Bug#848859: Lowering the bug`s severity

Santiago Vila sanvila at unex.es
Tue Dec 27 08:50:14 UTC 2016


On Tue, Dec 27, 2016 at 08:20:54AM +0100, Anton Gladky wrote:
> severity 848859 minor
> tags 848859 +unreproducible
> thanks
> 
> Hi  Santiago,
> 
> thanks for your bugreport! I tried to reproduce the failing
> test, building getfem on barriere.d,o 5 times but in all
> cases it did not failed. Also the last upload of getfem was
> successfully built on all supported platforms.
> 
> Nevertheless I am not closing the bug, but lowering its
> severity for the case if the bug will appear again.

The bug is probably still there, just that you didn't really bother to
reproduce it. I clearly said this in my bug report:

"The failure rate is around 5%, so if you try to reproduce this,
please try a *lot* of times."

This means, if you are bad at math, that the error will happen once
every 20 tries. If you only built it 5 times, you did not really tried
to reproduce it.

And then you didn't bother to Cc:me on this message, which I had to
read from debian-bugs-rc at l.d.o.

For this reason, I consider this downgrade of yours unacceptable and a
bad joke.

I'm going to repeat my tests in unstable, just for the case that some
already-fixed build-dependency was to blame for this, but if I still
reproduce this I will restore the original severity, as packages *must*
build from source *without* failure, this is not an invention of mine,
it's Release Policy.

Thanks.



More information about the debian-science-maintainers mailing list