Use mailmap in github stats

[jenshnielsen]

In the 1.4.0 release notes I manages to get my name in using 3 different variations http://matplotlib.org/users/github_stats.html. I thought that looked rather silly so this is an attempt to deduplicate this.

IPython deduplicates various spellings by using the .mailmap file so I have just copied this to our repository and added the relevant variations of my name to the mailmap

I think that at least @pelson and @pwuertz also appear more than once in the log so we might want to add them. This could be back ported to the 1.4.0 documentation branch.

[jenshnielsen]

If anyone thinks this makes sense, I am happy to go over the log and try to locate duplicates

[jenshnielsen]

Apparently the --use-mailmap is redundant when using a format string %aN already does that while %an does not. --use-mailmap is only relevant without the format string

[tacaswell]

👍

This should be back-ported to 1.4.0-doc.

[WeatherGod]

I see "Per" and "Per Parker" entries, which is likely to be a duplicate.
There is also a bunch of other first name only entries that might be dups.
We also have a bunch of username entries like "vagrant".

On Wed, Aug 27, 2014 at 1:20 PM, Thomas A Caswell notifications@github.com
wrote:

[image: 👍]

This should be back-ported to 1.4.0-doc.

—
Reply to this email directly or view it on GitHub
#3422 (comment)
.

[jenshnielsen]

Yes there are a number of additional duplicates. I will do a first attempt at this to see how many I can find.

[jenshnielsen]

I have added the names that I found from a first pass of the list.
I have mainly merged people with the same mail addresses but different names and
a few others that I am almost sure are correct. There are probably others. But when both the name and the mail address is different it is difficult to tell.

I have not merged anyone that only differs in mail address since I don't think we use this anywhere.
I.e. @tacaswell is registered with 4 different addresses

This reduces the total number of authors from 295 to 278

[jenshnielsen]

I don't plan to do any more on this. I think that it is ready to merge.

[tacaswell]

cherry-picked to v1.4.0-doc as 0eaedc8