[Reproducible-builds] [Popcon-developers] Bug#833695: popularity-contest: http://popcon.debian.org/all-popcon-results.txt.gz containso

[Holger Levsen]

reopen 833695
severity 833695 important
clone 833695 -1
reassign -1 python-popcon
severity -1 normal
retitle -1 python-popcon must not crash on invalid data
thanks

On Mon, Aug 08, 2016 at 11:34:43AM +0200, Chris Lamb wrote:
> > Please reassign to python-popcon.
> 
> Well, there are perhaps two bugs here, so simply reassigning won't be
> sufficient:
> 
> Firstly, all-popcon-results.txt.gz contains invalid data. I agree that
> the client should not be removed from testing but I could not find a
> suitable pseudo-package for the popcon "web" service.
>
> (The code that generates this file appears to be in the popularity-contest
> source package, hence why I filed it here)

Bill, following this I have reopened the bug and downgraded the
severity. 
 
> Where should this issue be raised? I don't care about the severity, more
> that it gets fixed. How can I help? :)

I also disagree with merely tracking an issue which breaks Debian
infrastrcuture on a mailinglist instead of a proper bug tracker.

Bill, if you really insist #833695 should not be assigned at
popularity-contest then please reassign it to qa.debian.org
pseudo-package.

> Secondly, src:python-popcon assumes valid data. Perhaps. But it would
> still have to raise an error if it could not parse the file so there's
> not much it could reasonably do. It *could* print a nicer exception so
> its easier to track down the problem, but that's purely aesthetic..

agreed, but that still would be nice, thus the cloned bug.


-- 
cheers,
	Holger
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 811 bytes
Desc: Digital signature
URL: <http://lists.alioth.debian.org/pipermail/reproducible-builds/attachments/20160808/ed23a3b7/attachment.sig>