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

Re: Release-critical Bugreport for March 3, 2000

In article <[🔎] 20000303174545.A16396@molehole.dyndns.org> you wrote:
> On 03-Mar-00, 12:26 (CST), Bdale Garbee <bdale@gag.com> wrote: 
>> > Package: ntp (debian/main)
>> > Maintainer: Bdale Garbee <bdale@gag.com>
>> >   59090  slink->potato upgrade (probably) delete config file
>> I'm unhappy about this, and soliciting inputs about whether having an xntp3
>> transition meta package is really worth bumping the epoch on the ntp v4 
>> sources?

> You're unhappy that I reported a bug?

No!  Not at all!  I'd even say your report was better quality than most that I
get, including as it did both an analysis of what the underlying problem was
and a proposed solution.  Most bug reports aren't nearly so useful!

I'm unhappy that the choices for solving this are all things I don't want to 
do...  splitting the xntp3 into a separate source package so that it can have 
its own version number (sigh), incrementing the epoch on the ntp source 
package (sigh), or tossing the xntp3 meta package (sigh).

The input I'm soliciting is which of these three about-equally-unpleasant-to-me
choices I should make.

> My first suggestion(which perhaps wasn't clear) was to
> simply make the version number of the xntp3 transition package higher
> than the last "real" xntp3 package. There's no requirement (that I'm
> aware of, anyway) that the xntp3 package version number match the ntp
> version number.

Since it's just a meta package, I made it be derived from the ntp source 
package.  It is therefore currently completely tied to the ntp version number.
If I had thought about the version number problem before I did this, I might
have made it a completely separate package... but I feel silly having a 
completely separate source package just to deliver a silly little transition
aiding meta package...


Reply to: