Re: [Tzdist] Next step

[Ken Murchison <[e-mail]>]

On 10/29/2015 04:53 AM, Eliot Lear wrote:
> <chair hat on>
>
> On 10/28/15 8:35 PM, Cyrus Daboo wrote:
>
>
> So if we drop that suggestion, that just leaves two possible things
> for the WG:
>
> 1) Registering a zoneinfo media type (which I think can be done
> without the need for a WG).
> Do others agree or disagree?

I agree that its worth registering a zoneinfo/tzdata/tzfile (or 
whatever) media type.  I don't know if this can/should be done outside 
of a WG.  I'll leave that for the chairs/ADs to decide. Having just 
written an iCalendar -> tzfile(5) translator, I'd be willing to help 
write a draft, but would defer to Paul on this.




>> 2) Something new to think about: should we propose that IANA run a
>> tzdist service as another facet of the time zone database support they
>> do? That would certainly be a way to "bootstrap" interest in the
>> protocol.
> The same sorts of questions:
>
> 1.  Would the community make use of such a service?  I think when it
> comes to a continuing operational cost, there would have to be very
> strong and broad demand.
> 2.  Is it appropriate for IANA to perform this function?  This one would
> require discussion once we get past (1).
> 3.  Is there a draft author?   Never too early to write a draft to
> define the service requirements and how it would be managed.  It's not
> all protocol.
> 4.  Are there reviewers?

I would think that admins and OS vendors would love to have "real-time" 
updates of tzdata rather than having to download, compile, and 
distribute that data themselves.  If the community were to switch to 
using tzdist for tzdata distribution, IANA seems to be an appropriate 
primary provider.  I'd be willing to contribute to and/or review a draft 
of the service requirements.

-- 
Kenneth Murchison
Principal Systems Software Engineer
Carnegie Mellon University