Skip to Content.
Sympa Menu

sm-commit - Re: [SM-Commit] PERFORCE change 76944 by Eric Sandall for review

sm-commit AT lists.ibiblio.org

Subject: Source Mage code commit list

List archive

Chronological Thread  
  • From: Jeremy Blosser <jblosser AT sourcemage.org>
  • To: Eric Sandall <eric AT sandall.us>
  • Cc: duane_malcolm <d.malcolm AT auckland.ac.nz>, Unet <unet AT sourcemage.org>, Gareth Clay <gareth AT caffeinefuelled.co.uk>, vladimir_marek <vlmarek AT volny.cz>, Juuso Alasuutari <iuso AT sourcemage.org>, Pol Vinogradov <vin.public AT gmail.com>, SM-Commit Daemon <sm-commit AT lists.ibiblio.org>, Arjan Bouter <abouter AT gmail.com>, Andra?? ruskie Levstik <ruskie AT mages.ath.cx>, Andrew Stitt <a AT t.armory.com>, Ladislav Hagara <ladislav.hagara AT unob.cz>, Arwed von Merkatz <v.merkatz AT gmx.net>, David Kowis <dkowis+smgl-p4 AT shlrm.org>, Ethan Grammatikidis <eekee AT eekee.is-a-geek.org>, Maurizio Boriani <baux AT member.fsf.org>
  • Subject: Re: [SM-Commit] PERFORCE change 76944 by Eric Sandall for review
  • Date: Wed, 29 Mar 2006 12:26:26 -0600

On Mar 29, Eric Sandall [eric AT sandall.us] wrote:
> On Tue, 28 Mar 2006, Jeremy Blosser wrote:
> > On Mar 29, Perforce Review Daemon [p4review AT smee.org] wrote:
> >> Change 76944 by eric_sandall@sandalle-moby on 2006/03/29 06:36:20
> >> * INSTALL: Added to install default global mpd.conf
> >
> > We need to establish some kind of policy around this. We just had a long
> > bug discussion (http://bugs.sourcemage.org/show_bug.cgi?id=10486) that
> > ended up with "if upstream doesn't install it during 'make install', we
> > should ask before installing it and default to 'n'".
> >
> > However, there are some obvious differences between a sample config that
> > _changes_ default behavior in a user app and a config file for a daemon
> > that needs some kind of config to operate. Is that where we want to draw
> > the line? We should establish what it is so spells can adhere to it and
> > users can get a consistent experience from spells.
>
> mpd won't run without some /etc/mpd.conf or ~/.mpdconf and no example
> is installed at all,

The mpd upstream sources contain a doc/ dir that has the mpdconf.example
file and a COMMANDS file. These should both get added to DOCS in the spell
so they get installed in /usr/share/doc/mpd/. (COMMANDS says it's for
'client developers', but mpd's architectures means 'client developer' may
include 'anyone that wants to talk to the daemon via telnet', this is a
good thing.)

> so I had no idea where to start in making a ~/.mpdconf.

'man 1 mpd' spells it out ;-)

> So I'd go with the second rule: Install config files only if the package
> needs them to run or they explicitly set the defaults

IMO we should still ask, but I'm an idealist about leaving things alone vs.
upstream (I'd probably prefer to ask before changing default install paths
even, or at least use a global 'enforce FHS' option for it).

Anyway, what you said is probably a reasonable way to do it, can this get
noted in the guru handbook somewhere?

> and if the default install doesn't install it then it's an upstream
> bug.

Hrm, maybe. Some projects don't do this because they want to force people
to not accept defaults and actually read the docs. Not an entirely bad
practice IMO (and upstream having any reason to do what they do is one
reason to not just install them ourselves, also IMO).

Attachment: pgpR64_guIDGk.pgp
Description: PGP signature




Archive powered by MHonArc 2.6.24.

Top of Page