Map contributors to use canonical real names and email addresses #482

[sbesson]

What this PR does

Since 2003, this codebase has substantially grown thanks to the input of several tens of contributors. Over two decades, commits from the same author are associated with multiple variants of the author line. In order to effectively review and audit contributions, it is necessary to normalize these variants.

The following rules are used for the construction of the .mailmap file:

• for all contributors who were also employees of the University of Dundee of Glencoe Software, the @dundee.ac.uk or @glencoesoftware.com email address is used as the canonical email address
• for all contributors with an executed CLA, the email address sent via the CLA is used as the canonical address
• for all contributors, the name used in https://www.openmicroscopy.org/teams/ or https://www.openmicroscopy.org/contributors/ is used as the canonical real name.

The .mailmap is constructed according to the official documentation using one of the two forms:

Proper Name <proper@email.xx> <commit@email.xx>

for mapping different email addresses and

Proper Name <proper@email.xx> Commit Name <commit@email.xx>

for mapping different real names.

Testing this PR

The unique list of commit authors can be generated and reviewing using git shortlog -se

All lines of the mailmap should be reviewed to make sure the email addresses and the mapping are valid. There are still a few historical users who I could not map so the knowledge of early day OME contributors will be very appreciated. The configuration can then easily be re-used for all the projects which were extracted (omero-model, omero-common....)

[melissalinkert]

Some of these are a bit debatable, but hopefully this helps someone else's memory from ~2004-2008. Mostly pulled from https://lists.openmicroscopy.org.uk/pipermail/ome-devel/, https://www-legacy.openmicroscopy.org/site/about/who-ome, and https://www-legacy.openmicroscopy.org/site/about/project-history:

1 (no author) <(no author)@6f2cb1de-eb0d-0410-b157-e593188b5901>

Probably Andrea Falconi given surrounding commits.

4 Weblitz User <staff@glencoesoftware.com>

Probably Carlos Neves.

229 brain <brain@05709c45-44f0-0310-885b-81a1db45b4a6>

Brian Loranger

50 david <david@05709c45-44f0-0310-885b-81a1db45b4a6>

David Whitehurst <david@glencoesoftware.com>

66 dcreager <dcreager@6f2cb1de-eb0d-0410-b157-e593188b5901>

Douglas Creager <douglas.creager@comlab.ox.ac.uk>

802 hsh <hsh@6f2cb1de-eb0d-0410-b157-e593188b5901>

Harry Hochheiser <harryh@pitt.edu>

319 jeffm <jeffm@6f2cb1de-eb0d-0410-b157-e593188b5901>

Jeff Mellen (couldn't find an address, see https://www-legacy.openmicroscopy.org/site/about/project-history/peter)

25 sfrank <sfrank@05709c45-44f0-0310-885b-81a1db45b4a6>

Stefan Frank <s.frank@dkfz-heidelberg.de>

1 tmacur1 <tmacur1@6f2cb1de-eb0d-0410-b157-e593188b5901>

Tom Macura <tm289@cam.ac.uk>

1 uimage <uimage@6f2cb1de-eb0d-0410-b157-e593188b5901>

git show 5bfa9728f23bf321 suggests it might be Paula Forbes <pforbes@computing.dundee.ac.uk>, or someone else from https://www-legacy.openmicroscopy.org/site/about/project-history/catriona

[jburel]

A few authors e.g. 2 former GS employees are missing from .mailmap list so the approach taken is not very clear. for example Paul van Schayck <polleke@gmail.com> is listed in .mailmap list and listed using git shortlog -se but Sam Hart <sam@glencoesoftware.com> is not listed .mailmap list

[sbesson]

Yes, my primary goal was to normalize scenarios where the email address and/or the real name was varying. In both examples above, this is not really the case

(base) sbesson@Sebastiens-MacBook-Pro-2 openmicroscopy % git log | grep pollek | uniq
Author: Paul van Schayck <polleke@gmail.com>
(base) sbesson@Sebastiens-MacBook-Pro-2 openmicroscopy % git log | grep sam@ | uniq  
Author: Sam Hart <sam@glencoesoftware.com>

I have not really made my mind on a strict policy but I am concerned about the maintenance burden of listing every contributor in the mailmap including those using a canonical real name and email address consistently. I pushed 7368ef8 to remove the superflous entry from the mailmap.

[jburel]

What about Weblitz User <staff@glencoesoftware.com>?

[sbesson]

Sorry https://github.com/ome/openmicroscopy/pull/6349#discussion_r1239897722 made me realise I got things wrong in https://github.com/ome/openmicroscopy/pull/6349#issuecomment-1604332401. There are two variants of the real name using with this email address so I pushed 7368ef8827574024b3b674e6aa2b4ca571c306fd away.

What about Weblitz User staff@glencoesoftware.com?

This was based upon https://github.com/ome/openmicroscopy/pull/6349#issuecomment-1598007980

[jburel]

The approach taken here is clearer. This is good to merge unless others think differently