Dashboard

[Lisa Sobierajski Avila]

Hello Ken,

Here are the failures I don't understand:

irix65: is this using the opengl version of the image mapper and text 
mapper, or is this using the x version. If it is using the x version, then 
the results make sense, and the solution is to simply create images for 
this platform. (x and win32 do not support the foreground / background 
location flag- just opengl) Of course, if it is using x, why is the 
TestFGBGPolyDataMapper2D.tcl test working?

solaris: is alpha being treated differently here when putting up images? I 
don't know that there is any sun-specific code, but maybe a difference in 
how opengl is implemented? It seems like the background is showing through 
the image in TestFGBGImageMapper.tcl. Also - it looks like all outlines are 
being drawn in the background in TestFGBGPolyDataMapper2D.tcl - not the 
expected behavior.

solarisCoverage: behaves differently than solaris. What is the difference 
with these builds? Maybe this is the same as the irix65 build in that it is 
using x versions of the text and image mappers? In which case we just need 
platform specific images. Again, why is TestFGBGPolyDataMapper2D .tcl working?

All the remaining problems with TestFGBGTextMapper can be fixed with 
generating platform specific images. The fonts are different on each 
platform. One interesting note - on linux the text is not centered correctly.

I believe that many of these problems do not have anything to do with the 
two line change I made, but rather existed already but were not ever 
tested. This is a case of where simple code coverage is not enough - we are 
testing vtkImageMapper and vtkPolyDataMapper2D extensively, but not in 3D 
windows.

Lisa



At 10:20 AM 3/23/00, Ken Martin wrote:
>For Y'alls Info,
>
>Looks like some Transform or LinearTransfrom code is leaking memory.
>
>Lisa, TestFGBG is failing plus a couple others.
>
>TestFGBGText is failing (need an IRIX image?)
>
>scalarConn is failing, I think that one is mine and I'll look into it.
>
>- Ken