Fix the commit hashes in ch10.2

[Smaug123]

This fixes the inconsistency introduced in
4f55250 (#1351), where commit messages were
changed without correspondingly updating the commit hashes.

This is incomplete because I don't yet know how to change the pictures.

The appropriate configuration to use to generate these hashes is:

export GIT_AUTHOR_DATE="2009-05-22T18:09:34+00:00-0700" \
export GIT_AUTHOR_EMAIL="[email protected]" \
export GIT_AUTHOR_NAME="Scott Chacon" \
export GIT_COMMITTER_DATE="2009-05-22T18:09:34+00:00-0700" \
export GIT_COMMITTER_EMAIL="[email protected]" \
export GIT_COMMITTER_NAME="Scott Chacon"

[Smaug123]

Sigh - there's a yak to start shaving with #1158 …

[Smaug123]

In fact this has to percolate irritatingly through the rest of this chapter. Would it be better to revert the appropriate section of the original commit? It amounts to changing a couple of commit messages back to lowercase, and the diagrams are already inconsistent (they have the old lowercase commit messages anyway); it would be a lot easier.

[ben]

I have two feelings about your effort here.

The first is that hey, this is just a book, and it's only meant to be an example. It's not a unit test, the hashes don't necessarily need to match up exactly.

The second is that this is the chapter where we generate hashes, and if you're following along, you should get the same answers we did.

Do we make a promise anywhere in the chapter that the user's answers should match the ones in the book? If so, then by all means, continue along your current path and fix them all. If not, I'd honestly sleep okay at night knowing that the book isn't 100.0% perfect on this. Everything has bugs. 😁

[Smaug123]

The book actually says explicitly that you shouldn't expect to get the same hashes because you're doing it at a different time; but it doesn't mention any other reason one might get different hashes, and I did spend multiple hours trying to work out why my hashes were different.

However, to do the fix properly would be a lot of work. Would you object instead to some sort of line like "The exact commit hashes in this section are representative examples only; while in principle any Git commit object can be reproduced precisely, historical details of this book's construction may mean the printed output does not correspond to a valid set of commits."?

[ben]

[…] to do the fix properly would be a lot of work. Would you object instead […]

Welcome to the wonderful world of technical book writing. 😄 I would absolutely accept a small note warning people that they may run into discrepancies. Probably want to enclose that in a [NOTE] block.

[Smaug123]

Thanks; I'm closing this in favour of #1424.

[HonkingGoose]

Hi @Smaug123, sorry for making your life difficult because of me changing the commit messages! 🤓

By the way, the hashes will be outdated/wrong later anyways, as Git is going to use SHA-256 hashes instead of the SHA-1 hashes they're using now.
So that would mean all hashes (and pictures with hashes in them) should be replaced to be technically correct again.

https://github.com/git/git/blob/v2.19.0/Documentation/technical/hash-function-transition.txt