[Teammetrics-discuss] Git Statistics

Scott Howard showard314 at gmail.com
Mon Jun 27 15:55:51 UTC 2011


Two comments might be fixed with one idea:

On Mon, Jun 27, 2011 at 8:12 AM, Sukhbir Singh <sukhbir.in at gmail.com> wrote:
> 2. There is no way to get the logs without cloning the repository. So
> first we have to clone the repository, get the logs and then extract
> the required information.

On Mon, Jun 27, 2011 at 10:47 AM, Andreas Tille <tille at debian.org> wrote:
> On Mon, Jun 27, 2011 at 07:56:19PM +0530, Sukhbir Singh wrote:
>> Hi,
>>
>> > The problem we have with Git is that it would be necessary
>> > to touch configuration of the teammetrics if *any* new package
>> > gets added by a team member.
>>
>> Can you give me an example as to what you are hinting at? From what I
>> have seen, most teams have a single central repository, be it Git or
>> SVN.

The git repos are all on alioth.debian.org in /git for example, the
debian science team is:
/git/debian-science/packages/

maybe you can make a read-only alioth account that your scripts can
ssh into alioth to parse for projects and performing "git log" on the
repos as they exist in alioth.

> Now comes the important part. Remember I said that we should include
> the lines committed as a metric? Here is what I propose: for each
> given author, let us get the total number of insertions and total
> number of deletions. The difference of those will be used as a metric.
> So if someone is adding lines and deleting very few of them, this
> means he is contributing more, right? There is only one problem with
> this approach -- sometimes when making edits, we delete more than we
> add. Should we bother about this then? If we do, we can't have a
> definite metric, then we need to have the metric of insertion and
> deletions as distinct metrics. Let's decide this.

I think that there are many contributions that fix bugs where the
number of lines added and deleted are equal, but they are
contributions (e.g. adding a B-D, fixing broken rules files) I think
you should try multiple ways and see which makes sense in real data
too.

~Scott


~Scott



More information about the Teammetrics-discuss mailing list