Oops, I meant to reply the other day. This is good.
--Mark On Tue, 12 Feb 2008, Nathan Bush wrote: > Date: Tue, 12 Feb 2008 17:05:47 -0800 > From: Nathan Bush <Nathan.Bush at Sun.COM> > To: Mark J. Nelson <Mark.J.Nelson at Sun.COM> > Cc: scm-migration-dev at opensolaris.org > Subject: Re: [scm-migration-dev] Please review # 421 > > Nathan Bush wrote: >> Mark J. Nelson wrote: >>> It's good enough, or if you felt like polishing a little bit more, you >>> could also typset OLDDIR/NEWDIR variables in build_old_new, because each >>> called function repeatedly uses the same constructs >>> >>> $WDIR/raw_files/old/$PDIR >>> $WDIR/raw_files/new/$DIR >> >> With variables set this way it would need to have $OLDDIR/$PF or >> $NEWDIR/$F to identify the file in the raw_files directory, but it >> would still need to have $PWS/$PDIR/$PF or $CWS/$DIR/$F to identify >> the file in the actual SCM directory. It's shorter, but it seems >> less readable to me -- needing to use a separate variable for the >> subdirectory in one place but not the other. >> >> However, it makes sense to typeset the original olddir/newdir in >> build_old_new, then the build_old_new_* variants can be mostly >> restored to the original versions, and a bit more code can be >> moved into build_old_new: >> >> http://cr.opensolaris.org/~nbush/scm-migration/421/webrev.v3/ > > Do you have any comments on the third version? I would like to push > this change soon. > > Thanks, > > --Nathan >
