I’ve been CVS-ing ROOT about once every three days starting from the 10.00 release. Today when I did, it broke in compile. This may not be serious if nobody is as scrupulous (or paranoid?) about the bleeding edge of HEP software. Anyway, I grabbed the complete CVS head and put it into another place to build and this time it didn’t break.
Here is the make output from my broken build (it didn’t get very far):
…
build/unix/depend.sh krb5auth/src/Krb5Auth.d “-pipe -Wall -W -Woverloaded-virtual -fPIC -Iinclude -D_REENTRANT” krb5auth/src/Krb5Auth.cxx > krb5auth/src/Krb5Auth.d
build/unix/depend.sh proofd/src/proofd.d “-pipe -Wall -W -Woverloaded-virtual -fPIC -Iinclude -D_REENTRANT” proofd/src/proofd.cxx > proofd/src/proofd.d
build/unix/depend.sh rootd/src/rootd.d “-pipe -Wall -W -Woverloaded-virtual -fPIC -Iinclude -D_REENTRANT” rootd/src/rootd.cxx > rootd/src/rootd.d
build/unix/depend.sh rpdutils/src/rpdutils.d “-pipe -Wall -W -Woverloaded-virtual -fPIC -Iinclude -D_REENTRANT” rpdutils/src/rpdutils.cxx > rpdutils/src/rpdutils.d
gmake: *** No rule to make target proof/inc/TPacketizer.h', needed by
proof/src/G__Proof.d’. Stop.
Again, no really big problem, at least not that I can see especially since I know so little about CVS (and build tools) and its (their) caveats. I’ll just use my new build from now on.
Justin