- tz - lists.iana.org

Re: Minute offsets
by ncm＠cantrip.org Jan. 17, 1997

Jan. 17, 1997

The argument that the time format must be obvious by example, so that idiots won't program it wrong, doesn't really apply any more. With the internet, it's easy to find the free code that implements the standard, and given correct free code, people too lazy to read the standard will also be too lazy to rewrite it. Nathan Myers ncm(a)cantrip.org

1 0

My comments on draft-newman-datetime-00.txt
by kuhn＠cs.purdue.edu Jan. 16, 1997

Jan. 16, 1997

Hello! I am the author of the ISO 8601 summary on http://www.ft.uni-erlangen.de/~mskuhn/iso-time.html and I just found your new Internet draft ftp://ds.internic.net/internet-drafts/draft-newman-datetime-00.txt about a standard date and time notation for the Internet. I agree that such a standard would be very useful and hope to be able to assist you in finishing it. Here are some comments that I wrote down while reading your first draft: 1) You wonder, what standard defines UTC and leap seconds. This standard is ITU-R Recommendation TF.460-4. Using your credit card, you can now order very easily an electronic copy from the International Telecommunication Union Web server http://www.itu.int/itudoc/itu-r/rec/tf.html You'll find there also other related standards regarding UTC and time signal transmission. 2) You suggest in section 3 to implement a rule that assumes for the 2-digit years 50-99 that they belong to the 20th century. Considering that Unix systems cannot represent in time_t any point in time before 1970-01-01 00:00:00Z, and that there are not any electronic documents on the Internet with a timestamp predating 1970, I suggest that 70 is a more reasonable border date. Gives us 20 more years and I expect to be almost dead or at least retired by then. 3) In section 4, you recommend the offset -00:00 to indicate UTC. One might consider the much shorter "Z" notation specified by ISO 8601 for this very common case. Saves a few bytes and is consistent with international radio practice (Z = Zulu = UTC). In section 5.4, you do already allow "Z", so here the text is inconsistent. 4) In section 5.4, you allow the hour value 24. Both 00:00:00 and 24:00:00 are allowed by ISO 8601, but the notation 24:00 identifying the midnight at the end of the day is useful only for time interval notations (e.g., 20:00--24:00) and should not be used for unique timestamps. Better restrict the hour range to 00--23. Times like 24:00:01 definitely violate ISO 8601, but are supported by your grammar. 5) You use the ugly "T" separator between date and time. ISO 8601 supports also the notion of separate date and time fields, so I suggest you simply define your format to consist of a date and a time field separated by a space. This is the ISO 8601 notation used by GNU RCS for instance (see option -zLT). I personally feel that separating date and time field by a space increases human readability and is preferable over the ISO 8601 combined dateTtime field format. 6) In section 6, you suggest an IANA registry of time zone names. Forget this. ;-) Time zone rules and names are thanks to the creativity of our politicians MUCH to dynamic and unstable. For very good reasons, ISO 8601 does not use time zone abbreviations at all. An excellent Internet list of current and historic timezone names that is constantly updated is already available in form of the tz database. You definitely should have a look at this first. <URL:ftp://elsie.nci.nih.gov/pub/tz*>. In addition, your proposal does not use any time zone names (which is a good thing!). 7) You could add a reference to the technical corrigendum 1 for ISO 8601, which fixes a few typos in the standard. 8) You could make the minute part of the time zone optional and require that time zones with an integral number of hours offset to UTC (which are almost all!) should only be represented by a 2-digit offset. That is what GNU RCS does already and I think having only a single colon increases human readability. 9) Feel free to copy my arguments against the ancient 12h am/pm time format from <URL:http://www.ft.uni-erlangen.de/~mskuhn/iso-time.html>, in order to help eradicate this silly notation from new software (for instance, sendmail still uses am/pm in Received: header lines, which I consider to be very bad Internet time practice). You should also not use the am/pm notation in section 5.5, as the am/pm notation is really a very US-specific idea. The rest of the world has used the 24h notation for many decades and other languages than English do not even have terms like am/pm. I highly recommend you to join the tz(a)elsie.nci.nih.gov mailing list, where you'll meet a number of very experienced Internet date, time, and especially timezone experts. Send e-mail to tz-request(a)elsie.nci.nih.gov to join the club and check ftp://elsie.nci.nih.gov/pub/tzarchive.gz for previous discussion. These are my suggestions for your draft. Please let me know when you have made any changes or have any questions. Thanks for your effort! Markus -- Markus G. Kuhn, Computer Science grad student, Purdue University, Indiana, USA -- email: kuhn(a)cs.purdue.edu

2 4

Re: C Code for draft-newman-datetime-00.txt
by kuhn＠cs.purdue.edu Jan. 16, 1997

Jan. 16, 1997

In message <32DD304D.40D(a)Renault.FR>, Antoine Leca wrote: > I didn't change anything, but this raises a point: wouldn't it be better > to pass a « time_t const *t » argument? > This is for consistency with other library functions, and portability > (particularly when time_t is a floating type). This rises an interesting question: Why does the C library functions always access time_t values by constant pointers and not by simply copying the value??? I never understood this part of the C lib API. How is portability increased by not using a call-by-value here? If there is any good reason that this has been done in libc, then I'd of course also add it to my code. I already did it for consistency with the existing library functions, but I'd like to know why this is supposed to be the right thing to do. > ¤ Comparing t with 0 doesn't take any sense (I know one system where > ¤ (time_t)0 is J2000,0, so negative t are allowed before 2000-01-01 12:00:00Z Ok, you are right. I just hope that the system you know does not encode 2000-01-01 11:59:59Z = (time_t) -1 as this value is reserved by the standard for error conditions. > ¤ Basically, you can't trust strftime to output what you want if you > ¤ don't set the locale to "C" before. > ¤ I know this code is overkill in the vast majority of cases. I just checked the standard again and it does not indicate that any of the conversion specifiers "%Y-%m-%d %H:%M:%S" depend on the locale. There ARE many conversion specifiers that depend on the locale, but not the above ones, so I understand that we do not have to switch to the "C" locale. > ****** 8601.c > if (prec > 0) { > sprintf(buf, ".%09ld", nsec); > buf += prec + 1; > ****** 8601wrt.c > if (prec > 0) { > int i=9; > while (i-->prec) nsec /= 10; > sprintf(buf, ".%0*ld", prec, nsec); > buf += prec + 1; > ****** > > ¤ I wrote this modification for avoiding a overflow of the buffer. > ¤ You may consider it as unnecessary (but you then may consider > ¤ issuing a warning in the man page). I checked already that nsec is not out of range previously, so I think this additional check is redundant. Thanks for your good comments and also for those I received from others. I'll post the next revision here soon. Markus -- Markus G. Kuhn, Computer Science grad student, Purdue University, Indiana, USA -- email: kuhn(a)cs.purdue.edu

4 3

minor fixes for tzcode1996o
by Todd C. Miller Jan. 16, 1997

Jan. 16, 1997

This fixes potential buffer overflows, corrects getopt(3) usage (getopt(3) returns -1, not EOF these days), and adds another PCTS tweak to pass an arguably bogus test. The diff is reversed, sorry about that. - todd --- time/localtime.c Mon Jan 13 20:17:09 1997 +++ /home/millert/tmp/tzcode1996o/localtime.c Tue Nov 5 11:50:08 1996 @@ -291,10 +276,10 @@ if (name[0] == ':') ++name; doaccess = name[0] == '/'; if (!doaccess) { if ((p = TZDIR) == NULL) return -1; - if ((strlen(p) + strlen(name) + 2) >= sizeof fullname) + if ((strlen(p) + strlen(name) + 1) >= sizeof fullname) return -1; (void) strcpy(fullname, p); (void) strcat(fullname, "/"); @@ -1379,14 +1364,7 @@ dir = tmcomp(&mytm, &yourtm); if (dir != 0) { if (bits-- < 0) -#ifdef PCTS - { - t += 2; - break; - } -#else return WRONG; -#endif if (bits < 0) --t; /* may be needed if new t is minimal */ else if (dir > 0) --- time/zdump.c Mon Jan 13 20:17:17 1997 +++ /home/millert/tmp/tzcode1996o/zdump.c Tue Nov 5 12:00:05 1996 @@ -224,7 +223,6 @@ show(argv[i], t, TRUE); tm = *localtime(&t); (void) strncpy(buf, abbr(&tm), (sizeof buf) - 1); - buf[(sizeof buf) - 1] = '\0'; for ( ; ; ) { if (cutoff != NULL && t >= cuttime) break; @@ -241,7 +239,6 @@ newtm = *localtime(&newt); (void) strncpy(buf, abbr(&newtm), (sizeof buf) - 1); - buf[(sizeof buf) - 1] = '\0'; } t = newt; tm = newtm; @@ -283,7 +280,6 @@ lotm = *localtime(&lot); (void) strncpy(loab, abbr(&lotm), (sizeof loab) - 1); - loab[(sizeof loab) - 1] = '\0'; while ((hit - lot) >= 2) { t = lot / 2 + hit / 2; if (t <= lot) --- time/zic.c Wed Jan 15 16:41:36 1997 +++ /home/millert/tmp/tzcode1996o/zic.c Fri Dec 27 19:42:04 1996 @@ -478,7 +476,7 @@ (void) textdomain(TZ_DOMAIN); #endif /* HAVE_GETTEXT - 0 */ progname = argv[0]; - while ((c = getopt(argc, argv, "d:l:p:L:vsy:")) != -1) + while ((c = getopt(argc, argv, "d:l:p:L:vsy:")) != EOF) switch (c) { default: usage();

1 0

Revised C Code for draft-newman-datetime-00.txt
by kuhn＠cs.purdue.edu Jan. 16, 1997

Jan. 16, 1997

Here comes the revised version of my example ISO 8601 output routine. It includes almost all suggestions that I have received plus some more and I have added test calls that should print on all POSIX systems the following: Show time in various forms: Fri Feb 13 18:59:59 2009 2009-02-13 18:59:59-05 2009-02-13 18:59:59.012-05 2009-02-13 23:59:59Z 2009-02-13 23:59:59.012Z 2009-02-13 23:59:60.012Z 2009-02-13 18:59:60-05 2009-02-13 23:59:59+00 2009-02-14 05:14:59+05:15 2009-02-14 05:14:59.012345678+05:15 Show a DST switch (Central Europe, October 1996): 1996-10-26 23:30:00Z = 1996-10-27 01:30:00+02 1996-10-27 00:00:00Z = 1996-10-27 02:00:00+02 1996-10-27 00:30:00Z = 1996-10-27 02:30:00+02 1996-10-27 01:00:00Z = 1996-10-27 02:00:00+01 1996-10-27 01:30:00Z = 1996-10-27 02:30:00+01 1996-10-27 02:00:00Z = 1996-10-27 03:00:00+01 1996-10-27 02:30:00Z = 1996-10-27 03:30:00+01 Thanks for the many valuable suggestions. Since I know now finally, why localtime() uses call-by-reference, I feel very comfortable with breaking the consistency with this legacy interface. Do you see any other portability or elegance problems? Markus -- Markus G. Kuhn, Computer Science grad student, Purdue University, Indiana, USA -- email: kuhn(a)cs.purdue.edu /* * Example implementation of the RFC xxxx profile of ISO 8601 * ASCII standard date and time representation in ANSI/ISO C. * * This implementation is only an informal appendix of RFC xxxx and not * part of the standard. This is an early draft version, do not use * this in products yet. * * Please send comments for improvement to * * Markus Kuhn <kuhn(a)cs.purdue.edu> * * Thanks for suggestions to: * * Chris Newman <Chris.Newman(a)innosoft.com> * Ken Pizzini <ken(a)spry.com> * Antoine Leca <Antoine.Leca(a)renault.fr> * * $Id: demo8601.c,v 1.1 1997-01-15 22:02:37-05 kuhn Exp $ */ #include <stdio.h> #include <stdlib.h> #include <time.h> /* * Create an ISO 8601 time string conforming to the RFC xxxx profile * * Input: * * t the time that will be written as returned by time() * utc if true, then UTC instead of local time will be shown * nsec nanoseconds since start of current second or 0 if unknown * prec number of digits (0..9) for decimal fractions of the * second (set to 0 if nsec is unknown) * * Output: * * s time string with strlen(buf) < 36 * * When time_t does not allow to encode leap seconds (like in POSIX), * you can use the value of second 59 in t and add 1000000000 to nsec * in order to represent a leap second. If input or libary errors * are detected, NULL will be returned, otherwise a pointer to the * output string. If buf is NULL, the time string will be written to * a local buffer. This function is not multithreading safe. * * Markus Kuhn <kuhn(a)cs.purdue.edu>, 1997 */ char * create_timestring(char *s, time_t t, int utc, long nsec, unsigned prec) { char *buf, buffer[36]; struct tm ltm, utm, *tmp; long offset; /* safety checks */ s = buf = (s == NULL) ? buffer : s; *buf = '\0'; if (nsec < 0 || nsec > 1999999999L) return NULL; if (prec > 9) prec = 9; /* convert time */ if ((tmp = gmtime(&t)) == NULL) return NULL; utm = ltm = *tmp; if (!utc) ltm = *localtime(&t); /* leap second handling */ if (nsec > 1000000000L) { ltm.tm_sec++; utm.tm_sec++; nsec -= 1000000000L; } /* basic format */ buf += strftime(buf, 20, "%Y-%m-%d %H:%M:%S", utc ? &utm : &ltm); if (prec > 0) { sprintf(buf, ".%09ld", nsec); buf += prec + 1; } /* time zone offset */ if (utc) { *(buf++) = 'Z'; *buf = '\0'; } else { /* we assume without check that the offset is less than 24 hours */ offset = ltm.tm_yday - utm.tm_yday; if (offset > 1) offset = -24; else if (offset < -1) offset = 24; else offset *= 24; offset += ltm.tm_hour - utm.tm_hour; offset *= 60; offset += ltm.tm_min - utm.tm_min; sprintf(buf, "%+03ld", offset / 60); /* hour offset */ buf += 3; if (offset < 0) offset = -offset; offset %= 60; if (offset != 0) sprintf(buf, ":%02ld", offset); /* minute offset */ } return s; } /* Some test code */ int main() { char buf[50]; time_t now; time_t last_dst_hour; int i; puts("Show time in various forms:\n"); now = 1234569599L; /* on POSIX this is 2009-02-13 23:59:59Z */ putenv("TZ=EST5"); puts(ctime(&now)); create_timestring(buf, now, 0, 0, 0); puts(buf); create_timestring(buf, now, 0, 12345678, 3); puts(buf); create_timestring(buf, now, 1, 0, 0); puts(buf); create_timestring(buf, now, 1, 12345678, 3); puts(buf); create_timestring(buf, now, 1, 1012345678, 3); puts(buf); /* leap */ create_timestring(buf, now, 0, 1012345678, 0); puts(buf); /* leap */ putenv("TZ=UTC0"); create_timestring(buf, now, 0, 0, 0); puts(buf); putenv("TZ=TST-5:15"); create_timestring(buf, now, 0, 0, 0); puts(buf); create_timestring(buf, now, 0, 12345678, 9); puts(buf); puts("\nShow a DST switch (Central Europe, October 1996):\n"); last_dst_hour = (((26L * 365) + 6 + 300) * 24) * 3600; putenv("TZ=CET-1CEST,M3.5.0/2,M10.5.0/3"); for (i = -1; i <= 5; i++) { printf("%s = ", create_timestring(NULL, last_dst_hour + i * 1800, 1, 0, 0)); puts(create_timestring(NULL, last_dst_hour + i * 1800, 0, 0, 0)); } return 0; }

1 0

C Code for draft-newman-datetime-00.txt
by kuhn＠cs.purdue.edu Jan. 15, 1997

Jan. 15, 1997

I have done a sample implementation of the draft-newman-datetime-00.txt ISO 8601 profile with the modifications suggested by Paul and myself already included. So far, this is only the write routine, the parser will follow later. Please scrutinize the code below such that it becomes highly portable and reliable ANSI/ISO C. I have some doubts whether this code will handle the local time form correctly during both DST switch hours. Does anyone know a nicer solution for determining the UTC offset of local time with portable library functions reliably? (May be, something like "utm.tm_isdst = ltm.tm_isdst" before mktime() is still necessary?) Well-engineered ISO C example code that can easily be included in new applications will certainly make the ISO 8601 profile RFC more popular and should definitely be added. I would also like to see new format specifiers in GNU date (and POSIX.2 date) that would allow me to easily generate the UTC offset fields. Markus -- Markus G. Kuhn, Computer Science grad student, Purdue University, Indiana, USA -- email: kuhn(a)cs.purdue.edu #include <stdio.h> #include <stdlib.h> #include <time.h> /* * Create an ISO 8601 time string conforming to the RFC xxxx profile * * Input: * * t the time that will be written as returned by clock() * utc if true, then UTC instead of local time will be shown * nsec nanoseconds since start of current second or 0 if unknown * prec number of digits (0..9) for decimal fractions of the * second (set to 0 if nsec is unknown) * * Output: * * buf time string with strlen(buf) < 36 * * When time_t does not allow to encode leap seconds (like in POSIX), * you can use the value of second 59 in t and add 1000000000 to nsec * in order to represent a leap second. If input or libary errors * are detected, buf will be set to be an empty string. This function * is not multithreading safe. * * Markus Kuhn <kuhn(a)cs.purdue.edu>, 1997-01-15 01:00:18-05, early draft */ void create_timestring(char *buf, time_t t, int utc, long nsec, int prec) { struct tm ltm, utm, *tmp; double off; long offset; /* safety checks */ if (buf == NULL) return; *buf = 0; if (t < 0 || nsec < 0 || nsec > 1999999999L) return; if (prec > 9) prec = 9; /* convert time */ if ((tmp = localtime(&t)) == NULL) utc = 1; /* fallback when library does not know local time */ else ltm = *tmp; if ((tmp = gmtime(&t)) == NULL) return; utm = *tmp; /* leap second */ if (nsec > 1000000000L) { ltm.tm_sec++; utm.tm_sec++; nsec -= 1000000000L; } /* basic format */ strftime(buf, 20, "%Y-%m-%d %H:%M:%S", utc ? &utm : &ltm); buf += 19; /* fractional seconds */ if (prec > 0) { sprintf(buf, ".%09ld", nsec); buf += prec + 1; } /* time zone offset */ if (utc) { *(buf++) = 'Z'; *buf = 0; } else { off = difftime(t, mktime(&utm)); if (off < -0.01) { *(buf++) = '-'; offset = -off + 0.5; } else { *(buf++) = '+'; offset = off + 0.5; } offset = (offset + 30) / 60; sprintf(buf, "%02ld", offset / 60); /* hour offset */ buf += 2; offset %= 60; if (offset) sprintf(buf, ":%02ld", offset); /* minute offset */ } return; } int main() { char buf[50]; time_t now; now = time(NULL); puts(ctime(&now)); create_timestring(buf, now, 0, 0, 0); puts(buf); create_timestring(buf, now, 0, 12345678, 3); puts(buf); create_timestring(buf, now, 1, 0, 0); puts(buf); create_timestring(buf, now, 1, 12345678, 3); puts(buf); create_timestring(buf, now, 1, 1012345678, 3); puts(buf); create_timestring(buf, now, 0, 12345678, 9); puts(buf); return 0; }

2 1

Re: -00:00 on draft-newman-datetime-00.txt
by David Keegel Jan. 14, 1997

Jan. 14, 1997

To quote some previously private email (with permission): ] On Sun, 12 Jan 1997, David Keegel wrote: ] > Perhaps when ISO-8601 allows different formats (eg: +hh:mm, +hhmm, or +hh) ] > and RFC822 only allows one of these formats, then it would be appropriate ] > to choose the common format (eg: +hhmm). Chris Newman wrote: ] For the "Internet profile", I selected the +hh:mm form. ISO does allow ] the +hhmm form, but one reading of the spec would say only if all ] punctuation is omitted from the rest of the date time format. If all the ] punctuation is omitted, the format becomes basically unreadable. I agree omitting all punctuation would be a Bad Thing. So the question hinges on whether ISO-8601 says you can omit punctuation from the time- zone, without omitting punctuation from everything else. Would anyone on the tz list like to interpret ISO-8601 to answer this? __________________________________________________________________________ David Keegel <djk(a)cyber.com.au> http://www.cyber.com.au/ +61 3 9642-5997 Cybersource P/L: Unix Systems Administration and TCP/IP network management

1 0

[J. Wunsch: hey (fwd)]
by Garrett Wollman Jan. 13, 1997

Jan. 13, 1997

This is something of a security bug report, but the security problem is not necessarily a bug in the timezone code. However, I do think it's a bug that our code (which is not the very latest tzcode) will crash if the TZ variable is set to point to a non-TZ file. I can think of two ways to combat this: 1) Disallow selection of files containing `.' or starting with a leading `/' unless the name is the same as TZDEFAULT. 2) Use some of the tzh_reserved[] bytes for a magic number. (2) is more flexible, but still allows for a malicious user to cause a core dump. I prefer (1) and will probably implement that unless someone can provide a good reason not to. ------- start of forwarded message (RFC 934 encapsulation) ------- From: J Wunsch <j(a)ida.interface-business.de> To: wollman(a)freebsd.org Subject: hey (fwd) Date: Fri, 10 Jan 1997 12:28:15 +0100 (MET) Hi Garrett, can you deal in time with this problem? Otherwise, i'd look into it myself. - ----- Forwarded message from Adam Kubicki ----- Message-Id: <199701092155.WAA12271(a)innocence.interface-business.de> From: Adam Kubicki <mikee(a)solozzo.tele.pw.edu.pl> Subject: hey To: joerg_wunsch(a)interface-business.de Date: Thu, 09 Jan 1997 23:02:44 MET In-Reply-To: <199612171142.MAA11937(a)ida.interface-business.de>; from "J Wunsch" at Dec 17, 96 12:42 (noon) X-Mailer: Elm [revision: 112.2] hi, there is a bug in tzset() function. setting TZ environment variable to some file, you can cause program to dump core - variables read from this file are used as various offsets in settzname() thus you get sigsegv. Because suid programs dont dump core, its not so dangerous, but you can export TZ in telnet and force login to dump core. gettimeoftheday() in login.c is called after loging in, but before setuid(uid) so you will get login.core in you home directory. this core file will follow symlink allowing you to overwrite any file on system. And, setting TZ to /etc/master.passwd you will find whole master.passwd in core file (touch login.core first to fool default umask/owner core flags). I'be be glad to get a smart patch from you, as quick fix i disabled TZ in telnetd. - -adam - ----- End of forwarded message from Adam Kubicki ----- - -- J"org Wunsch Unix support engineer joerg_wunsch(a)interface-business.de http://www.interface-business.de/~j ------- end -------

3 4

comments on draft-newman-datetime-00.txt
by Paul Eggert Jan. 9, 1997

Jan. 9, 1997

I liked Markus Kuhn's comments on your Internet Draft ``Date and Time on the Internet'' <URL:ftp://ds.internic.net/internet-drafts/draft-newman-datetime-00.txt> (December 1996). I have the following further comments and suggestions, some of them echoing Kuhn, some providing more details. Please let me know if you have any questions about these comments, and I'd appreciate hearing about any updates to your draft. Section 2 This section should mention that all dates use the Gregorian calendar. Section 3 Section 3 suggests that 00-49 should map to 2000-2049, and that 50-99 should map to 1950-1999. This contradicts existing practice in one important set of implementations: the XPG4 standard says that 00-68 should map to 2000-2068, and 69-99 should map to 1969-1999. This accommodates nonnegative Posix timestamps which start at 1970-01-01 00:00:00 UTC (equivalent to 1969-12-31 in some time zones). Any suggestion for mapping 2-digit to 4-digit dates is bound to be controversial. Perhaps it would be best to remove this recommendation entirely. If you must suggest a range, the XPG4 range is a bit better than the range in the December draft, as it matches existing practice better (and it will give you another 19 years to upgrade your software...). The recommendation ``Three digit years MUST be interpreted by adding 1900'' seems to be designed only for interfacing to buggy software (e.g. software that uses C's tm_year value without changing it). It is also incompatible with ISO 8601, which uses 3-digit strings for day-of-year. I suggest removing this recommendation. Section 3 mentions ``two digit'' years and ``three digit'' years without making it clear that we are talking about lexical syntax, not numeric value. This should be clarified. For example, the string `0096' should denote the year 96, not 1996; otherwise there would be no way to specify the year 96. Section 4.1 A good reference for local time zone rules is the tz database maintained by Olson et al (<URL:ftp://elsie.nci.nih.gov/pub/>, updated regularly). Section 4.2 When the local offset is unknown, the offset "-00:00" MAY be used to indicate that the time is in UTC and the local offset is unknown. This is worded a little confusingly -- could you please clarify? Is it common to have situations where UTC is known but local time isn't? Without more motivation, it's hard to see why this suggestion is needed. Section 4.3 An alternative would be to show a list of the timezone labels defined in [section XXX]. I'm not sure what is meant here. Surely by ``timezone label'' you do not mean commonly used strings like `EST', since such strings are ambiguous in practice. E.g. `EST' has one meaning in the US, another in Brazil, and yet another in Australia (where the meaning of `EST' also depends on whether it is winter or summer!). Also, there is controversy about what the time zone labels ought to be -- e.g. should Canadian Eastern Standard Time be called `EST' (English) or `HNE' (French)? If by ``timezone label'' you mean some other identifying label, then I suggest using the Posix TZ string as extended by the Olson package (again, see <URL:ftp://elsie.nci.nih.gov/pub/>). This is common practice, is widely used, and there are multiple implementations; e.g. the source code to another implementation of the client library for the Olson TZ extensions can be found in <URL:ftp://prep.ai.mit.edu/pub/gnu/glibc-1.09.1.tar.gz>. For example, the Olson TZ string `America/Los_Angeles' stands for local time in Los Angeles; the Posix TZ string `CST-8' stands for the time zone named `CST' that is always 8 hours ahead of UTC. The advantage of `America/Los_Angeles' is that it can work correctly for all times in the past (the tz database has entries back to 1883, when Los Angeles adopted standard time), and requires no user modification to work correctly in the future, as the tables can be updated automatically by administrators as needed. Section 5.4 time-hour's range should be 00-23, not 00-24. ISO 8601 allows 24 but people are sometimes confused by it, it makes sorting a bit trickier, and it provides no useful functionality in this context. time-second allows leap seconds (value 60), but many systems do not support leap seconds. (E.g. Posix requires that the host not support leap seconds in time_t values.) The RFC should recommend what to do on a system that lacks leap second support when it is given a time stamp containing a leap second. I suggest that such systems treat values >= 60 as 59.999... (the number of 9s after the decimal point being the maximum allowed by the host). time-secfrac should use "." for the decimal point, as this is more commonly used in computer applications. Usurping "," for the decimal point might cause problems in applications that use "," for other punctuation. ISO 8601 allows "." for decimal point. time-numzone should not require specification of minutes (and seconds) as they are normally zero. E.g. change it to: time-numzone = ("+" / "-") time-hour [":" time-minute] date-time should not require a "T" between the date and the time; it should allow a space as well. This is easier to read, is allowed by ISO 8601, and is common practice. ISO 8601 provides no way to represent years before the year 0000, or after the year 9999. This makes it difficult to represent timestamps in some historical applications. To fix this, you might extend the syntax for date-fullyear to: date-fullyear = ["-"] 4*DIGIT where the years are numbered ..., -0002, -0001, 0000, 0001, 0002, .... (Note that unlike the traditional Julian calendar, there is a year 0 in the modern Gregorian calendar.) Section 6 I don't know of any ``IANA Registry of Timezone Names''. But please see my comments on Section 4.3 above for more details about an existing registry that could be used as the basis of an IANA registry. Appendix A Why is this section needed? Appendix B I don't see why this section is needed, since this draft RFC doesn't care about the day of the week. But if you think it's needed, here's the canonical reference for Zeller's congruence (written in German): Chr. Zeller, Kalender-Formeln, Acta Mathematica, Vol. 9, Nov. 1886 And here is code that is derived directly from Zeller's paper and uses Zeller's notation. #include <stdio.h> /* Return day of week, with 0 meaning Saturday and 1 meaning Sunday. See Chr. Zeller, Kalender-Formeln, Acta Mathematica, Vol. 9, Nov. 1886. */ int zeller (int year, int month, int day) { int jan_or_feb = month < 3; int y = year - jan_or_feb; int J = y / 100; /* century number */ int K = y % 100; /* year number within the same */ int m = month + 12 * jan_or_feb; /* month number */ int q = day; /* day number in the month */ int h = (q + ((m + 1) * 26) / 10 + K + K/4 + J/4 - 2*J) % 7; /* weekday number (1 is Sunday) */ return h; } char *dayofweek[] = { "Saturday", "Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday" }; int main () { int year, month, day; printf("Enter the year (0001-9999): "); scanf("%d", &year); printf("\nEnter the month (1-12): "); scanf("%d", &month); printf("\nEnter the day of the month (1-31): "); scanf("%d", &day); printf("The day of the week is: %s\n", dayofweek[zeller (year, month, day)]); return 0; }

5 10

Re: Precise date of Millenium
by Jamie Zawinski Jan. 8, 1997

Jan. 8, 1997

Eric Ulevik wrote: > > Otherwise, the argument is that there was no year zero > in the Julian calendar, and we should celebrate after 2000 years. > Thus 1 January 2001. But people forget that the Gregorian calendar > skipped a number of days, so this is wrong! > > So if you wish to be pedantic, please celebrate on 1 January 2001 > on the Julian calendar, which is 14 January 2001 Gregorian. Well that depends on what event you're celebrating! 2000 years since the first second (apex?) of the Julian calendar? 2000 years since the first second of the Gregorian calendar? (As reckoned in the Julian calendar? Weird, but I think that's what you described...) 2000 years since the Julian calendar was introduced? That's still some centuries away. 2000 years since Jesus was born? Nobody really knows, do they? But hey, any excuse for a party... -- Jamie Zawinski jwz(a)netscape.com http://www.netscape.com/people/jwz/

2 2