[comp.sys.ncr] Sys V Rel. 2 ZONEINFO

jsmithso@aut.UUCP (Jim Smithson) (03/12/90)

Has anyone worked with the NCR timezone software(/etc/zoneinfo files)
under NCR System V Release 2.01.01? It all seems very good except that
I cannot figure out which library the tzsetwall() function ( See ctime(3C) )
is in.  Any clues?

From other articles in this newsgroup I see that the NCR is doing
away with the /etc/zoneinfo stuff in System V release 3.
Can anyone say why it was removed? Did it work?


-- 
James R. Smithson	Ascom Autelca AG      CH-3073 Guemligen/Switzerland
E-mail: jsmithso@aut.UUCP   FAX: +41 31 527 745   Voice : +41 31 529 214
UUCP: ...!uunet!mcsun!cernvax!hslrswi!aut!jsmithso

wescott@Columbia.NCR.COM (Mike Wescott) (03/14/90)

In article <1177@telsys.aut.UUCP> jsmithso@aut.UUCP (Jim Smithson) writes:
> Has anyone worked with the NCR timezone software(/etc/zoneinfo files)
> under NCR System V Release 2.01.01? It all seems very good except that
> I cannot figure out which library the tzsetwall() function ( See ctime(3C) )
> is in.  Any clues?

Try libcieee.a, in localtime.o.

> From other articles in this newsgroup I see that the NCR is doing
> away with the /etc/zoneinfo stuff in System V release 3.
> Can anyone say why it was removed? Did it work?

It worked fairly well.  There were some backwards compatibility and
installation problems involving programs that had old ctime(3) routines
compiled in.

Basically, the zoneinfo approach is a better idea but in our hurry to get
SVID compliant SysVr3 to market we dropped it in favor of the AT&T SysVr3
approach.  Only to see the zoneinfo stuff appear in SysVr4.
--
	-Mike Wescott
	 mike.wescott@ncrcae.Columbia.NCR.COM

art@pilikia.pegasus.com (Art Neilson) (03/14/90)

In article <1177@telsys.aut.UUCP> jsmithso@aut.UUCP (Jim Smithson) writes:
>
>Has anyone worked with the NCR timezone software(/etc/zoneinfo files)
>under NCR System V Release 2.01.01? It all seems very good except that
>I cannot figure out which library the tzsetwall() function ( See ctime(3C) )
>is in.  Any clues?
>
>From other articles in this newsgroup I see that the NCR is doing
>away with the /etc/zoneinfo stuff in System V release 3.
>Can anyone say why it was removed? Did it work?
>
>
>-- 
>James R. Smithson	Ascom Autelca AG      CH-3073 Guemligen/Switzerland
>E-mail: jsmithso@aut.UUCP   FAX: +41 31 527 745   Voice : +41 31 529 214
>UUCP: ...!uunet!mcsun!cernvax!hslrswi!aut!jsmithso

I can tell you this, the zoneinfo package is NOT NCR's own work.  It's a
*public domain* package, the latest incantation of which can be had from
volume18 of comp.sources.unix, look for the package called localtime3.
BTW, the tzsetwall() function is not in X3J11's work, which the timezone
package closely follows.  The package by default builds an object library
called 'libz.a', which is installed in /lib.  If you don't have that library,
NCR did something different.  IMHO NCR shouldn't implement such things
in their system without notifying the user of the deviation.  NCR manuals
give NO credit to ado@ncifcrf.gov, and users who don't know any better
will think this is a standard UNIX facility.  Mind you - I'm not against
NCR distributing some of the great PD software provided us by the net
community, I'd just like them to make it *optional*, perhaps have it on
a separate tape and allow users to install the things they want, and give
credit where it's due.  I'm personally glad if they removed it from OS 3.0.
-- 
Arthur W. Neilson III		| ARPA: art@pilikia.pegasus.com
Bank of Hawaii Tech Support	| UUCP: uunet!ucsd!nosc!pegasus!pilikia!art

guy@auspex.auspex.com (Guy Harris) (03/14/90)

>From other articles in this newsgroup I see that the NCR is doing
>away with the /etc/zoneinfo stuff in System V release 3.

Were they to do so, it would be *quite* amusing, since if they pick up
System V Release 4, which has the Arthur Olson time zone code in it,
they're just going to put it back in again.... 

guy@auspex.auspex.com (Guy Harris) (03/15/90)

>Basically, the zoneinfo approach is a better idea but in our hurry to get
>SVID compliant SysVr3 to market we dropped it in favor of the AT&T SysVr3
>approach.  Only to see the zoneinfo stuff appear in SysVr4.

Coulda just merged 'em and had something SVID-compliant *and* Olsonish,
y'know....

wescott@Columbia.NCR.COM (Mike Wescott) (03/16/90)

In article <3037@auspex.auspex.com> guy@auspex.auspex.com (Guy Harris) writes:
>>Basically, the zoneinfo approach is a better idea but [...] we dropped it

> Coulda just merged 'em and had something SVID-compliant *and* Olsonish,
> y'know....

That's what I wanted to do but we didn't have the time (no pun intended).

--
	-Mike Wescott
	 mike.wescott@ncrcae.Columbia.NCR.COM

guy@auspex.auspex.com (Guy Harris) (03/16/90)

>Mind you - I'm not against NCR distributing some of the great PD
>software provided us by the net community, I'd just like them to
>make it *optional*, perhaps have it on a separate tape and allow
>users to install the things they want, and give credit where it's due.

Uh, while that may apply to something that's a bit of code that can be
treated as a self-contained chunk, e.g. a text editor, it doesn't apply
quite so well to something that *replaces* existing chunks of software,
such as the Olson time zone code.  It makes eminent sense to make it the
standard time conversion code in "libc", since having it as some optional
library means you may have a mix of programs that use two different
methods, which means the system administrator has to set both methods up
independently.

If you have most programs linked against a shared C library, and offer a
choice of one or the other, that might be acceptable, although it still
seems to represent a triumph of concern for form over concern for
content....