Item5247: parse and aggregate svn log files for release information.

see TWiki420SvnLog

-- TWiki:Main/SvenDowideit - 14 Jan 2008

I am concerned here

I have spent 1000s of hours testing new features and old features and have opened quite many bug reports where I spent 8-20 hours analysing the problem. And still it seems I have a minor role in the development of the release.

We cannot base the credits on SVN checkins. One feature can be a well tested feature with one checkin. Or it can be done with 100s of small incremental checkins.

It is not a fair way to evaluate contributions. I should be on the top 5 getting credit for 4.2.0 code with all the time I have spent testing.

-- TWiki:Main.KennethLavrsen - 14 Jan 2008

When I wrote above I had seen an earlier version of TWikiHistory.

Looks better now.

Maybe we also need to count the bug reporters and list the top core/default plugin bug reporters. Testing is as important as coding. And there are a few important contributors that have reported quite many bugs and for sure have spent many hours.

Also make sure only to count since release 4.1.2. We should not credit accumulated because then new contributors drown.

-- TWiki:Main.KennethLavrsen - 14 Jan 2008

I agree that we can't just use svn commits, it is just one metric. but if we measure no metrics, we're just pissing in the wind. I'd like to get to a point where we have several automatic metrics, and can make a knwon algorythm to combine them.

I also think there is still something not right with the HallOfFame svn stats - so I thought I'd play with making a more detailed alternative.

-- TWiki:Main.SvenDowideit - 14 Jan 2008

closing as unloved.

Sven