2% {31688ms} ByteSymbol(String)>>=
>> 37.8% {30556ms} SystemOrganizer(Categorizer)
>> >>numberOfCategoryOfElement:
>> 15.4% {12449ms} ByteString(String)>>compare:caseSensitive:
>> 3.3% {2668ms} ByteSymbol(Symbol)>>=
>> 3.0% {2425ms} Array(SequenceableCollection)
>> >>indexOf:startingAt:ifAbsent:
>> **Memory**
>>     old            +0 bytes
>>     young        +298,768 bytes
>>     used        +298,768 bytes
>>     free        -298,768 bytes
>> **GCs**
>>     full            0 totalling 0ms (0.0% uptime)
>>     incr        1948 totalling 3,992ms (5.0% uptime), avg 2.0ms
>>     tenures        0
>>     root table    0 overflows
>> On 7/16/07, Derek Arndt <[hidden email]> wrote:
>>> > I have not done any profiling to find out, but my guess is that  
>>> the
>>> > GLSL compiler just isn't very swift.  I have seen some comments to
>>> > that effect on some of the OGL/GLSL support lists.  It seems to  
>>> get
>>> > noticably longer with longer shader programs, which is a bad thing
>>> > since you can't really save the compiled programs.
>>>
>>> Hey David,
>>>
>>> In the C++ programs I've written they load and compile almost
>>> instantly (at least on the Mac) for even some decently long shaders.
>>> When working with shaders in Croquet I notice performance is roughly
>>> equal for everything but creating the first Program Object (eg the
>>> first piece of code to be created when setting up shaders).  I  
>>> assume
>>> this has to do with deeper elements of Croquet preparing/
>>> initializing
>>> the GL layer in some way before the GL driver comes into play.
>>>
>>> So my question is, what's going on there.
>>>
>>> Cheers,
>>> Derek
>>>
>>> >
>>> > On 7/16/07, Peter Moore <[hidden email]> wrote:
>>> >> Looks great David. Any idea why it takes so long to set up?
>>> >>
>>> >> On Jul 14, 2007, at 10:10 AM, David Faught wrote:
>>> >>
>>> >> > Another new version.  Let me know what you think.  I crave
>>> >> > feedback, +1 or
>>> >> > -100.  From the package comments:
>>> >> > -----
>>> >> > Some error handling that actually works and a few fragment  
>>> shaders
>>> >> > to choose
>>> >> > from.
>>> >> >
>>> >> > Requires the OpenGL-OpenGL package from the PublicContribs
>>> >> > repository to be
>>> >> > installed first.  It is necessary to manually do "OGLExtManager
>>> >> > initialize"
>>> >> > after loading the OpenGL-OpenGL package.
>>> >> >
>>> >> > This package also depends on a fix for Mantis #6553, which can
>>> >> > temporarily be a
>>> >> > pretty trivial do-it-yourself kind of thing. Just change
>>> >> > VectorColor#asVector4
>>> >> > so that it returns a Vector4 instance instead of a Vector3.
>>> >> >
>>> >> > Install both packages, do the couple of manual items above,  
>>> then in
>>> >> > Squeak open
>>> >> > a Tweak Project, enter it, get a Workspace from the Tools menu,
>>> >> > enter "GPUInfo
>>> >> > new chkGLSLSupport" and doIt.  If this says that you have GLSL
>>> >> > support, then
>>> >> > enter "Wrinkle1 open" and doIt.
>>> >> >
>>> >> > A Tweak Project Window can be used instead of a Tweak  
>>> Project for
>>> >> > somewhat
>>> >> > better debugging, but the offsets are screwed up a bit.
>>> >> > -----
>>> >> > It takes a couple of minutes to do the GLSL compiles on my  
>>> machine.
>>> >>
>>> >>
>>>
>>>

> David Faught wrote:
> >> In any case, let me know what the problem is.
> >
> > Thanks for your help!  The error that I get in your oglBounds method is:
> >
> > MessageNotUnderstood: CWindow>>boundsInWorld
>
> Oh. You must have launched this from within Tweak project builder. This
> is not what I meant. The example was meant as a sort of "standalone app"
> in Morphic that gets run without the overhead implied by
> CProjectBuilder. In other words you need to launch it from Morphic not
> from inside Tweak.