Re: [PATCH v5 2/2] worktree: add 'list' command

> Michael Rappazzo <rappazzo@gmail.com> writes:
>> +             strbuf_strip_suffix(&head_ref, "\n");
>> +
>> +             if (starts_with(head_ref.buf, ref_prefix)) {
>> +                     /* branch checked out */
>> +                     strbuf_remove(&head_ref, 0, strlen(ref_prefix));
>> +             /* } else {
>> +              *  headless -- no-op
>> +              */
>> +             }
>> +             printf("%s  (%s)\n", path, head_ref.buf);
>
> Is this new command meant to be a Porcelain?  This would not work as
> a plumbing that produces a machine-parseable stable output.
>
> I am not saying that it _should_; I do not know if we even need a
> 'list' command that is driven from an end-user script that gives
> anything more than "where the work trees are".
>
> My inclination is to suggest dropping the "which branch" code
> altogether and only give "path_only" behaviour.