[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: RFC: some new deb package flag: "upgrade-conflicts"

On Thu, Oct 03, 2002 at 09:54:17AM -0500, Steve Langasek wrote:
> > > -- if you
> > > think leaving this stuff around doesn't make maintenance more difficult,
> > > I have some scripts to show you... :)
> > Would you mind?
> Ok.


> So, the file looks a bit cluttered.  If you can suggest a way to make a
> shell script do a lot of stuff without looking cluttered (which I believe
> is an inherent property of shell scripts), I'm all ears. :)  

One possibility, since it's your postinst, is to move it to an entirely
separate file, and just have something like:

	if dpkg --compare-versions "$2" lt "2.2.3a-6"; then
		# upgrading from pre-woody version, so files may be in all
		# sorts of weird places. there's lots of crufty junk here


> Otherwise, I
> view supporting potato->sarge upgrades as more cumbersome than not
> supporting them.  Maybe not prohibitively so -- if Policy says we have to
> do it, it can be done -- but given the choice, I think there are more
> productive things that I can spend my development time on, especially
> when it's so easy to introduce a bug in the script that can render this
> little-tested upgrade path completely useless.

Yeah, well, we already have enough problems getting regular upgrades
well tested, so that's unlikely at the moment. Oh well. It just seems
a shame to cull working code for something users actually want.

> # Generate configuration file if it does not exist, using default values.
> [ -r "${INITCONFFILE}" ] || {
> 	echo Generating ${INITCONFFILE}... >&2

(Oh dear)


Anthony Towns <aj@humbug.org.au> <http://azure.humbug.org.au/~aj/>
I don't speak for anyone save myself. GPG signed mail preferred.

 ``If you don't do it now, you'll be one year older when you do.''

Reply to: