--- Martin Bähr <[EMAIL PROTECTED]>
skrev:
> On Tue, Oct 03, 2006 at 11:25:06PM +0200, Axel
> Liljencrantz wrote:
> > One note here: The way I read the part of the
> standard
> > you quote, it does not make any mention either of
> the
> > use of either the '.d' or the 'rc' suffix
>
> right, that was my personal opinion in tht matter.
>
> > _My_ interpretation is that both are allowed
>
> i agree, sorry, should have made that more clear.
>
> > , though I have been told that the '.d' suffix
> > is a recent invention and should is frowned upon
> as
> > non-Unixy.
>
> wierd. although i don't know anything but unix so i
> can't verify this,
> but i doubt that this is true, given the historical
> existance of
> /etc/init.d and /etc/rc{0,1,s,3,4,5,6,}.d
True. I'm not an expert on Unix history either, so I
don't know how traditional the '.d' suffix is.
>
> > Unfortunatley, I think that it's a bit late for
> > removing the '.d', I expect there are quite a few
> > users that have their own functions and
> completions in
> > ~/.fish.d.
>
> yes, i'll buy that argument :-)
>
> > Even worse, there are definitely many users with a
> ~/.fish file, so a
> > new directory name would cause a clash.
>
> true, the upgrade might be a bit tricky, but i don't
> think it is
> impossible. fish could detect the situation at
> startup and then warn the
> user, and ask him/her to move things around (or even
> offer to do it
> (just don't do it without asking))
It's possible. Fish already has done a major
move-around of files where all the scripts that where
considered part of the base system where moved to
/usr/share/fish. But I strongly dislike moving the
users files. It's bad for your health.
>
> > The current naming is partially there for
> historical reasons.
>
> i thought (and i hope) fish is still in the stage of
> you being willing
> to break compatibility to make things nice and
> consistant.
> so i would hope that you would consider ignoring
> historical reasons.
Absolutely, but there is still _some_ resistance
against breaking compatibility. I'm not yet convinced
that the reasons given are enough motivation for a
break. I'm not even convinced yet that the layout you
propose is better than the one I propose.
>
> > This arrangement is modelled after the bash layout
> > using /etc/profile and /etc/profile.d/.
>
> note that /etc/profile is not a bash but an sh
> configfile (and is used
> by bash because it is sh compatible)
Yes, I know, but I always assumed that the profile.d
part was a bashism which is why I wrote it the way I
did. Haven't really investigated this, so my original
post may have been wrong or misleading.
>
> > I feel that a ~/.fish configuration file makes
> much more sense than a
> > ~/.fish.d/fish file, it's too much to write for a
> file that you
> > occasionally want to edit
>
> i disagree. i find it much more irritating to have
> config files in
> different directories. if they were all in one, i
> could cd to there, and
> then have all fish config files at my disposal for
> convenient access,
> while otherwise i'd have to think first about which
> file i need to work
> with before deciding in which directory i should be.
I guess we work differently, then. I usually just do
an 'echo BLAH BLAH BLAH >>~/.fish', to append
something to the config file, or launch it inoan
editor for a more complex addition. Both of these
would require more key strokes with a new filename.
Different usage patterns, but both are reasonable, I
guess. It's hard for me to see a solution that fits
both patterns well.
>
> also the other advantages of being able to have one
> item to copy around,
> backup, remove, etc far outweigh the disadvantage of
> having a slightly
> longer path for the main config file.
Agreed, this is an advantage.
>
> > As a sidenote, I'll mention that emacs uses
> ~/.emacs
> > as a user accessible configuration file, and uses
> > ~/.emacs.d for 'strange' files. And everybody here
> > uses emacs and trusts the judgement of the FSF in
> all
> > things, right? ;-)
>
> no.
> i use vi (and i even switched from emacs ;-)
> seriously, that argument is flawed in as much as
> emacs is so old that
> ~/.emacs isa fixpoint on a unix system. and likely,
> ~/.emacs.d is still
> optional. (/etc/profile.d is optional too)
Yeah, the 'everybody loves emacs and trusts fsf' bit
was meant as a joke, sorry about that. Most fish users
seem to use vi, and there is a lot of shouting between
the FSF and the kernel devs right now, so I couldn't
help myself. :-/
>
> .fish.d is not optional so the arguments for other
> .d dirs do not really apply.
The .emacs.d directory is automatically created when
you enable certain emacs features. profile.d is
created automatically by a basic install of Fedora,
and very probably several other Linuxes (Linuces?
Linuses?). But None of that matters, because 'other
people do it' is an exteremly weak argument for doing
something stupid. The real question is if a
.fish+.fish.d directory layout is stupid.
>
> greetings, martin.
> --
> cooperative communication with sTeam -
> caudium, pike, roxen and unix
> offering: programming, training and administration
> - anywhere in the world
> --
> pike programmer travelling and working in europe
> open-steam.org
> unix system- bahai.or.at
> iaeste.(tuwien.ac|or).at
> administrator (caudium|gotpike).org
> is.schon.org
> Martin Bähr http://www.iaeste.or.at/~mbaehr/
>
--
Axel
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users
