- tz - lists.iana.org

Re: c-common.c; strftime.c; question
by alex＠agsm.unsw.edu.au Aug. 23, 1996

Aug. 23, 1996

>And a year-2000 related question: might it be best to change strftime so >that "%x" yields, for example, > 1996-08-22 >rather than > 08/22/96 >as is currently true? Absolutely! It infuriates me that such a result should ever be returned. Any concession to the US-only we'll-do-it-our-way-and-never-change-and-claim-to-be-best-no-matter-how-bad- it-is all-numerical way of writing dates should be eliminated. Alex LIVINGSTON Macintosh Support Computing & IT (C&IT) Australian Graduate School of Management (AGSM) The University of New South Wales (UNSW) Sydney NSW 2052 AUSTRALIA E-mail : alex(a)agsm.unsw.edu.au; cit(a)agsm.unsw.edu.au (C&IT) Facsimile: +61 2 93137279 Telephone: +61 2 99319264

2 1

c-common.c; strftime.c; question
by ado Aug. 22, 1996

Aug. 22, 1996

A new version of the GNU C compiler's file c-common.c is available: ftp://elsie.nci.nih.gov/pub/c-common.c This version does strftime format checking against the formats recognized by an impending release of GNU's strftime.c; as suggested by Paul Eggert, it's also been retooled to minimize the differences between it and the stock c-common.c (at the cost of a bit of obscureness). Also available is ftp://elsie.nci.nih.gov/pub/strftime.c which is a first attempt at run-time strftime format checking. If the environment variable CHECK_STRFTIME_FORMATS_FOR_TWO_DIGIT_YEARS is defined, strftime formats are checked at run time and diagnostics are written to stderr about formats that produce two-digit years in all locales, in the current locale, or in some locales. And a year-2000 related question: might it be best to change strftime so that "%x" yields, for example, 1996-08-22 rather than 08/22/96 as is currently true? Feedback on the new files and the question most welcome! --ado

1 0

Re: c-common.c
by Ulrich Drepper Aug. 20, 1996

Aug. 20, 1996

From: Paul Eggert <eggert(a)twinsun.com> Subject: Re: c-common.c Date: Mon, 19 Aug 1996 19:42:12 -0700 > Does c-common.c and/or the strftime in version 2.0 of glibc address > these points [about ISO 8601 week notation]? > > glibc 2.0 strftime doesn't. I am preparing a fix along the lines of > the tz strftime. The fix will also make GNU strftime comply with > Posix.2 section 4.15.4, and fix a few other bugs that I've discovered > by code inspection (e.g. sometimes it returns 0 when it shouldn't, and > it assumes 32-bit time_t's in a place or two). Paul, you will correct me, but a few weeks ago you tested the "week" code in strftime and you hopefully saw that I have no %U, %V, and %W implemented (all three kinds of week representation from XPG4 and therefore ISO 8601). If think %W is the representation Jim van Zandt asked for. -- Uli --------------. drepper(a)cygnus.com ,-. Rubensstrasse 5 Ulrich Drepper \ ,--------------------' \ 76149 Karlsruhe/Germany Cygnus Support `--' drepper(a)gnu.ai.mit.edu `------------------------

2 1

Re: c-common.c
by Ulrich Drepper Aug. 20, 1996

Aug. 20, 1996

From: ado(a)elsie.nci.nih.gov (Arthur David Olson) Subject: c-common.c Date: Mon, 19 Aug 96 17:25:58 EDT Hi, I took a fast look at the source and it seems to be really fine. Only one thing is not the best we could get: ** Only format characters recognized by glibc 1.09.1's strftime appear below; 1.09.1 is the last official release, but we have version 2.0 near to a date of release and the strftime function was changed. I would happily send you the last version so that you could add the other formats (mostly GNU extensions). Thanks, -- Uli --------------. drepper(a)cygnus.com ,-. Rubensstrasse 5 Ulrich Drepper \ ,--------------------' \ 76149 Karlsruhe/Germany Cygnus Support `--' drepper(a)gnu.ai.mit.edu `------------------------

1 0

UNIX System Administration Handbook CDROM
by Evi Nemeth Aug. 19, 1996

Aug. 19, 1996

August, 1996 Dear Arthur: A CD-ROM of useful software, tools, and facts for system administrators has replaced the appendices as part of the second edition of UNIX System Administration Handbook (Prentice Hall, 1995). We are about to do a new printing of the book and CD and would like to include your software package "tzcode, tzdata" on the CD. Please excuse the form-letter-ness of this message, it's due to the number of software packages in question and a lack of sophistication in my PERL programming skills. My purpose in writing covers three issues: availability, correctness of our index entry, and continuing permission to include your software. 1) We would like to confirm that "tzcode, tzdata" is in fact freely redistributable and that you have no objection to its inclusion on our CD. The CD is included with the book at no cost; the current version is available as an update for a modest distribution cost. The image of the CD can also be accessed by anonymous ftp from cs.colorado.edu. 2) We want to make sure that our page-length "index entry" from the CD for your software is accurate and complete. Both the index entry and the tentative table of contents of the CD are included below. 3) In order to expedite future updates of the CD we'd like you to choose a permission model from those listed below (or suggest your own): a) OK to include the software forever; don't bother me again b) OK to include the software, but ask each time you revise the CD c) Please DO NOT distribute my software We intend to upgrade as software evolves, but new releases of the CD will track printings of the book and so the timing is not under our control. If you agree to a "don't ask me again" model, it would help us keep our index pages current and correct if you would make the README associated with your software include any changes in the maintainer or primary ftp site. Please use email to respond; if we don't hear from you we will assume that it is OK to include your software and that our index entry is correct (permission model b), but we would prefer an explicit answer. Thank you, -evi Evi Nemeth, Assoc Prof Computer Science Dept University of Colorado Boulder, CO 80309-0430 303-492-5385

1 0

Sri Lanka situation
by Chris Carrier Aug. 17, 1996

Aug. 17, 1996

Paul Eggert > INTERNET:eggert@twinsun.com wrote: >despite the phrases used in the news articles, the Sri Lanka time switch >isn't daylight time in the usual sense -- it's a permanent change Are you sure of this? Is it 'for the duration of the emergency' or is it meant to permanent and for all time, or is it year-round until a definite date in the future? (As an aside, YRDST in the United States, when implemented in 1942, was to last until six months after the end of the war, but was modified by another bill in the summer of 1945, after the war was won, to end on the last Sunday in September, the ending date for most jurisdictions observing seasonal DST in the twenty years before the war. A lot of war-related legislation had a 'duration plus six months' clause in it.) I had imagined that the change would be year-round and not seasonal, because of SL being so near the equator; I ran some sunrise/sunset data for Colombo and noted there was only a 36 minute difference between the latest (February 2, 6:29 GMT+5:30) and earliest (May 26, 5:53 GMT+5:30) sunrises of the year.

1 0

Sri Lanka, Arthur C. Clarke
by Chris Carrier Aug. 16, 1996

Aug. 16, 1996

Did anyone get a response from anyone in Sri Lanka concerning the daylight- time situation? I've been subscribing to soc.culture.sri-lanka and haven't seen anything there other than Paul's posting, as the ng seems to be mostly about Sinhalese and Tamils living outside Sri Lanka cheering on their respective home teams and accusing each other of war crimes. Also, I wonder if Arthur C. Clarke, who lives in Sri Lanka and has made comments on time zoning, has had anything to say about it. The future of time zoning is of interest to me, especially since I live in California, and Clarke in the novel IMPERIAL EARTH (1976) projected that in 2276 we would have only four time zones. Although not explicitly stated, I believe that the times he used were GMT-5 for both American continents, GMT+1 for Europe and Africa, GMT+7 for Asia, and either GMT+13 or -11 for Australia/NZ/Pacific Ocean. Chris Carrier

2 1

Year 2000
by ado Aug. 12, 1996

Aug. 12, 1996

One thought is to provide the GNU folks with code to support a compilie-time option that checks strftime arguments and warns about warn about two-digit-year format specifications (along the lines of the checking now done of printf and scanf formats). Another thought is to set up strftime to warn about two-digit-year format specifications if a sufficiently obscurely named environment variable is present. Does anyone know of prior art in these (or other) areas? --ado

1 0

Re: Supplying names for timezones in the data files
by ado Aug. 5, 1996

Aug. 5, 1996

> # ZONE-DESCR America New_York United_States Eastern time > > Good idea, but how about a format like this instead? > > America/New_York +404251-0740023 US US Eastern time > > where the columns are: > > * zone name in tz database > * latitude and longitude (in ISO 6709 format) of the zone's main location > * country code (in ISO 3166 2-letter format) > * an (English) comment describing the zone One approach would be to, for example, Link America/New_York Coords/+404251-0740023 Link America/New_York Descrips/US_Eastern_time --ado

3 2

S.u.S.E. Linux 4.2 for Developers
by Burchard Steinbild Aug. 5, 1996

Aug. 5, 1996

To: tz(a)elsie.nci.nih.gov Date: Mon, 5 Aug 1996 18:06:23 +0200 (MET DST) X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Content-Length: 2242 Hello, first of all: Sorry, if this mail seems to be a little bit unpersonal. It will be sent to a lot of people - so I don't have the time to write so many emails. It is also possible, that you get this message more than once on different email addresses, just delete the following ones ;-) Maybe you know that we produce Linux-CDROMs (some prefer: Linux based GNU System CDROMs) - maybe you've already gotten one :-) . In any case, our new triple-CD (the first one with a complete own distribution, the former releases always based on Slackware, Thanx to Pat for his great work) has been derived from your work and we would like to say "Thank You" by sending you a FREE DEVELOPER'S COPY of our CDROMs. So if you care to get hold of one, please send me your snail mail address. Don't worry about shipping costs - we'll try to find someone in every edge of this planet to carry the parcels to a local post office :-) We will allow 2-3 weeks for your response to reach us before we start shipping outside of Germany. So please be patient if it takes a little while until the packet reaches you. So, if you're interested, just reply to this mail ;-) - if you're not interested and don't want this offer anymore, drop me a line as well. Thanx again & cu ...bs PS: I think I've taken great care with the email addresses, but never- theless it is possible, that I catched an address of a mailing list. If this has happened, please be fair - this offer is only meant for developers, not for the plain user. PPS: If you answered the previous developer's mailing, and didn't get your copy until now, please do not despair! Probably the address was faulty or incomplete. Try again, so we can sort out what went wrong. PPPS: Take a look at http://www.suse.de, if you're interested in more information or simply ask me by email ;-) PPPPS: If you're interested in a list who got this offer (1000+), drop me a line. Maybe you'd like to check, if I've forgotten someone ;) -- My tip for this evening: Zappa: In New York --- Burchard Steinbild, S.u.S.E. GmbH, Gebhardtstr. 2, 90762 Fuerth, Germany Tel: +49-911-74053-0, Fax: +49-911-7417755, Email: bs(a)suse.de

1 0