Re: [PATCH v5 9/9] submodule: support reading .gitmodules when it's not in the working tree

On Thu, Sep 27, 2018 at 7:44 AM Antonio Ospite [off-list ref] wrote:

If you end up touching get_oid() please CC me.

noted. I am not sure I'll touch it anytime soon, though.

Are you suggesting to look into super-prefix for any reason in
particular?

No, I misread the intent of that part of your message

quoted

So I think based on ff6f1f564c4 it is safe to remove all calls to
repo_read_gitmodules.

Thanks for confirming.

OK, so the plan for v6 is:

  - avoid the corruption issues spotted by Gábor by removing the call
    to repo_read_gitmodules in builtin/grep.c (this still does not fix
    the potential problem with nested submodules).

  - add a new test-tool which better exercises the new
    config_from_gitmodules code,

Sounds good.

  - add also a test_expect_failure test to document the use case that
    cannot be supported yet: nested submodules without .gitmodules in
    their working tree.

Personally I would want to live in a world where we don't *have* to nor
*want* to support submodules without .gitmodules in the respective
superproject.

We did support some use cases historically that I would make sure to
continue to support, but I am not sure how much effort we want to spend
on supporting further use cases of incomplete submodules.

Feel free to do so, as such tests help to document the boundaries.

Stefan