On Fri, 29 Mar 2024, Andy Fiddaman wrote:
>
> On Thu, 28 Mar 2024, Miroslav Lichvar wrote:
>
> > On Mon, Mar 25, 2024 at 07:10:31PM +, Andy Fiddaman wrote:
> > > How does the attached patch look? This adds `local threshold` as you
> > > suggested, that defaults to 0 which means it is not
On Fri, Mar 29, 2024 at 01:44:34AM +, Andy Fiddaman wrote:
> I've attached the latest patch.
>
> I'm happy to go back and do something here if you'd like, but also
> please feel free to adjust what I've done here to better fit the
> goals of the project.
I applied the patch with some fixes.
On Thu, 28 Mar 2024, Miroslav Lichvar wrote:
> On Mon, Mar 25, 2024 at 07:10:31PM +, Andy Fiddaman wrote:
> > How does the attached patch look? This adds `local threshold` as you
> > suggested, that defaults to 0 which means it is not used. If it's set,
> > then the local reference is not
On Mon, Mar 25, 2024 at 07:10:31PM +, Andy Fiddaman wrote:
> How does the attached patch look? This adds `local threshold` as you
> suggested, that defaults to 0 which means it is not used. If it's set,
> then the local reference is not activated until the root distance drops
> below the value
On Thu, 21 Mar 2024, Miroslav Lichvar wrote:
> > The first, "local oncesynced" prevents the local reference being activated
> > unless time was synchronised at some point in the past.
>
> This makes sense to me, although I think it might be better to
> generalize it as an activating root
