Can anyone who is more familiar with ive ReaderWriter comment on the following observations.
Setting the includeImageFileInIVEFile option when writing ive files causes very large files to be written even if the scene contains few textures. It appears that whenever an image object is referenced in the graph a copy of it is placed in the output stream along with a full copy of the image data. Am I correct in this assumption? Looking at the ive input code it appears that new image objects are created unconditionally without checking for duplication whenever they are encountered. It appears that this results in many copies of the same image in the graph. Am I talking complete nonsense here? I was just wondering whether it would be an easy change to make to get the code that reads the mage data from the stream to tag the data with the filename and put it in the cache. This data could then be used on subsequent reads. Comments anyone? I will have a go at this if it seems feasible, although a more elegant solution that stops the duplicate data getting into the file in the first place may be better. I suspect most of the other duplications that occur can be removed with the optimiser. I will not be offended if I am told the above is utter nonsense! Roger
_______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
