On Mon, Jan 30, 2023 at 09:58:40AM -0600, Rob Herring wrote: > On Sun, Jan 29, 2023 at 01:11:32PM +0100, Jacopo Mondi wrote: > > On Sun, Jan 29, 2023 at 12:40:03PM +0100, Krzysztof Kozlowski wrote: > > > On 28/01/2023 12:27, Jacopo Mondi wrote: > > > > Add the bindings documentation for Omnivision OV5670 image sensor. > > > > > > > > Signed-off-by: Jacopo Mondi <jacopo.mondi@xxxxxxxxxxxxxxxx> > > > > --- > > > > v6->6.1 > > > > - Use additionalProperties: false for endpoint properties from > > > > video-interfaces.yaml > > > > - List 'remote-endpoint' among the accepted endpoint properties > > > > now that we use additionalProperties: false > > > > > > b4 diff '20230128112736.8000-1-jacopo.mondi@xxxxxxxxxxxxxxxx' > > > Could not create fake-am range for lower series v1 > > > > > > Can you send patches in a way it does not break out workflows? Why > > > making our review process more difficult? > > > > Because it's a nit on a 10 patches series with no other changes > > requested ? > > So? Think of patch series as an 'email transport' for your git branches. > If you rebase your branch, that's a whole new branch to send. > So if a series has a single comment and could be then collected as it is but one patch I saw it happening multiple times on the ML and I thought it was an accepted practice. > > What is difficult exactly ? > > In addition to 'b4 diff', if a maintainer is applying this series, for a > v7 they just do: > > b4 shazam msgid-of-v7 > > For v6.1, they do: > > b4 shazam msgid-of-v6 > git rebase -i ... > <stop on patch 1> > git reset --hard HEAD^ > b4 shazam msgid-of-v6.1 > git rebase --continue > > Which one makes the maintainer's life easier? > With b4 it now certainly makes a difference. As I save patches from my mail client and apply them manually I never really considered picking one patch over the other from the same thread "more difficult". I should have noticed when Krzysztof mentioned b4 in his first reply. > If it's a CI job trying to apply and test this, there's no way it's > going to do the second case. > That's another point yes. Got your message, I'll stop :) Don't think a v7 is needed for this on though (if not other comments ofc) > Rob
