On Mon, Sep 10, 2018 at 3:32 PM Junio C Hamano <[email protected]> wrote:
>
> Stefan Beller <[email protected]> writes:
>
> > I separated this from the other series, making it into 2 patches:
> > This first patch adds tracing for string lists and the next patch that
> > removes the unused function from the string list API.
> > That way we can decide on these two patches separately if needed.
>
> Of course, even though these are 1/2 and 2/2, only one of them and
> not both would apply.
Or you could squash them once we reach consensus that both are good.
> Thanks for sticking to the topic.
>
> Given how simple that "dump them to standard output" code is, I am
> inclined to say that anybody who needs to inspect the contents of
> string list at various points in the code under development can
> create one from scratch even if we did not have this implementation,
> so perhaps 2/2 is a better choice between the two.
This sounds like the consensus is not to take both but only 2/2,
which I'd be happy with, too.
> It is not costing us much to leave it in the code. It's not like
> the function costed a lot of maintenance burden since it was added
> in 8fd2cb40 ("Extract helper bits from c-merge-recursive work",
> 2006-07-25), so the alternative #3 might be to do nothing.
True, but ...
> somebody else in the future to propose removing
is what is actually happening here already, see
https://public-inbox.org/git/[email protected]/
> I am inclined to say we'd take
> 2/2 ;-)
>
> Thanks.
Feel free to take Alexanders patch from 2015 instead.
Thanks,
Stefan
