2f6b50d146016205047834165a81810d7849efce removed contributors

[ghost]

@schacon created this repository, https://github.com/progit/progit2, as a separate one from https://github.com/progit/progit (i.e. manually copied the contents across, which is a Git antipattern IMO).

As such, people who contributed to https://github.com/progit/progit, and whose contributions were manually copied to https://github.com/progit/progit2 as part of the latter's initial commit, do not necessarily show up in git shortlog when it is run on this repository. It was presumably for this reason that @schacon credited them manually by adding book/contributors.asc in 7270a692ab7ef88db5a3d7955bfe12d4e5f95250. (I expect that file was generated by running git shortlog on https://github.com/progit/progit .)

However, 2f6b50d146016205047834165a81810d7849efce deleted book/contributors.asc, thus erasing that credit from subsequent versions of Pro Git 2 (aside from people who were also present in the Pro Git 2 shortlog).

I see at least three possible ways to fix this:

1. Restore book/contributors.asc and add a (post-receive?) hook to expand it each time a new contributor's commit is merged.
2. Restore book/contributors.asc and merge its contents with the git shortlog output automatically during the build process.
3. Instead of populating the contributors section of the book by running git shortlog solely on https://github.com/progit/progit2, the build process should also run it on https://github.com/progit/progit and merge the output from the two invocations.

[ben]

I want to pop in to clarify a bit of history. I agree that it's an anti-pattern to take a snapshot of one git repo and use that as the initial commit of a separate repo, but it was done for good reasons.

The people who contributed to the 1st edition never licensed their changes to Scott for the purposes of making a print edition. When the time came to start working on 2e (with the intent of selling a dead-tree version), we couldn't in good conscience use all of their work without permission, so we had to start from a snapshot where Scott owned all of the IP. We still wanted to thank them in the print edition, and that's how contributors.asc came to be.

And yes, when the contributors list became generated at build time, we lost the list of contributors to the 1st edition. However, at this point the second edition consists of original work by myself and Scott, and work by the authors that appear in the shortlog of this repo. So that change seemed like a move to a more accurate representation at the time.

I'd accept a PR that includes both lists, whether merged together or separate. I'm still grateful to the people who helped out with 1e, and it'd be nice to give them credit in 2e.

[ghost]

I agree that it's an anti-pattern to take a snapshot of one git repo and use that as the initial commit of a separate repo, but it was done for good reasons. ... When the time came to start working on 2e (with the intent of selling a dead-tree version), we couldn't in good conscience use all of their work without permission, so we had to start from a snapshot where Scott owned all of the IP.

Ah, that's useful to know :)