Technology and Trends
 USENET Archives
  
From: s900...@minyos.xx.rmit.EDU.AU (Craig Macbride)
Subject: Timezone silliness
Date: 1995/06/09
Message-ID: <3r9ken$14m@aggedor.rmit.EDU.AU>#1/1
X-Deja-AN: 104152442
organization: Royal Melbourne Institute of Technology, Melbourne, Australia.
nntp-posting-user: s900387
newsgroups: comp.unix.sco.programmer

Would anyone like to hazard a guess (be as cruel as you like) as to
why some software (eg. VSI*FAX) fails to use the routines within SCO
to manipulate dates and times, but instead reads $TZ directly from
the environment and then makes totally incorrect assumptions about the
contents, such as the following?

* That all timezones are of form "XXXNXXX".
* That N above is never negative.
* That all timezones are in the Northern hemiphere.
(etc)

--
 _--_|\	    Craig Macbride	<cr...@rmit.edu.au>
/      \		I am Pentium of Borg.
\_.--.*/		Division is useless.
      v			You will be approximated.

From: j...@jpr.com (Jean-Pierre Radley)
Subject: Re: Timezone silliness
Date: 1995/06/09
Message-ID: <m0sK5rX-0004FqC@jpr.com>#1/1
X-Deja-AN: 104152370
sender: n...@xenitec.on.ca (xenitec.on.ca News Administrator)
submit-to: sco...@xenitec.on.ca
content-length: 668       
content-type: text/plain; charset=US-ASCII
precedence: bulk
resent-from: m...@xenitec.on.ca
organization: [resent by] The SCOMSC gateway and Propagation Society
mime-version: 1.0
newsgroups: comp.unix.sco.misc

Craig Macbride writes:
> 
> Would anyone like to hazard a guess (be as cruel as you like) as to
> why some software (eg. VSI*FAX) fails to use the routines within SCO
> to manipulate dates and times, but instead reads $TZ directly from
> the environment and then makes totally incorrect assumptions about the
> contents, such as the following?
> 
> * That all timezones are of form "XXXNXXX".
> * That N above is never negative.
> * That all timezones are in the Northern hemiphere.


How is a timezone in the Southern Hemisphere any different, apart perhaps
from its abbreviation?


-- 
Jean-Pierre Radley   j...@jpr.com     XC custodian    Sysop, CompuServe SCOForum

From: Bela Lubkin <be...@sco.COM>
Subject: Re: Timezone silliness
Date: 1995/06/09
Message-ID: <9506091352.ZM23790@sco.com>#1/1
X-Deja-AN: 104152412
sender: n...@xenitec.on.ca (xenitec.on.ca News Administrator)
references: <m0sK5rX-0004FqC@jpr.com>
submit-to: sco...@xenitec.on.ca
content-type: text/plain; charset=us-ascii
precedence: bulk
resent-from: m...@xenitec.on.ca
organization: [resent by] The SCOMSC gateway and Propagation Society
mime-version: 1.0
newsgroups: comp.unix.sco.misc

Jean-Pierre Radley wrote:

> How is a timezone in the Southern Hemisphere any different, apart perhaps
> from its abbreviation?

Daylight savings time would be during a different set of months,
wouldn't it?

>Bela<

From: j...@jpr.com (Jean-Pierre Radley)
Subject: Re: Timezone silliness
Date: 1995/06/09
Message-ID: <m0sKBI5-0004FvC@jpr.com>#1/1
X-Deja-AN: 104152414
sender: n...@xenitec.on.ca (xenitec.on.ca News Administrator)
submit-to: sco...@xenitec.on.ca
content-length: 465       
content-type: text/plain; charset=US-ASCII
precedence: bulk
resent-from: m...@xenitec.on.ca
organization: [resent by] The SCOMSC gateway and Propagation Society
mime-version: 1.0
newsgroups: comp.unix.sco.misc

Bela Lubkin writes:
> 
> Jean-Pierre Radley wrote:
> 
> > How is a timezone in the Southern Hemisphere any different, apart perhaps
> > from its abbreviation?
> 
> Daylight savings time would be during a different set of months,
> wouldn't it?

So it would. How do SCO time routines deal with that, quite aside from the
initial question about the FAX program's timezone problems?

-- 
Jean-Pierre Radley   j...@jpr.com     XC custodian    Sysop, CompuServe SCOForum

From: Bela Lubkin <be...@sco.COM>
Subject: Re: Timezone silliness
Date: 1995/06/09
Message-ID: <9506091424.ZM24106@sco.com>#1/1
X-Deja-AN: 104152415
sender: n...@xenitec.on.ca (xenitec.on.ca News Administrator)
references: <m0sKBI5-0004FvC@jpr.com>
submit-to: sco...@xenitec.on.ca
content-type: text/plain; charset=us-ascii
precedence: bulk
resent-from: m...@xenitec.on.ca
organization: [resent by] The SCOMSC gateway and Propagation Society
mime-version: 1.0
newsgroups: comp.unix.sco.misc

> > Daylight savings time would be during a different set of months,
> > wouldn't it?
> 
> So it would. How do SCO time routines deal with that, quite aside from the
> initial question about the FAX program's timezone problems?

You use a different and much more complex timezone specification.  I
don't know the syntax (someone from one of those parts might be able to
explain it).  The effect is that you can specify something like "the
timezone is called XST except during daylight savings, when it's XDT.
Daylight savings starts on November 20th at 2am and ends on Jan 15th at
1am.  Standard time is GMT+400 and daylight savings is GMT+530" -- all
in a little string of about 40 chars of gibberish.

>Bela<

From: Han Holl <jeh...@euronet.nl>
Subject: Re: Timezone silliness
Date: 1995/06/11
Message-ID: <3reqtq$f68@news.euro.net>#1/1
X-Deja-AN: 104223745
references: <m0sKBI5-0004FvC@jpr.com> <9506091424.ZM24106@sco.com>
content-type: text/plain; charset=us-ascii
organization: Euronet Internet
x-url: news:9506091424.ZM24106@sco.com
mime-version: 1.0
newsgroups: comp.unix.sco.misc
x-mailer: Mozilla 1.1N (X11; I; Linux 1.2.9 i486)

Bela Lubkin <be...@sco.COM> wrote:
[...cut...]
>1am.  Standard time is GMT+400 and daylight savings is GMT+530" -- all
>in a little string of about 40 chars of gibberish.
>
>>Bela<

You don't have to be in the southern hemisphere to have gibberish in your TZ:
Here's the Dutch one:
TZ=CEST-1EEST-2;M3.5.0,M9.5.0/3

Han

From: egg...@twinsun.com (Paul Eggert)
Subject: Re: Timezone silliness
Date: 1995/06/16
Message-ID: <3rtft9$1k8@light.twinsun.com>#1/1
X-Deja-AN: 104598236
references: <m0sKBI5-0004FvC@jpr.com> <9506091424.ZM24106@sco.com> 
<3reqtq$f68@news.euro.net>
organization: Twin Sun Inc, El Segundo, CA, USA
newsgroups: comp.unix.sco.misc

Han Holl <jeh...@euronet.nl> writes:

>You don't have to be in the southern hemisphere to have gibberish in your TZ:
>Here's the Dutch one:
>TZ=CEST-1EEST-2;M3.5.0,M9.5.0/3

That string might work _this_ year, but it won't work _next_ year,
since the Dutch time zone rules will change.

And when you change the string next year,
you'll bollix up times recorded this year.

SCO should adopt the system used in SVR4, Solaris, Linux, NeXTSTEP,
etc.  Then you could set your TZ environment variable to something more
readable like 'Europe/Amsterdam' or 'Australia/Sydney' and never have
to worry about setting TZ again (unless you move).  This is much more
convenient for users than the Posix TZ gibberish.  It handles recorded
dates correctly, unlike the Posix-only solution.  It's upwards
compatible with Posix.  And it's easy to implement, since the source
code and tables are public domain (see <ftp://elsie.nci.nih.gov/pub/tz*>).