Greg Stein wrote: > > My point was that *anything* could change and create a source/binary > compatibility problem. And in all likelihood, something *will* change > that we rely upon. One more argument to one function would bring down > the house. That is all it takes.
Yes of course. My point was more about the various apr-ish additions that exist within SVN that might become part of apr 2.0, and with the whole "portability vs. helper stubs vs. utility" resolved with the big rollup / merger of aprutil, what is useful becomes the overriding simple decision, not 'should we add this, and where would it go' :) And totally agreed that svn 1.7 isn't the timeframe to focus on any shift to an entirely new [2.0-ish] flavor of apr.
