[Scilab-Dev] http://bugzilla.scilab.org - Severity of bugs

[Sylvestre Ledru]

> This is VERY good news. Anything that can help in understanding and 
> setting priorities looks good to me. As already discussed with you, it 
> would really help if rules in the Scilab project would exist, be 
> written down, and be exposed publicly. This is a vast subject, on 
> which bugs priority is only one side, and that any free software 
> project has to deal with I think.
Yep, your are right in theory but each bug is particular. It will be
hard to do general rules.
Some of the bug are obvious. For example
Problem in the display of UTF-8 char in the Java Console => Blocker
Scipad debugger broken => Blocker
Dynamic link broken in the binary => Blocker
etc

Some others are not.


> Also, who decides of whether a bug will be major, minor, etc.? Rules 
> again.
The guy who submit the bug. If anyone disagrees, he can change it.
Anyway, it is quite subjective.

> > Consequently, please, take a few seconds to update this field. Pick up
> > the severity you estimate to be the right one.
> 
> So the OP decides? Perhaps as a first step only?
At least the first step, usually in other FOSS people have a good vision of the severity of a bug.

> 
> The latest Bugzilla also introduced a new field: Priority. How to set 
> it properly? What do the five priority levels mean?
I don't know.

> 
> > In the mean time, we are going to reopen all bug classified as "LATER"
> > and use the tag/keyword system:
> > http://bugzilla.scilab.org/describekeywords.cgi
> 
> Also a VERY good idea IMO. I was wondering for years what LATER means, 
> especially when associated with "RESOLVED".
Well, pour rendre à César ce qui est à César, it is your idea and we agree with it ;)

> Maybe a suggestion on the keywords: as an example, perhaps 
> Fix_me_for_next_Scilab would be more appropriate than 
> Fix_me_for_Scilab_5.x. The reason I see is keywords updating when you 
> release a version, and keywords outdating. Just my 2cts.
Huhu, funny keyword! But yep, you are right.

> Also, again, who decides whether a bug should be tagged with a given 
> keyword, for instance "Fix_me_for_Scilab_5.0-beta"? 
At least someone who has the knownledge of what is involved by this bug.
And you know that it is not because a guy (opteam or or) tagged a bug
that this bug will be fixed.

> What are the applicable rules in case of a disagreement between people on the 
> severity, priority, etc.?
On a case by case basis I would say.

> On the same subject, what about bugs tagged as "REMIND"? Will you also 
> reopen them?
Yep, forgot this one.

> And since we are here, do you have any opinion about voting for bugs? 
> Bugzilla has this capability and it would allow for better sensing 
> feelings of the community I think, although it's perhaps a second step 
> to consider when the rest above gets more mature.
Well, as you know, some bugs are still open because they are pretty hard
to fix or involve a huge work or redesign. Voting won't change that.
I don't really like this idea:
Voting for a bug could suggest that we know how to fix a bug but we
don't do it for any reasons. 
And moreover, if a well known bug has many votes and we don't fix it
because it is too hard or too risky, it is not a very good signal that
we will send.


Sylvestre