begin Peter Jay Salzman quotation: > 1. when you go into a cvs directory and issue a 'cvs update', how does > cvs "know" where to pull data from? is the remote cvs server info stored > in some file in the local cvs directory? or do i need to reset CVSROOT > every time i want to update my copy of a particular cvs?
Ever notice the 'CVS' directories in the directories you checked out? These contain files such as 'Repository' and 'Root' which keep the state info you mentioned (CVSROOT and which module it's a member of). I've even moved a repository to a new server and done a recursive edit of these files to reflect the server name change, and it's come off without a hitch. So no re-export of CVSROOT is necessary if you're cwd contains a 'CVS' subdir. > 2. in the lines: > > cvs server: Updating camlibs/agfa-cl18 > cvs server: Updating camlibs/barbie > cvs server: Updating camlibs/canon > P camlibs/canon/canon.c > P camlibs/canon/psa50.c > cvs server: Updating camlibs/canon/ps350 > U libgphoto2/libgphoto2.pc.in > cvs server: warning: libgphoto2/libgphoto-2.0.pc.in is not (any longer) > pertinent > > what do the P and U prefixes mean? should "non-pertinant" files be > deleted by hand? IIRC, P means patched, U means updated (didn't exist prior to update). > 3. after issuing 'cvs update' for a cvs that uses autoconf, is it necessary > (or is it good practise) to remake the make files by running configure > again? I'd recommend it. You never know what might have changed in the configure.in file... -- G
