Bug#880411: sqldeveloper-package version forking
On 19 November 2017 at 20:51, Adam Borowski <kilobyte@angband.pl> wrote:
>On Sun, Nov 19, 2017 at 04:55:42AM +0000, Lazarus Long wrote:
>> The fact is that the NMU became a fork of an older version, that is out of,
>> and irrelevant for, my timeline.
>
> Well, the history not having a record of a previous upload will confuse bug
> version tracking. I think that, barring a manual intervention of setting a
> "fixed" version in the bug that's a part of its current version history,
> the bug will never go away.
>
> But I'm not sure what will happen.
Yes, both the NMU and proposed-updates versions are forks, both of
which are correctly recorded as Closes [868673]. As a result, if you
click on the version graph and then "Don't ignore boring", both
branches end in green and are also listed as "Fixed in versions".
As for what will happen with v0.4.4, it would be nice if all the
intermediary changelog entries were processed, but by the looks of
[accepted], only the specifically uploaded one is. 20/20 hindsight
therefore says v0.4.4 changelog should have included something like
the following (split out according to this [guidance]):
* Incorporates unreleased fixes now uploaded to unstable:
- JVM path (Closes: #693798)
- build error (Closes: #868673)
- Included debhelper on the build dependency list (LP: #588458)
- Replaced dh options with overrides (LP: #998258)
[868673]: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=868673
[accepted]: https://tracker.debian.org/news/887940
[guidance]: https://lists.debian.org/debian-devel-announce/2003/06/msg00016.html
I documented this partly for my own learning, but Lazarus please can
you send the relevant metadata emails, or let me know you don't have
time to? I believe this is all you need to fix the version graph (for
each bug):
"Control: fixed -1 0.4.4" (at top of email to 868673-done)
https://www.debian.org/Bugs/server-control#fixed
--
Phil Morrell (emorrp1)
Reply to: