[GEM-dev] possible fix for the [separator] memory leak

I think I have found a fix for the separator memory leak. It looks like there is really no case where the number of texture coordinates held in the GemState is ever not 4. So basically I put a check to see if the number of coordinates differ between the cached state and the passed GemState in the postrender function. If they differ it does the whole new allocation and the leak ensues, but I have yet to encounter a situation where that triggers.

I might be missing something rather big and obvious, so let me know before I commit this.

cgc

---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.