Source Mage code commit list

[Arwed von Merkatz <v.merkatz AT gmx.net>]

On Tue, Sep 05, 2006 at 11:04:47AM -0700, seth AT swoolley.homeip.net wrote:
> On Tue, Sep 05, 2006 at 08:00:38PM +0200, Arwed von Merkatz wrote:
> > On Tue, Sep 05, 2006 at 10:14:08AM +0200, Thomas Orgis wrote:
> > > Am Mon, 4 Sep 2006 20:14:23 +0200
> > > schrieb Jaka Kranjc <lynx AT mages.ath.cx>: 
> > > 
> > > > > + ?config_query ?CREATE_DEV_TREE ?"check for / create static /dev 
> > > > > tree" y
> > > > Shouldn't this query default to n, as most users do want udev?
> > > 
> > > I tried not to change the logic of the spell: Formerly, it was
> > > 
> > > if  [ "$DEVICES" != "devfs" ]; then
> > >    check_static_dev
> > > fi  &&
> > > 
> > > So both udev and static mode triggered the check_static_dev function, 
> > > while this function then asked for creation of the static tree, 
> > > defaulting to y.
> > > What I did was to move that question out of the function and out of 
> > > INSTALL (!!) into the configure stage, but not changing default 
> > > behaviour.
> > > I must admit that I feel safer when there are some device nodes as 
> > > fallback even when using udev. It doesn't hurt, normally, but it 
> > > shouldn't be necessary.
> > > 
> > > To change this default behaviour, that is a question that perhaps needs 
> > > a bit of discussion. When ppl agree, I'm not opposing.
> > 
> > It defaults to yes because there were some install isos that installed a
> > completely empty /dev directory and used devfs. You can't boot an udev
> > system with such a /dev, so that check was added.
> 
> When we do hacks like this, it's god to put a little comment in that 
> says why we're doing something out of the ordinary.  Other coders aren't 
> omniscient, and they can't read minds, either. ;)
 
True, if we all were omniscient, this list would be really superfluous
;)

-- 
Arwed v. Merkatz                              Source Mage GNU/Linux developer
                                                    http://www.sourcemage.org