The options for darcs changes (aka log) seem to have accumulated in a
rather random fashion. In particular it is unclear how these options
interact when more than one of them are given simultaneously:

--human-readable
--machine-readable (currently not available to users)
--count
--number
--xml-output
--interactive
--all/--no-interactive

The code in Darcs.UI.Commands.Log is quite inscrutable due to the
explicit flag testing in many different places. --count is currently
broken when used in combination with --interactive or --no-interactive:
in both cases, --count is ignored. I think it should instead override
any interactivity options. I think this broke when the new options
system was added, but before I fix it, we should have a clearer idea
what the specification of (the interaction of) these options is.

Forgot to mention:

--context

Perhaps the --human/machine-readable options were meant to affect (only)
the output of --context? Who remembers when and why they have been added?

Some interesting related issues:

* issue1457 (--foo, --no-foo)
* issue1158 (--match vs --patch)
* issue1534 (--context vs --match)

I could have sworn there was an “we need to overhaul the internal flag
management code” ticket lying around.

(while not really participating in darcs dev as I'd like to be, I'm
always happy to see the UI being improved)

On 9 November 2014 01:26, Ben Franksen <bugs@darcs.net> wrote:
>
> Ben Franksen <benjamin.franksen@helmholtz-berlin.de> added the comment:
>
> Forgot to mention:
>
> --context
>
> Perhaps the --human/machine-readable options were meant to affect (only)
> the output of --context? Who remembers when and why they have been added?
>
> __________________________________
> Darcs bug tracker <bugs@darcs.net>
> <http://bugs.darcs.net/issue2412>
> __________________________________
> _______________________________________________
> darcs-devel mailing list
> darcs-devel@darcs.net
> http://lists.osuosl.org/mailman/listinfo/darcs-devel



-- 
Eric Kow <http://erickow.com>

Hi Eric

This is exactly what I tried to get rolling with the Options redesign. I
already said so elsewhere: the reason I brought this issue up is because
I despaired trying to reverse-engineer a specification from the existing
code...