Re: Bisection results/Regression changesets useful?

[Vincent van Ravesteijn]

John McCabe-Dansted schreef:
I have a few questions.
1) Should I do a bisection on essentially every regression I find,
and/or are there specific bugs that you would like me to perform a
bisection on?
  
I don't know what to answer here. It's often easy to find out what 
change caused the bug, but sometimes there are bugs which are hard to 
find, then the bisection result might be very useful.

2) Does it help to include the committer of the changeset the caused
the regression, e.g. because they are most likely to understand the
changeset?
  
You can do it, but it won't change the world I guess. Some devs are no 
longer actively involved, sometimes a correct change only reveals 
another bug .. etc.

3) Do you already do bisections on bugs, if so what tools do you use?

  
Usually it is clear in which piece of the code the bug is, then "svn 
blame" can help you. Apart from that, our memory of what has been 
changed can help us.


Vincent