[vtk-developers] VTK library version numbers and verdict

[Philippe P. Pebay]

Hello David

I don't remember why we did that when we integrated Verdict into VTK. There might have been a reason, which I forgot. 
Some folks are using the standalone version of Verdict. I am forwarding your question to the verdict-dev list.

Thanks
P

David Cole wrote:
> Hey all,
> 
> 
> The VTK libraries (with the exception of verdict) all use the following 
> chunk of code from the main VTK CMakeLists.txt file to specify thier 
> VERSION and SOVERSION properties:
> 
>   # Append the library version information to the library target
>   # properties.  A parent project may set its own properties and/or may
>   # block this.
>   IF(NOT VTK_NO_LIBRARY_VERSION)
>     SET(VTK_LIBRARY_PROPERTIES ${VTK_LIBRARY_PROPERTIES}
>       VERSION "${VTK_VERSION}"
>       SOVERSION "${VTK_MAJOR_VERSION}.${VTK_MINOR_VERSION}"
>       )
>   ENDIF(NOT VTK_NO_LIBRARY_VERSION)
> 
> 
> The verdict library instead uses this chunk of code from 
> VTK/Utilities/CMakeLists.txt:
> 
>   SET( VERDICT_LIBRARY_PROPERTIES VERSION "vtk${VTK_VERSION}" )
> 
> 
> I (think I) understand the "vtk" is intended to distinguish it from when 
> it is built as a standalone library and not as part of VTK...
> But the net effect of this discrepancy is that the install tree contains 
> many libraries with three files each like this...:
>   libvtktiff.so -> libvtktiff.so.5.4
>   libvtktiff.so.5.4 -> libvtktiff.so.5.4.0
>   libvtktiff.so.5.4.0
> 
> ...and then the verdict library with only two files that do not follow 
> the same convention:
>   libvtkverdict.so -> libvtkverdict.so.vtk5.4.0
>   libvtkverdict.so.vtk5.4.0
> 
> 
> So here's the question:
> Isn't the name "libvtkverdict" sufficient to distinguish the library 
> from its standalone non-VTK library...? Can I make libvtkverdict 
> consistent with the other VTK libraries by changing the code to this: ??
> 
>   SET( VERDICT_LIBRARY_PROPERTIES
>     VERSION "${VTK_VERSION}"
>     SOVERSION "${VTK_MAJOR_VERSION}.${VTK_MINOR_VERSION}"
>     )
> 
> Or is there some other reason that I am missing that it needs to be the 
> way it is?
> 
> 
> Thanks for any insight / feedback......
> 
> David Cole
> 


-- 
Philippe Pébay
Sandia National Laboratories