[2311] in Release_Engineering
suggestion
daemon@ATHENA.MIT.EDU (geer@crl.dec.com)
Fri May 11 09:27:11 1990
Date: Fri, 11 May 90 09:26:52 EDT
From: geer@crl.dec.com
To: rel-eng@ATHENA.MIT.EDU
i heard this suggestion from Jim McGinness of UEG:
when doing a full build, first do a
make -n all > make.commands
then perform the make by doing this
sh make.commands > make.log
the argument for why this is better is that the
file ``make.commands'' is a a deterministic,
reproducible, self-usable record of what was
done
seems like a good idea to me
--dan
