On Fri, Aug 15, 2014 at 4:07 PM, Gerald Combs <gerald@xxxxxxxxxxxxx> wrote:
> On 8/15/14 12:17 PM, Jeff Morriss wrote:
>> Hi folks,
>>
>> One thing I've noticed about the Petri-Dish flag in Gerrit is that every
>> time I set it to +1 I get an error saying "needs code review." Since
>> Chris just noticed and complained about that too (on change 3624) I
>> thought I'd mention it here.
>
> Are you pressing the "Publish Comments" or "Publish and Submit" button
> when this happens? We're currently using the default MaxWithBlock
> setting[1] for the Code-Review label, which means that submission
> requires at least one "+2" set and no "-2"s.
>
> It would be nice if Gerrit disabled "Publish and Submit" when the
> required conditions aren't met but its development team seems to prefer
> error messages. For what it's worth I've been trying to reduce this
> behavior in the Qt interface.[2]
FWIW, I'm using the "new" review screen (you can enable it in
preferences) where there is no "Publish and Submit" button so this
problem doesn't exist.
I quite like the new screen but it still has a few quirks, which is
probably why it isn't default yet.