HI Andrew, On Tue, Jun 9, 2009 at 6:52 PM, Andrew Cunningham<[email protected]> wrote: > Call it what you like (bug, not optimal behavior, I don't care) but I would > expect that in a code snippet
Well I do care. Your first post was impersonal and un-informative - absolutely no conclusion could be made without asking further questions. Please remember the rest of the community are not automaton but people, and we aren't mind readers. So if you want to get on you don't bark out your demands, but you explain the issue you are seeing. Please go back to your original post and score it out of 10 in terms in information and politeness. For this particular issue the fact that is resets to packing to 1 isn't something I would categorizes as bug - the code works perfectly, it doesn't crash, it doesn't overwrite over the end of the line, and it never advertised that it would honor an independently set parameter and was never designed this way. Adding the later would a new feature and behavior of this particular method. The change of behavior is not something that should be troublesome for current end users and as I mentioned above the packing hasn't been advertised as being managed in any particular way - it's a black box, then changing it's behavior shouldn't effect end users as they shouldn't be making any assumptions right now. It's worth note that this method has been part of OSG for many years without anyone taking issue so I presume most users don't require tight control over packing. Given the two options of amending the readPixels API to add a packing paramert vs just reusing the _packing parameters in the allocateImage() call in readPixels(), personally I'd suggest the former - the API change to keep the allocateImage, setImage and readPixels methods all consistent, one could provide a default value of 1 to make sure the code worked as before. Feel free to make addition of the packing parameter to readPixels and send in the changes to osg-submission as whole files. Cheers, Robert. _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
