Thanks, David.  This helps a bunch in trying to track things down on my 
MacOSX/Pismo.  (I am curious if anybody else is having any success on 
this front.)  I am not sure how this works or what it means, but it 
suggests to me that either I haven't "hooked things up right," or the 
Rage128 on the Pismo simply doesn't make the cut somehow.  the results 
of the TGraphicsTest was this:

VM: Mac OS - Squeak3.3alpha of 30 January 2002 [latest update: #4971] 
Squeak VM 3.2.3b2
Image: TeaSqueak3.2 [latest update: #276]

B3DHardwareEngine(Object)>>error:
[] in WorldState>>displayWorldSafely:
[] in BlockContext>>ifError:
MessageNotUnderstood(Exception)>>handlerAction
MessageNotUnderstood(Exception)>>signal
B3DHardwareEngine(Object)>>doesNotUnderstand:
OpenGLMorph>>glRenderOn:
OpenGLMorph>>renderOn:
OpenGLMorph>>drawAcceleratedOn:
OpenGLMorph>>drawOn:
FormCanvas(Canvas)>>draw:
FormCanvas(Canvas)>>drawMorph:
[] in OpenGLMorph(Morph)>>fullDrawOn:
FormCanvas>>roundCornersOf:in:during:
FormCanvas(Canvas)>>roundCornersOf:during:
OpenGLMorph(Morph)>>fullDrawOn:
FormCanvas(Canvas)>>fullDraw:
FormCanvas(Canvas)>>fullDrawMorph:
[] in WorldState>>drawWorld:submorphs:invalidAreasOn:
Rectangle>>allAreasOutsideList:startingAt:do:
Rectangle>>allAreasOutsideList:do:
[] in WorldState>>drawWorld:submorphs:invalidAreasOn:
Array(SequenceableCollection)>>do:
WorldState>>drawWorld:submorphs:invalidAreasOn:
[] in WorldState>>displayWorld:submorphs:
FormCanvas>>roundCornersOf:in:during:
FormCanvas(Canvas)>>roundCornersOf:during:
WorldState>>displayWorld:submorphs:
PasteUpMorph>>privateOuterDisplayWorld
PasteUpMorph>>displayWorld
[] in WorldState>>displayWorldSafely:
BlockContext>>on:do:
BlockContext>>ifError:
WorldState>>displayWorldSafely:
WorldState>>doOneCycleNowFor:
WorldState>>doOneCycleFor:
PasteUpMorph>>doOneCycle
[] in Project class>>spawnNewProcess
[] in BlockContext>>newProcess


ANALYSIS:
This hardware has NO accelerated OpenGL support whatsoever (bad news).
If the hardware is indeed 3D capable, you will need to obtain OpenGL 
drivers from your vendor.
This hardware has no sufficient depth buffer support (bad news).
A depth buffer of at least 16 bit is required by Tea.
This hardware has NO stencil buffer support (bad news).
Stencil buffer support is required by Tea.

SUMMARY: This system should be capable of running Tea just fine.


On Saturday, November 9, 2002, at 04:14 PM, David A. Smith wrote:

> That is correct - I kept the OpenGL morph around just in case I needed 
> to perform some very simple OpenGL tests. The GLX class has changed a 
> lot recently, so it is a bit out of date. In any case, not really 
> useful.
>
> There is a class that tests the OpenGL setup. Just run:
>
> TGraphicsTest open
>
> and click on the "run tests" at the top of the resulting window.
>
> David
>
>
> At 10:00 PM 11/9/2002 +0100, you wrote:
>> On Sat, 9 Nov 2002, Ned Konz wrote:
>>
>> > I built the VM as Bert described. I have GLX installed on my 
>> system, as
>> > well as libffi.
>> > When I drag the OpenGL morph from the Objects tool, I get the 
>> following
>> > walkback.
>> > It appears that aRenderer is a B3DHardwareEngine but should be a 
>> GLXUnix
>> > instead.
>>
>> As far as I can tell, the OpenGL morph is a left-over from debugging. 
>> You
>> should use the Teapot morph instead.
>>
>> -- Bert
>
>