I have put a gpc snapshot on my GPC web page http://www.math.uni.wroc.pl/~hebisch/gpc
This is my (unofficial) "stable snapshot": I have included Delphi and Mac objects patch and collected fixes (most fixes appeared on the mailing list). On i386 Linux using gcc-3.4.4 as a backend it is supposed to build without problems and pass all the tests. On 64-bit targets there are still multiple test failures (19 on my Amd64).
I tried to be conservative: new features are all related to Delphi classes and Mac objects. Also, I postponed fixing some problems where immediate fix was too risky.
I did routine testing ot this snapshot and I am not aware of any problem not present in earlier gpc versions. However, it is up to you to decide how good this snaphot is.
Why this shapshot? The last official gpc release was 7 months ago. There were more then dozen of fixes posted on the mailing list. I would like to reduce resultiong confusion collecting all the fixes together. Since some fixes are to backend diffs and another one renames a file it somewhat inconvenient to have the fixes as a patch, so I created a new tarball.
More detailed locations: the tarball:
http://www.math.uni.wroc.pl/~hebisch/gpc/gpc-20051104.tar.bz2
Minimal diff (does not include generated files):
http://www.math.uni.wroc.pl/~hebisch/gpc/curr4.diff.gz
News file (includes older news too):
http://www.math.uni.wroc.pl/~hebisch/gpc/NEWS-20051104
List of fixed bugs:
http://www.math.uni.wroc.pl/~hebisch/gpc/Fixed-20051104
Enjoy