Re: Bug#640195: Change the bug 640195 severity to normal
Dear Anton,
Here is a summary of the status of paraview from my point of view.
1. I can no longer build vtkedge. This has been reported as 640488
(severity serious) by doko. The bug should be extremely clear, as you
are now working on paraview package. It is also clear that the issue
is in the paraview package itself. Which explain the reason of bug
640195. Why did I prefer two bugs instead of one ? In the past I had
numerous reports (with an 's') against gdcm or vtkedge because the
cmake import rules for vtk package or paraview package were bogus.
This is why I prefer to keep one (and only one) bug open with a clear
title to prevent any new duplicates, and set up a very nice and clear
block relation, this did work quite well in the past.
Correct me if I am wrong here, but I believe that 640488 is severity
serious and as such 640195 should be serious.
I do believe that all paraview package 3.10.1-* so far are in bad
shape. This should also be relatively clear from the numerous private
emails we had on the subject. paraview package is not only a client,
but it is also an extensible platform (as in the past on the 3.8
release).
2. I can no longer build gdcm/paraview extension. You should know that
as we have discussed about this, sorry to repeat myself. You may ask
why there is still no gdcm-paraview module in debian. The reason is
quite simple, gdcm has been stuck in unstable for months, because of
506992. And for other reasons, including my fault when I messed up the
source upload of paraview 3.8.0. Don't get me wrong but when you say:
...
> No reasons for this bug to be a serious [1] and not to let
> testing-users to get 3.10-version soon.
...
I see it from my prospective, as trying to get a half-baked paraview
in testing, which means this will get my job on package such as gdcm
or vtkedge even harder as paraview has entered testing.
3. Again when I quote your message:
...
> Please, discuss on the list before changing the severity.
...
Could you please confirm, this is the first time you hear about
640488, could you also please confirm you do not see any relation in
between 640488 and paraview package. Did I not mention this type of
issue will arise. Did I not communicate on public mailing lists [1],
[4] ?
4. You sent an email on Fri, 9 Sep 2011 20:34:54 +0200 [2], asking the
reason why the bug was set to important (see above if this is still
not clear). Then on Sat, 10 Sep 2011 13:23:51 +0200 [3] you decided I
did not need to answer, and therefore you decided to change it back to
normal. Did you see *any* email from me lately stating that I was
working pretty hard on paraview to get it back into shape. I believe
you did see those emails as can be seen from the git-commit history of
paraview. I would really hope now, that you stop calling me a liar,
and acknowledge that I do communicate, early and often and in the very
mailing list you referred to in your statement. I would like also to
point out that you left me a little less than 17 hours to respond to
your inquiry, would you agree, this is very little time ?
Don't get me wrong on this, all I would like is have vtk & paraview in
much better shape as they are. But I cannot concentrate on working on
paraview or vtk if people from my own team are working against what I
believe are the goals of debian.
As per your request, I did express my opinion publicly (I did not know
you preferred this over private emails). Now, I would like to have
your opinion on paraview package. I did ask for your opinion in the
past [1], I am still looking for answer on the following issues:
1. paraview versus convenient copy of VTK. Since you are running to
get paraview in testing, you must have solved the very difficult issue
of duplicate vtk import rules in paraviewconfig.cmake file. Could you
please elaborate on that ? Could
2. Could you please also elaborate on the status of paraview-python
package ? When should users prefer this package over the one from vtk
? Can a user extend vtk independently of paraview ? How should one
deal with the the fact that paraview does not use VTK 5.8 ABI ?
(1) and (2) are the two main reasons which makes me thing paraview
3.10-* are not currently ready for testing.
Thanks for your reading,
-M
[1] http://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg938336.html
[2] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=640195#65
[3] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=640195#70
[4] http://lists.debian.org/debian-science/2011/08/msg00055.html
On Sat, Sep 10, 2011 at 1:23 PM, Anton Gladky <gladky.anton@gmail.com> wrote:
> severity 640195 normal
> thanks
>
> No reasons for this bug to be a serious [1] and not to let
> testing-users to get 3.10-version soon.
> Please, discuss on the list before changing the severity.
>
> [1] http://www.debian.org/Bugs/Developer#severities
> Anton
>
>
>
> --
> debian-science-maintainers mailing list
> debian-science-maintainers@lists.alioth.debian.org
> http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/debian-science-maintainers
>
--
Mathieu
Reply to: