On Mon Sep 7, 2026 at 1:20 PM JST, John Hubbard wrote:
On 9/6/26 8:13 PM, Alexandre Courbot wrote:
quoted
On Sat Aug 22, 2026 at 10:54 AM JST, John Hubbard wrote:
<...>
I was contemplating merging this patch early as it is a reasonable
change on its own, but unfortunately these two introduced (and unused
for now) members prevent me from doing so. They are also not related to
moving the Radix3 code around. Can you introduce them when they are
actually used?
Sure.
quoted
This would allow me to merge this one and patch 6 (which btw is
strangely not placed right after this one, with unrelated changes in
between) early and lighten the series a bit.
I think patch 11 can also go in early, but I refrained from picking it
up now to spare you a conflict upon rebasing this series. Ideally this
one would also have been at the beginning of the series.
Don't worry about conflicts, because lately I can resolve those pretty
much automatically, with today's tools.
I do worry because today's tools tend to introduce churn when they meet
a speedbump. :')
Should I sent a separate 2-patch series, with this patch merged in with
patch 6, and patch 11, then? I also would like this series to be
shorter, of course.
A separate series would be the shortest course, I think we could merge
it quickly in this case.
Let's keep the move to a new module and zero-padding as separate patches
though, it makes the history cleaner.