The previous check simply compared whether the last commit of the branch
was the same in the local and the remote repository.

This commit also checks whether the remote commit is an ancestor of the
local commit. This fixes the case where the local repository has new
commits after the last published commit.

Reference: https://www.zsh.org/mla/users/2022/msg00067.html

Co-authored-by: Philippe Troin <phil@fifi.org>

Fixes #9699

Fixes #10587

Fixes #10448

Fixes #10404

Fixes #8387

Fixes #10187
Co-authored-by: NoPreserveRoot <NoPreserveRoot@pm.me>

This patch removes the 'exit 1' status code when the update check
finishes or is ended with CTRL+C.

Fixes #10033

* Don't error on upgrade no-op

No error code is required for a non failure scenario.

* Manually check whether changes were pulled in `omz update`
Co-authored-by: Marc Cornellà <hello@mcornella.com>

Fixes #9495

* Suppress the problematic trap output in check_upg

The newly added trap, in systems where `rm` is aliased to `rm="rm -v"`,
shows a message stating that "update.lock" has been removed each time `zsh` is called.
I simply suppressed it with directing the output to `/dev/null`.

* Use `command` instead of >/dev/null to suppress

Co-authored-by: Marc Cornellà <marc.cornella@live.com>

This early return made it so the lock wasn't removed, therefore locking out
the upgrade script from ever entering the upgrade routine.

Fixes #6138 

Note: the logic needs some rework.

`[Oh My Zsh] Would you like to check for updates? [Y/n]: ` does not make sense, 
since answering yes will download/apply the new updates instead of checking for them.

This makes things work even on system lacking /bin/sh, such as on
non-rooted Android systems.

Reverts #2296, but mostly #1883.

There is no need to source ~/.profile when this script is read. oh-my-zsh writes no configuration data in ~/.profile.

If the user wishes to use data within ~/.profile, then they should source it in another place.

Fixes #2315

Some environments (such as Android) does not have /usr/bin.