[tz] TZ Coordinator candidate

Zefram zefram at fysh.org
Sat Mar 10 08:38:53 UTC 2012


With ado's retirement now almost upon us, I'd like to volunteer for the
role of TZ Coordinator.  Although I've only been directly involved, on
the mailing list, for half a year, I've been peripherally involved for
much longer.  I'm very comfortable working in this space, and I believe
I have the right personal qualities for the job.

Firstly, I have a strong and unshakeable inclination to attend to detail,
a determination to be thorough, and a decisive concern for correctness.
(The former, at least, is an autistic trait, one of several that turn
out to be useful to engineers.)  I believe these qualities have been
apparent in my contributions to the mailing list in recent months.
For further substantiation, see my published code and contributions to
other mailing lists (links below).

I have a calm temperament, which kre rightly identified (on 2009-09-04)
as a vital quality for this role.  This can be seen in action on the
various other mailing lists that I have inhabited in recent years.
Particularly see the leap seconds mailing list, where in the past couple
of years many discussions have regrettably become much more heated
than befits the calibre of participant, and perl5-porters, where I've
occasionally dealt with a prickly CPAN author.

I'm at the right point in my career to take on this sort of role,
potentially for as long as ado has served.  I'm mature enough to take
on a long-term commitment, while still young enough to have a long term.
I was born in 1977, and so am some thirty years away from pensionable age.
I can't promise to actually serve for that long, but there isn't anything
standing in the way of me doing so.

I have, and use, network access nearly every day, so I can respond
promptly to urgent issues.  I read email, as it comes in, during most of
my waking hours, and I'm similarly available on IRC.  I tend to spend
my weekends and holidays (from paid work) programming.  I don't make
a habit of travelling, though I presently attend an annual conference
that occupies me for five consecutive days.  (I haven't so far attempted
to access the net while at a conference, but it is available at these
conferences, and I ought to get with the times.)

Although it's not strictly necessary in a TZ Coordinator, I'm technically
well qualified to look after the timezone code.  C was my main programming
language for about ten years, 1988 to 1998.  Since then my main language
has been Perl, but I still regularly work in C.  About half of the Perl
modules that I've published are largely implemented in C, for one reason
or another, and I'm a contributor to the Perl core, which is written
in C.  The quality of my code can be discerned by perusal of my published
Perl modules.

I have a deep and abiding interest in horology, and particularly in
computer implementation of it.  I've been on the leap seconds mailing list
for some years, actively involved in many discussions there.  In the Perl
world much of my published code is concerned with time, and last year
I gave a well-received talk about time (but not timezones) at two Perl
conferences (video link below).  Looking more specifically at timezones,
some of my Perl code is about timezones, much of that specifically about
using the TZ database.  I've become intimately familiar with the database
through my work on this code.

As a bonus, I have some plans for how the database should develop from
where it is now.  Chief among them, I'd like to see it become more
historically comprehensive, but there are obvious problems for users
that would come from multiplying up the number of timezones on the
present infrastructure.  I've got some clear ideas about how to solve
these problems, principally some concepts for a new tzfile format that
would let zones share representation.  I'm confident of the feasibility,
and I'd like to develop in this direction with a view to supplanting the
present implementation, applying all due caution for stability issues.
It's not necessary for me to be the TZ Coordinator in order to proceed
with such a project, and I hope to do so anyway, but there's some level
of convenience in the roles being colocated.

In the spirit of openness, I'll now also list potential downsides of me
as TZ Coordinator.  Most obviously, I haven't been on the TZ mailing list
very long.  I tend to pick things up quickly, but this is a defect in the
experience I'd bring to the role.  Secondly, I have a bit of a concern
that my existing role as a consumer of the TZ database (via authorship
of relevant Perl modules) might compromise objectivity in developing
the API (loosely speaking) that the database provides to the world.
(On the other hand, it certainly provides a useful perspective.)  Thirdly,
I'm of the firm opinion that timezones ought to be abolished, in favour
of universal use of Universal Time.

Here are links to relevant resources.  Other mailing lists that I'm on:

    http://six.pairlist.net/pipermail/leapsecs/ (leap seconds)
    http://www.nntp.perl.org/group/perl.datetime/ (time data in Perl)
    http://www.nntp.perl.org/group/perl.perl5.porters/ (Perl core)

Published Perl modules:

    https://metacpan.org/author/ZEFRAM

Video of talk about time:

    http://www.presentingperl.org/lpw2011/why-time-is-difficult/

Miscellaneous other bits on my web page:

    http://www.fysh.org/~zefram/

-zefram



More information about the tz mailing list