[darcs-users] darcs patch: Refactor "darcs show authors" help.

Trent W. Buck trentbuck at gmail.com
Sun Mar 15 07:15:20 UTC 2009


Simon Michael <simon at joyful.com> writes:

>> + "An author's name or email address may change over time.
>
> Also, this is for tidying up the inevitable silly mispellings and
> inconsistencies that crop up due to unfamiliarity, unconfigured darcs,
> recording patches on strange machines etc.

I didn't want to mention that, because I think it sounds kinda
demeaning, and dealing with legitimate name changes provides an adequate
use case without saying "this is also useful for working around cock-ups
that you accidentally accepted into your repo."

>> + "when multiple author strings refer to the same individual, create an\n" ++
>> + "`.authorspellings' file in the root of the working tree.  Each line in\n" ++
>> + "this file begins with an author's canonical name and address, and may\n" ++
>> + "be followed by a comma and zero or more extended regular expressions,\n" ++
>
> One or more, a comma and zero regexps wouldn't make sense.

Fair enough.

>> + "Any patch with an author string that matches the canonical address or\n" ++
>
> Any patch with an author string that contains the email address
> extracted from the canonical address. That's why I recommend the Name
> <email> form, the canonical email won't be noticed if it's not in
> angle brackets.

Bah, before you were correcting me when I referred to the entire
name-addr as "address", and now you're parsing the new version that way!
My use of "address" above is supposed to be clearly just the addr-spec
part, because of "begins with [...] name and address".

I didn't want to say "Name <email>" because it's conventional to use
italics (which we don't have) or angle brackets to markup the
non-terminals "name" and "email", so it would transliterate to ASCII as
"<name> <<email>>", which is just confusing.



More information about the darcs-users mailing list