On Tue, 12 Dec 2000 [EMAIL PROTECTED] wrote: > Date: Tue, 12 Dec 2000 20:24:43 -0500 > From: [EMAIL PROTECTED] > To: [EMAIL PROTECTED] > Cc: [EMAIL PROTECTED], [EMAIL PROTECTED] > Subject: Re: Some updates to the sysvinit/initactions spec > Resent-Date: Wed, 13 Dec 2000 02:26:11 +0100 > Resent-From: [EMAIL PROTECTED] > > Date: Thu, 7 Dec 2000 00:18:41 -0500 > From: "Theodore Y. Ts'o" <[EMAIL PROTECTED]> > > Good point. The original place which I had suggested was > /etc/init.d/lsb-init-functions, but some folks objected to this on a > number of grounds (a) they didn't like it in /etc, and (b) > lsb-init-functions isn't an actual init.d script, but rather a fragment > of one. > > Does anyone care to suggest another location? > > No one has commented, so I'll offer a few suggestions: > > /lib/lsb/init-functions > /etc/init.d/lsb-init-functions > /etc/lsb-init-functions > > If no one comments, I'll choose /lib/lsb/init-functions. That seems to > make the most amount of sense. my first choice would be /etc/lsb/init-functions, this makes sense to me, and has a good taste to my mouth :), my second choice /lib/lsb/init-functions > > 2. killproc basename [signal] and pidofproc basename: > > Can we replace basename with "path" ? If we have the full path > to a program it is much easier and safer to identify the right > process. > > This seems reasonable to me, although it's a change from what the > SysVinit package does. Comments? > > I've heard no comments on this either. If someone thinks this is a bad > idea, speak now or forever hold your peace.... The point is that it goes against the tradition we saw in sysV. that is OK, if we are confident that sysadmins soon will become confortable with this change, otherway, we can stay in peace with the tradition.
anyway i agree with this change. Luigi Genoni
