Re: [PATCH v4 0/3] implement fetching of moved submodules
Heiko Voigt writes: > The previous RFC iteration can be found here: > > https://public-inbox.org/git/20171006222544.GA26642@sandbox/ > > This should now be in a state ready for review for inclusion. > > The main difference from last iteration is that we now also support > unconfigured gitlinks for push and fetch for backwards compatibility. > > To implement this compatibility we construct a default name for gitlinks > if there is a repository found at their location in the worktree. I do not remember the details of the patch in the previous round that corresponds to PATCH 2/3 here, so I cannot comment on the incremental improvement between the two, but the fallback in 2/3 looks like a sensible thing to do. Let's see what others, especially those who are interested in the "--recurse-submodules" area, say. Thanks.
[PATCH v4 0/3] implement fetching of moved submodules
The previous RFC iteration can be found here: https://public-inbox.org/git/20171006222544.GA26642@sandbox/ This should now be in a state ready for review for inclusion. The main difference from last iteration is that we now also support unconfigured gitlinks for push and fetch for backwards compatibility. To implement this compatibility we construct a default name for gitlinks if there is a repository found at their location in the worktree. Cheers Heiko Heiko Voigt (3): fetch: add test to make sure we stay backwards compatible implement fetching of moved submodules submodule: simplify decision tree whether to or not to fetch submodule-config.h | 3 + submodule.c | 200 +--- t/t5526-fetch-submodules.sh | 77 - 3 files changed, 210 insertions(+), 70 deletions(-) -- 2.14.1.145.gb3622a4
