Re: [netmod] consensus all: timezone-location and draft-ietf-netmod-iana-timezones-03

[Andy Bierman <[e-mail]>]

>
> ....
> >
> > Having two definitions of timezone names that are not synced is the
> > worst of all solutions. I want 3rd party NETCONF implementations to
> > function on open systems where the packaging and maintenance of the
> > timezone database is separate from the packaging and maintenance of
> > the NETCONF implementation.
>
> Even if the tzdata package is upgraded and the module isn't, I fail to see
> the reason why it should stop functioning, except in really singular cases.
>
>

It is important that the data type be correct 100% of the time.
The failure mode will always be singular cases here.  We would
like to use enumeration, but it doesn't work for this particular typedef.
Identity names cannot have forward slashes or whitespace.
The only base type that will work 100% is string.



> >
> > We started with the idealistic idea there could be a maintained
> > enumeration and it turns out this is not realistic to establish and
> > likely also difficult to implement on open systems. So we are going
> > with a string which I believe will work just fine for a large part of
> > the world. If we find out later that this is not workable, we can add
> > additional objects to report which time zone names are available. At
> > this point in time, we have to ship things.
> >
> > I leave it to Tom to determine concensus. But I note that concensus
> > can be rough.
>
> Sure, if everybody else is happy with the string type, then go ahead with
> it.
>
> Lada
>
> >
> > /js
>


Andy