[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: lynx-dev Re: [-dev.12] experimental text entry fields patch (updated
From: |
David Combs |
Subject: |
Re: lynx-dev Re: [-dev.12] experimental text entry fields patch (updated) |
Date: |
Mon, 11 Jan 1999 14:40:00 -0800 (PST) |
> From address@hidden Mon Jan 11 06:48:57 1999
> From: address@hidden
> Subject: Re: lynx-dev Re: [-dev.12] experimental text entry fields patch
> (updated)
>
> In a recent note, Kim DeVaughn said:
>
> > Date: Sun, 10 Jan 1999 16:35:07 -0800
> >
> > can't just do (an equivalent of) "rm -f some-dir" and have it and all
> > its contents blown away ... you still need to unlink() the individual
> > files in it before rmdir()'ing the directory.
> >
> That's what "rm -rf some-dir" does. Don't know about portability.
> ^
> ^^^
It's so damned dangerous, doing that. A key-bounce, eg between
a dot and a name, is just horrible.
What I have learned (the hard way!) to do is to create
a triple-deep directory todelete/todelete/todelete, then
"mv" all to-be-deleted files to that directory,
then cd to there (easy: cd !$), l (alias for ls -AFs),
be sure, then do a "sh" (to get away from adding
to a history list for csh),
do a ls -lAts, AND ALSO a pwd,
and ONLY THEN the rm *, etc.
If I screw up, I am protected by two OTHER layers of
dummy directories.
Of course, I've aliased rm to rm -i.
David (who once did the fastest-ever control-c <repeated>
after noticing a boo boo on my prior method!)
Oh, I also, from time to time, do a "du -a / >! allunixdua.out",
which helped then to find out how far it had gotten.