Re: [netmod] technical discussion around draft-ietf-netmod-iana-timezones-03

[Martin Bjorklund <[e-mail]>]

Paul Eggert <[e-mail]> wrote:
> Eliot Lear wrote:
> > It is possible to traipse through each of the files and grep for
> > "Zone"
> > and "Link".
> 
> Sure, one could do that, and that would sidestep many
> of the political problems.  This would generate a longer list,
> one that I expect doesn't match up to CLDR or with
> any other widely-used time zone resource.  And the list
> would omit commonly-used names like 'UTC0'.  But these
> issues may well not matter to users of the YANG module.

We want the YANG module to match what the underlying system is
actually using, provided they support the TZ Database available from
IANA.

Due to my ignorance I thought that zone.tab provided this list.  But
the underlying idea doesn't change: the YANG module should list all
timezone names in the TZ Database, including the links.

I did the grep excerice and found 579 names marked as Zone or Link,
and 415 in zone.tab.  UTC0 is not in the list, but UTC is.  I think it
is ok that the system actually would internally accept e.g. UTC0, even
if it is not available in the YANG module.

This puts a heavier burden on IANA I guess.  It is of course still
trivial to write the program that takes as input the current YANG
module and the new tzdata<vsn>.tar.gz file and produces a new YANG
module.  If this is what we want to do, I can write that program.  But
the question is if this procedure is ok for IANA?


Also, you asked previously about the POSIX TZ format.  The YANG
module does not support that, but it would be trivial to add, if we
believe that it would be useful.  Currently the module support a
choice of a location string or utc-offset.


/martin