- tz - lists.iana.org

time zone object
by ncm＠cantrip.org Sept. 30, 1997

Sept. 30, 1997

Paul Eggert <eggert(a)twinsun.com> wrote: > From: ncm(a)cantrip.org (Nathan Myers) > > we also need to ... specify caching of retrieved information with > preference for the cached data up to an expiry date. This requires > representing the expiry date in the cached file. > > I don't think it's realistic to insist on an expiry date for time zone > rules. In practice, these rules can change with very little notice. Many (most?) countries make a practice of announcing changes well in advance, and the overwhelming majority of time zone queries will concern times in those countries. Hence, an expiry period of fraction of the conventional period for such a country (e.g. a month) could radically reduce the TCP traffic for zone queries overall. > For example, last year the Sri Lankan government gave > less than 24 hours' notice of an emergency rule change > that moved the clocks ahead by an hour. See ``Sri > Lanka advances clock by an hour to avoid blackout'', > http://www.virtual-pc.com/lankaweb/news/items/240596-2.html > (1996-05-24). Of course some countries do not make a practice of forewarning us, and the expiry period for their cached records could be as little as 24 hours. A policy with such a short expiry period would *still* radically reduce net traffic for queries against that record, compared to an "always check the authoritative network source" policy. In practice, times for places that change with such short notice have low reliability anyway, so there seems to me little danger in giving them what might seem an unreasonably long expiration time (e.g. a week). Anyway, the minimum expiry period of most current TZ installations is measured in years, so even with the occasional blip the aggregate reliability of the times produced would still be vastly improved over the status quo. Time zone conversions are never better than a "best guess" in any case, for the reasons Paul mentions. > In general, time zone rules are ``until further notice'', > and any automated system for propagating time zone rules > will just have to take this into account. "In general", the expiry period must be assumed to be short. In specific common important cases there is good reason to make it long. The longer it is, the quicker is the response time for users, and the lower is the overall network traffic load. [Perhaps the client should be allowed to specify how long it is willing to wait for a time zone conversion, so that if the network response is too slow the cache is used regardless of its expiry date. (The daemon should update the cache anyway, if it eventually succeeds.)] Nathan Myers ncm(a)cantrip.org

2 1

RE: time zone object
by Olson, Arthur David Sept. 30, 1997

Sept. 30, 1997

As John Cowan pointed out, HTTP would suffice as a low-level protocol for access to zone tables across the net, but we also need to identify authoritative sources of up-to-date information; and specify caching of retrieved information with preference for the cached data up to an expiry date. This requires representing the expiry date in the cached file. Does the "zic" format permit a datestamp? There are bytes "reserved for future use" at the front of the file. --ado

1 0

Re: time zone object
by ncm＠cantrip.org Sept. 29, 1997

Sept. 29, 1997

John Cowan <cowan(a)drv.cbc.com> > Nathan Myers wrote: > > a "double", relative to Jan 1, 2001, ... > > would give microsecond accuracy for now, with > > precision decreasing as we pass and recede from that date; or we > > could use a 64-bit integer count of microseconds from practically > > any choice of epochs. Each representation has its advantages. > > First, note that the term "epoch" refers to the moment of time > represented as zero, not the range of expressible times. The > Unix epoch is 1997 Jan 1 00:00:00 GMT. I believe I used the term "epoch" correctly. Of course (as I said) the choice of epoch is of no interest given a fixed-precision time value. A floating-point time has greatest precision near the epoch. With a 48 bit mantissa, precision would still be better than a millisecond for dates throughout recorded history. > ... I favor the Java convention: a 64-bit signed integer > representing milliseconds, with the same epoch as Unix. That > provides sufficient resolution for normal purposes (anything > that *requires* microsecond resolution probably requires > microcode, embedded programming, or the like), and has a range > clear back to the Carboniferous Period (~300 My B.P.) Some of us do write microcode and embedded programs, and code in such environments tends make more use of time quantities than the is conventional (hence the term "real-time"). Still, the current Java convention seems tolerable for most uses. > > The second familiar problem is the crossover periods for time changes. > > Converting from microseconds-past-epoch to local time is well-defined; > > but conversion back demands accommodation to local time that doesn't > > exist (during "spring forward" gaps); and [2] local time that is > > ambiguous (before or after "fall back"). Since there isn't any really > > satisfactory solution to these, they just need to be visible (and > > unavoidable) in the interface. > > The ADO package uses the is_dst field to resolve the ambiguity. > Except during "fall back" overlaps, this field isn't required when > converting from broken-out local time to absolute time, but it is > used to resolve the ambiguity in that case. The is_dst field fails to solve the problem. The problem is that the caller of the conversion function is generally not equipped to "resolve the ambiguity", or even to determine whether there is an ambiguity to resolve. The overwhelming majority of programs do not (and probably cannot) correctly or meaningfully use the is_dst field. The conversion function itself can tell whether there might be an ambiguity, but has no way to communicate it back to the caller, or to identify the set of alternatives. I believe that the conversion function must return a value that represents 0, 1, or 2 time values, so that the caller is equipped to (and forced to) deal with the ambiguity after all the necessary information is available. We also need a time representation equivalent to "NaN" that permits us to talk about times shortly before the epoch. > > The third problem is that the TZ database on line is available only > > in a form that would be very clumsy to handle in an on-line transaction. > > That is, if it changes at all (which happens frequently), there is no > > way to check for rule changes in a particular locality without downloading > > the whole thing and unpacking it, which would introduce an unacceptable > > delay in an interactive application. > > Why so? A server can load single "zoneinfo" binary files as needed > from a conventional HTTP server: several could be provided around > the Internet, and intranets could have one or two available also. This seems like the beginning of a reasonable approach. Better failure tolerance than "Error 200" seems needed. Nathan Myers ncm(a)cantrip.org

3 2

time zone object
by ncm＠cantrip.org Sept. 26, 1997

Sept. 26, 1997

On Wed, 24 Sep 1997, John Cowan wrote: > I would be interested in hearing from people > who want to help improve the currently rather feeble Java 1.1 > implementation of time zones. Jeffrey Sean Connell wrote: > This couldn't possibly have come at a better time. I agree, this would be the proverbial Good Thing. I need a a similar interface for a C++ locale facet. If the level of traffic gets too great for this mailing list, we can take it off-line. I see several problems in current implementations that I would not want to propagate into an object interface. First, a 32-bit epoch would be a big mistake; we can use a "double", relative to Jan 1, 2001, which would give microsecond accuracy for now, with precision decreasing as we pass and recede from that date; or we could use a 64-bit integer count of microseconds from practically any choice of epochs. Each representation has its advantages. The second familiar problem is the crossover periods for time changes. Converting from microseconds-past-epoch to local time is well-defined; but conversion back demands accommodation to local time that doesn't exist (during "spring forward" gaps); and [2] local time that is ambiguous (before or after "fall back"). Since there isn't any really satisfactory solution to these, they just need to be visible (and unavoidable) in the interface. The third problem is that the TZ database on line is available only in a form that would be very clumsy to handle in an on-line transaction. That is, if it changes at all (which happens frequently), there is no way to check for rule changes in a particular locality without downloading the whole thing and unpacking it, which would introduce an unacceptable delay in an interactive application. To check only locally depends on "system administrators" to install updates, but most Java platforms will have (for practical purposes) no system administrator. One solution would be a network of TZ servers that can hand out up-to-date fragments of the database on demand, similar to DNS service. A good time zone class would adapt to different sources of zone information in much the way IP name lookup happens, but with fallback to a local cache when the network is inaccessible. Probably the IETF should be involved in this last bit ... do any IETF members monitor this list? Nathan Myers ncm(a)cantrip.org http://www.cantrip.org/

4 3

FW: South Africa
by Olson, Arthur David Sept. 25, 1997

Sept. 25, 1997

I don't know the basis of the "official abbreviaton" phrase in Mark's letter. --ado ---------- From: Mark R V Murray[SMTP:markm@iafrica.com] Sent: Thursday, September 25, 1997 1:49 PM To: tz(a)elsie.nci.nih.gov Hi In your timezone info for Africa, you have "invented" SAT for South African Time. Please will you change this to the official abbreviation of SAST for South African Standard Time. Countries outside South africa use the CAT, EAT, WAT etc names. Thanks! M -- Mark Murray Senior Network Engineer, UUNET Internet Africa Join the anti-SPAM movement: http://www.cauce.org

1 0

None
by Mark R V Murray Sept. 25, 1997

Sept. 25, 1997

Hi In your timezone info for Africa, you have "invented" SAT for South African Time. Please will you change this to the official abbreviation of SAST for South African Standard Time. Countries outside South africa use the CAT, EAT, WAT etc names. Thanks! M -- Mark Murray Senior Network Engineer, UUNET Internet Africa Join the anti-SPAM movement: http://www.cauce.org

1 0

Java class ADOTimeZone
by John Cowan Sept. 25, 1997

Sept. 25, 1997

I am in the process of implementing a Java 1.1 class called "ADOTimeZone" which provides support for the ADO timezone information. The different zones, compiled on a Unix system, are being included with the Java class as class-specific resources. Currently I have compilable code, but I have not yet begun component testing. When the code works, I will be giving it away freely. I would be interested in hearing from people who want to help improve the currently rather feeble Java 1.1 implementation of time zones. -- John Cowan http://www.ccil.org/~cowan cowan(a)ccil.org e'osai ko sarji la lojban

2 1

None
by Devang R. Vaidya Sept. 16, 1997

Sept. 16, 1997

Hello, Does anyone maintain a list of setting for the tz (timezone) parameter for various regions. For the eastern timezone of US, TZ=EST5EDT,4,1,0,3600,10,4,0,7200,3600 is being set. Thanks, Devang

1 0

Re: ISO C strftime() extention for ISO 8601
by kuhn＠cs.purdue.edu Sept. 16, 1997

Sept. 16, 1997

"Clive D.W. Feather" wrote on 1997-06-30 13:14 UTC: > You may be interested to know that WG14 adopted last week working paper > N733, which adds the following items to the strftime() function: > > %f is replaced by the weekday as a decimal number (1-7), where > Monday is 1 (the ISO 8601 weekday number). > %F is equivalent to "%Y-%m-%d" (the ISO 8601 date format). > %T is equivalent to "%H:%M:%S" (the ISO 8601 time format). > %V is replaced by the ISO 8601 week number of the year (weeks > begin on a Monday, and week 1 is the week that includes both > January 4th and the first Thursday of the year) as a decimal > number (00-53). Thanks a lot for forwarding this interesting working paper. Please forward the following comments to the authors of N733 and whoever else might be interested. If possible, I would like to review the final text of the new strftime() definition (will the current draft be on the Web somewhere?). Comments on WG14 proposal N733 ------------------------------ Markus Kuhn <kuhn(a)cs.purdue.edu> -- 1997-06-30 (= 1997-W27-1) I appreciate this proposal and I have the following related comments. a) The valid range for %V numbers is 01-53. There is no week 00 in the ISO 8601 week numbering scheme. The week before a week 01 is either week 52 or week 53 of the previous year. b) %V alone is not sufficient to be able to use the ISO week numbering system. Each week is associated with a year, the year in which the majority of the days of this week fall, and this is not necessarily the year in which all days of the weeks fall. For instance: The week 1999-W52 goes from 1999-12-27 to 2000-01-02, in other words, the day 2000-01-02 has a week notation of the form 1999-W52-7. There should be another format descriptor for the year to which the current ISO week belongs, preferably both in 4-digit (%G) and 2-digit (%g) form. If you implement an algorithm for %V, you'll get the value of %G anyway as a by-product very easily, and therefore it should be made available to the strftime() user. c) Existing practice: The Olson tzcode package <ftp://elsie.nci.nih.gov/pub/> contains a widely used strftime() implementation that supports already: %u ISO 8601 week day number (1 = Monday, 7=Sunday) %V ISO 8601 week number (01-53) %G ISO 8601 year of current week, 4-digits %g ISO 8601 year of current week, 2-digits Unless there is a good rationale for the characters suggested by N733, I would suggest to stick with %u instead of %f for the weekday number, and I hope that you will add %G and %g as used in the Olson package and Arnold Robbins' strftime version 3.0. Therefore, if I evaluate on 1977-01-02 the string %G-W%V-%u, I should get 1976-W53-7, and on 1975-12-29 I should get 1976-W01-1. Not directly related to N733, but affecting the same part of the standard, I have a number of other suggestions: d) The range for %S and tm_sec is currently defined to be 00-61 to provide for "as many as two leap seconds". This was based on a serious missunderstanding and there can never be two leap seconds per day as it becomes very obvious by reading ITU-R Recommendation TF.460-4 (I can send you a copy if you are interested). Since this 00-61 range is being widely quoted in other standards, this error should be fixed just to stop spreading this serious missconception of how leap seconds work. The correct range is 00-60. This is not an interoperability problem, but fixing this would make WG14 look like they know what they are doing, and it is therefore a good idea. e) I wonder whether %W is anywhere used and whether this field could be dropped to simplify implementation and memory cost. Countries that start the week with Monday normally use ISO 8601 week numbers (%V) and not the scheme defined by %W. I suspect %W was defined based on a missconception of how week numbers work in Europe. Unless anyone can come up with an example where %W is used or needed, I suggest to drop it as it looks completely useless to me (and please don't quote standards that just copied the %W from ISO C). f) In the definition of %y and %Y, the first two digits of a four digit year are refered to as the "century", which is problematic, since the years 1999 and 2000 belong to the 20th century, but 2001 belongs to the 21st century. Suggested better wording: "%y is replaced by the last two digits of the year as a decimal number (00-99)". Again, not a serious interoperability problem, but it makes WG14 look like they know what they are doing. g) mktime() is the inverse function of localtime(), but there exists no portable inverse function for gmtime() that converts a struct tm given in UTC into time_t. This is a serious problem, and the addition of a new function (e.g., mkgmtime() might be a possible name) should be considered seriously. It is not possible to invert gmtime() in a 100% portable way in an application program, and in practice, I have encountered awful hacks like binary searches over the time_t range to invert gmtime() in an as portable as possible way. See <http://www.ft.uni-erlangen.de/~mskuhn/iso-time.html> for further info. Markus -- Markus G. Kuhn, Computer Science grad student, Purdue University, Indiana, USA -- email: kuhn(a)cs.purdue.edu

2 1

Free Submission to 620 Search Engines
by 88847271＠15831.com Sept. 13, 1997

Sept. 13, 1997

Now when you sign up for web hosting with http://www.VirtualServing.com you get free use of The Central Registry to submit your site to over 620 search engines.* http://www.VirtualServing.com provides high speed virtual website hosting services for $40.00 per month including the following: - Automatic site retrieval and conversion to our servers for new customers (in most cases) - Microsoft Front Page Server extensions - Support for Active Server Pages & Microsoft Visual InterDev - Free use of The Central Registry to submit your site to over 620 search engines* - Comprehensive monthly web site statistics - Choice of primary server country - Unlimited bandwidth usage - 5 inbound E-Mail boxes with E-Mail forwarding - FTP access with one username and password - Microsoft Access 97 - Microsoft SQL Server 6.5 - Microsoft Index Server - Microsoft Transaction Server - Microsoft Distributed Transaction Coordinator - Microsoft Certificate Server - Coming with MS Internet Information Server 4.0 in October 1997 - Ability to Administer your web site over the web - Coming with MS IIS 4.0 in October 1997 - CGI program support - Complete daily backups - One DNS mapping for your domain name - Access to raw history files - ISAPI support - 20 MB of disk storage This is a limited time offer so come to http://www.VirtualServing.com and sign up today! Thank you. * Clients must prepay the first six months of service to qualify for one time free use of The Central Registry

1 0