Hi All,

There have been quite a few major changes to the OpenSceneGraph over the
last six months, all these changes are in preparation for 2.0.  I haven't
really discussed at a high level where I'm going with all these changes, so
with this post I'm hoping to flag up what all these disparate change are all
about.   Principally about making the OpenSceneGraph simpler to use and
maintain, but also behind the scenes there are also changes that will make
the OpenSceneGraph more scalable and powerful.  Over the weekend I reflected
on the convergence towards the next stable release and came up with what
could possible be the rallying call for the effort towards 2.0 - the best I
could up with is "Simpler yet more Powerful".

Under the Simpler to use and maintain front the changes we have are;

  1) The QuickStartGuide book :-)

  2) Unified build system - Cmake

  3) Reduction of non optional and non standard external dependencies from
Three to Zero!

      a)  OpenThreads integrated into core OpenSceneGraph build
      b) osgProducer+Producer replaced by osgViewer
      c) GDAL dependencies removed by moving database building function out
to VirtualPlanetBuilder

 4) More coherent and flexible high level viewer classes - osgViewer with
its SimpleViewer, Viewer and CompositeViewer

 5) Better support for 3rd party windowing library and view integration
(only part done on this one :-)

 6) Use of SVN replacing CVS.

 7) Weekly developer releases

On the "yet more Powerful" front we have:

 8) osgShadow library (still in development)

 9) osgManipulator library

 10) osgTerrain - new height field rendering framework (still in
development)

 11) osgViewer library with its support for new threading models and more
complex viewer configurations

 12) Support for distortion correction as a configurable part of the viewer
(still in development)

 13) New intersection routines that support line segment, polytope and
plane interesection as well as high
      level ElevationSlice, HeightAboveTerrain and LineOfSight classes in
osgSim which support querries
      against paged databases.
--

We still have a way to go before 2.0, the new build system needs to be
thoroughly tested and optimized, the new osgViewer library is not yet
feature complete, as are the osgShadow and osgTerrain.  We also need to test
things thoroughly and really ground out as many bugs and oddities as we
can.  To this end I'd encourage the community to move to start using the SVN
version or the weekly dev releases of the OpenSceneGraph and don't feel
afraid of pitching in a coding up solutions to problems you see. My
intention is to keep the SVN versions and the 1.9.x dev series as stable as
possible.

As we get nearer to 2.0 I will start upping the dev release rate, using dev
releases instead of having 2.0 release candidates.  This way the processes
that are in action today with the dev releases will be identical to that of
2.0, the only difference is that I'll be eventually by setting the version
number to 2.0 once one of the 1.9.x series looks solid enough - in effect
one of 1.9.x will be 2.0.  My intention is to keep the SVN versions and the
1.9.x dev series as stable as possible, while there might be an occasional
build or run-time hiccup the quality bar needs to be kept high, and keep on
getting higher as we converge to 2.0.

I have added a TODO list item at the bottom of the LatestDevelopments page,
please feel free to add critical items that are missing such as bugs.


http://www.openscenegraph.com/osgwiki/pmwiki.php/Documentation/LatestDevelopments

My hope is that 1.9.3 is already more stable than 1.2, and is certainly
already simpler to get to grips with thanks removal of the external
dependencies and the QuickStartGuide.  I do wonder whether we should annex
the 1.2 stable release for new users, and point them straight to the
1.9.xseries.  Thoughts? Experiences?

To sum up, the systems and process that will take us to 2.0 are now in
place, the guiding princpal for this wok has been and should remain "Simpler
yet more Powerful".  Convergence to 2.0 also needs the community to rally
around this effort, but rather than a all out effort, I see the drive
to 2.0as being an every day part of the project and something that we
can all play
a role in.

Cheers,
Robert.
_______________________________________________
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Reply via email to