Hello,

I maintain a small astrological chart calculator that converts birth
dates and times to UTC, so historical offsets for Argentina matter to us
in one-minute increments. While checking our results against the ACS
Atlas (through Solar Fire), we found a case where tzdata and that atlas
disagree, and where tzdata's own comments suggest the data may need a
second look. I am not asking for a specific change -- just flagging what
we found, with sources.

1. The 1991 -04 period in America/Argentina/Cordoba

southamerica currently has:

    # Cordoba (CB), Santa Fe (SF), Entre Rios (ER), Corrientes (CN),
    # Misiones (MN), Chaco (CC), Formosa (FM), Santiago del Estero (SE)
    Zone America/Argentina/Cordoba ...
                -3:00   Arg     %z      1991 Mar  3
                -4:00   -       %z      1991 Oct 20

So all eight provinces are at -04 from 1991-03-03 to 1991-10-20.

Two things argue against applying that to the whole group:

(a) The file's own comment, a few lines above, says Shanks & Pottenger
    claim that four of those eight provinces were NOT at -04 in that
    period:

    # - Formosa switched to -3:00 on 1991-01-07.
    # - Misiones switched to -3:00 on 1990-12-29.
    # - Chaco switched to -3:00 on 1991-01-04.
    # - Santiago del Estero switched to -4:00 on 1991-04-01,
    #   then to -3:00 on 1991-04-26.

(b) For Entre Rios, the ACS Atlas gives -03, not -04, in that same
    period. A chart cast for Parana, Entre Rios, 1991-03-31 06:45 local
    is computed by Solar Fire at UTC-3. For contrast, the same program
    gives UTC-4 for San Salvador de Jujuy on 1991-06-30, matching
    America/Argentina/Jujuy. So the atlas does distinguish between these
    provinces, and tzdata's grouping flattens that distinction.

Since the header note says "Unless otherwise specified, data entries are
from Shanks & Pottenger through 1992", and the Cordoba zone comment says
Shanks & Pottenger split that region into six subregions during
1991/1992 ("we haven't verified this yet"), it looks like the -04 line
may be carrying Cordoba's own history for provinces that did not share
it.

2. The national legal time was -03 throughout that period

Argentina's official time is published by the Servicio de Hidrografia
Naval, the body that keeps and distributes legal time:

    https://www.hidro.gov.ar/observatorio/LaHora.asp?op=3

Its historical table lists, without any provincial qualification:

    from 1991-03-03 to 1991-10-19   -3
    from 1991-10-20 to 1992-02-29   -2
    from 1992-03-01 to 1992-10-17   -3
    from 1992-10-18 to 1993-03-06   -2

So -04 in 1991 was in every case a provincial departure from legal time,
not the national rule. That does not by itself refute the -04 entries
(provinces did depart from national time in 2004 and 2008, as the file
already documents), but it does mean each -04 needs province-level
evidence, and right now eight provinces share one line.

3. A second period worth checking: DST 1992-93

For the 1992-93 summer, tzdata applies -02 (the national DST) to every
Argentine zone except San_Luis. Spanish-language secondary sources,
citing contemporary press, say that most provinces did not adopt DST
that summer and that only Buenos Aires, Jujuy, Tucuman, Catamarca,
Entre Rios, Santa Cruz and Tierra del Fuego did:

    https://es.wikipedia.org/wiki/Hora_oficial_argentina

The same article documents the 1991-92 summer as the first one with
widespread non-adoption, and documents Mendoza's provincial law 5566
(standard time at -04, DST at -03, in force in 1991 and 1992, repealed
by law 5841 in April 1992), which is consistent with the current
America/Argentina/Mendoza entries.

I realise Wikipedia is a secondary source. I mention it because it cites
newspaper coverage from the period and because it lines up with (1):
these were years when provincial practice diverged from national law,
and the current grouping of Cordoba cannot represent that.

What would help

If anyone on the list has access to provincial decrees or newspaper
archives from 1991-1993, the specific questions are:

  - Did Entre Rios and Santa Fe follow Cordoba to -04 in 1991, or did
    they stay at -03 as the ACS Atlas indicates for Entre Rios?
  - Which provinces observed DST in the 1991-92 and 1992-93 summers?

I am happy to help chase Argentine sources (I am based in Argentina) if
that is useful, and to send the chart data behind the Solar Fire
comparisons.

Thanks for maintaining this -- the per-province detail already in the
file for 2004 and 2008 is far better than what most sources have.

Best regards,
Sebastián Bru
sebabru.com

Reply via email to