Why is depthfirst option of pm3d done in batches?

[sfeam (E. Merritt) <eam...@gm...>]

Petr:

Could you take a look at bug report #3004793 on SourceForge?

https://sourceforge.net/tracker/?func=detail&atid=102055&aid=3004793&group_id=2055

It seems from the code that pm3d deliberately chooses not to merge
multiple pm3d surfaces using "depthfirst" unless they are adjacent
in the splot command line. That is, an intervening non-pm3d object
causes the quadrangles making up the pm3d surfaces _not_ be sorted together.
There may be a good reason for this, but I can't think of what it is.
Do you recall?

	Ethan