[vtk-developers] RE: [new tree] what happened to vtkImplicitModeller :)

[Ken Martin]

Hello Folk,

Sorry about the dependency problems. I've been making the changes on
Linux with static builds and it doesn't care about resolution until
link time. I have moved the MergePoints(2D) back into Filtering. In a
momentary lapse of reason I thought they were filters and moved them
into Graphics. That should fix the IO issues, I checked in new
versions of IO and removed the depends on Graphics. Here is what I am
shooting for the depend tree

Common - depends on nothing
Filtering - depends on Common
IO Graphics Imaging - depends on Filtering and Common
Rendering - depends on Graphics and Imaging (and via them Filtering
and Common)
Hybrid, Parallel, Patented - depend on all the above

ImplicitModeller was moved into Imaging because it produces images
(structured points) If it requires a graphics filter to operate then
I'll move it into Hybrid. Here is the basic kit descriptions

Common - core data structures and system support files
Filtering - support classes useful for filtering but no actual filters
if possible
Imaging - any filters or sources that produce images (image spoints)
Graphics - any filters or source that produce (poly, ugrid, sgrid,
rgrid)
IO - all IO classes except those that require Rendering
Rendering - all rendering classes
Hybrid - classes that depend on multiple toolkits
Parallel - classes typically used for parallel VTK
Patented - any patented classes

Thanks
Ken

> >>
> >>Whilst you're at it, there are dependency problems with
> >>vtkStlReader
> >>vtkUgFacetReader
> >>vtkMarchingCubesReader
> >>
> >>need vtkMergePoints. My local cmakelists has them
> commented out, I didn't
> >>look to see where problem was (ie which lib was missing)
> >
> >yes, I know, there are also pb with the ImplicitModeler needing
> >vtkClipPolyData...
> >I'll see what I can do.
>