package: qa.debian.org

Hi,

http://packages.qa.debian.org/m/munin.html says "A new upstream version is
available: 2.1.2, you should consider packaging it." while it also says that
experimental has 2.1.2-1, that's annoying.

Thanks for your work on the awesome PTS!


cheers,
	Holger

<pabs> h01ger: should it be changed to 'please move version X from
experimental to unstable'?
<pabs> I've always considered it a reminder to do that
* h01ger would prefer it to be quiet
<h01ger> that version is perfectly fine in experimental
* h01ger dislikes automatic todo entries where one cannot or doesnt want to do
something about it
<h01ger> if you think munin 2.1.2-1 belongs to unstable, file a wishlist bug.
<h01ger> (you as in $you)
<h01ger> like it is, the PTS conveys the message that the munin maintainers
are slacking, while we are not. (or maybe we are, but not in regards to
2.1.2-1)
<pabs> h01ger: is munin 2.1.2 a beta or dev version or something? what is
holding it out of sid?
<pabs> if so I think the right answer is for the watch file in sid to not
match dev versions
<h01ger> pabs, how is that relevant? (its a dev version)
<h01ger> http://sf.net/munin/ munin-([\d+\.]+)\.tar\.gz
<h01ger> is the content of the watch file (+version=3)
<h01ger> so currently (AIUI) it looks like i want to remove the watch file
entirely
<taffit> http://sf.net/munin/ munin-(2\.0[\d+\.]+)\.tar\.gz  would do
<h01ger> pfft
<pabs> http://sf.net/munin/ munin-(\d\.[02468]\.[\d+\.]+)\.tar\.gz
<h01ger> i rather remove it then maintain a useless file
<themill> I think we perhaps need to accept that all automated checks are
going to occasionally going to offer up false positives and we accept that
they are right 90% of the time and just learn to ignore them the other 10% of
the time.
<jcristau> that would just make the watch file less useful
* h01ger thinks there is too much guessing involved
<h01ger> 2.1.2-1 is packaged
<jcristau> themill: the problem is when people don't realize their checks are
wrong 50% of the time
<h01ger> so why claim its not
<themill> jcristau: 50%? really?
<jcristau> whatever
<jcristau> even 10% would be too much
<h01ger> i could live with the noise about "please move the package to
unstable" but claiming 2.1.2-1 is not packaged and we should do our work, is
plainly wrong
<taffit> indeed
<themill> h01ger: yeah, that check should use max-version not sid-version
<h01ger> (pabs, its a total guess that future stable releases will be 2.2 -
they could also become 3.0 or munin 13.10)

Holger Levsen <holger@layer-acht.org> writes:

Or provide a way for the package maintainer to say "this todo entry is
wrong, please go away" for those entries known to have false positives,
like the "newer version is available" one.

It's best to let computers do things they're good at rather than asking
humans to do it.  Computers are much better at ignoring things they've
been told to ignore than humans are.  (More implementation work, though.)

Hallo,

Ich habe dir eine Mail geschickt, aber keine Antwort von dir, warum?

Eddie