Re: [PATCH v5 0/9] sequencer: don't fork git commit

[Junio C Hamano <gitster@pobox.com>]

Phillip Wood <phillip.wood@talktalk.net> writes:

> From: Phillip Wood <phillip.wood@dunelm.org.uk>
>
> I've reworked the config handling since v4. It now stores the default
> values in struct replay_opt rather than using global variables and
> calls git_diff_basic_config(). Unfortunately I've not had time to
> modify git_gpg_config() to indicate if it successfully handled the key
> so git_diff_basic_config() is called unnecessarily in that case. Within
> git_diff_basic_config() userdiff_config() also suffers from the same
> problem of not indicating if it has handled the key.

Ouch.  I thought we agreed that we were ready to go incremental and
the topic was merged to 'next' earlier last week.

After scanning the difference between the two rounds, it seems that
the more important difference is the rework of the configuration,
which looks better thought out than the previous round, and with
associated change to use replay_opts fields instead of free variables
to carry gpg-sign and cleanup-mode settings around, which also looks
sensible to me.

Can you make these differences into incremental "that earlier one
was suboptimal for this and that reasons, let's make it better by
doing this" patches to queue them on top?

Thanks.