I have a few complaints and observations about this:
- The patch order is screwed up in the series. Without knowing what
the --recode-patch option does, the reader is forced to look-ahead
before judging this patch.
- No documentation in the same patch as the feature is added. I am
guessing that the new configuration variable (and the new option we
will see laster) means "the patchfile I got is in this encoding but the
mail header does not mark it as such, so I am giving what encoding it
is", but this forces the reader to look-ahead.
- "It will be used by ... only", says who? In an environment where
people send patches in a local encoding but want to keep their
repository in a different encoding, it may not be totally implausible
to wish "format-patch" to pay attention to this variable to _produce_
the output in that encoding, especially given the name of the variable
that does not say anything about in which direction it is used, no?
- Assuming that I guessed the meaning of this option and parameter right,
I am not sure if this should be a configuration variable. It implies
that the majority of patches, if not all, are in this single local
encoding that is different from the encoding used in the repository.
Is it common? I dunno.
Junio C Hamano <gitster@pobox.com>