Re: [PATCH v6 06/11] ref-filter: implement '--merged' and '--no-merged' options

Karthik Nayak [off-list ref] writes:

quoted

I also have a feeling that compared to an implementation based on
paint_down_to_common(), including is_descendant_of(), this may be
less precise (e.g. it would be confused with clock skew that lasts
more than SLOP commits).  If we are inventing a new helper (as
opposed to moving an existing one), we'd probably be better off
doing a thin wrapper around paint_down_to_common() than calling
revision-walk machinery.

I'll have a look and get back to you.

Not as part of this series, now I know it is a straight-forward port
of what we already have for "branch --merged".  This series is not
yet about improving counting logic but first unifying the three.

Thanks.

On Tue, Jun 30, 2015 at 9:28 PM, Junio C Hamano [off-list ref] wrote:

Karthik Nayak [off-list ref] writes:

quoted

quoted

I also have a feeling that compared to an implementation based on
paint_down_to_common(), including is_descendant_of(), this may be
less precise (e.g. it would be confused with clock skew that lasts
more than SLOP commits).  If we are inventing a new helper (as
opposed to moving an existing one), we'd probably be better off
doing a thin wrapper around paint_down_to_common() than calling
revision-walk machinery.

I'll have a look and get back to you.

Not as part of this series, now I know it is a straight-forward port
of what we already have for "branch --merged".  This series is not
yet about improving counting logic but first unifying the three.

Thanks.

Sure, added this to my personal TODO :)

-- 
Regards,
Karthik Nayak