Need a correct way to resume an aborted clone

[harendra-kumar]

Currently when a clone aborts due to file name conflict it suggests to do a pull to finish the clone. But that is not equivalent. If the clone had finished successfully it would have run several hooks (e.g. post-clone, post-merge etc.) but a pull does not run those same hooks. On top of that pull might run very different hooks. This leads to unexpected and different behavior.

We should perhaps remember the merge pending state and finish the clone by running the command again. Any other command in that state should fail with a message asking the user to finish the clone first.

[RichiH]

Alternatively, the failed clone could clean up after itself and then abort.

I am somewhat torn on what would be less surprising/annoying.