Loading ...
Sorry, an error occurred while loading the content.

Leap Seconds

Expand Messages
  • johnmsteele
    OK, obviously, no one was interested in the TAI 64 format. However, I was really trying to start a discussion of leap second handling, and TAI64 was a specific
    Message 1 of 18 , Mar 16, 2004
    • 0 Attachment
      OK, obviously, no one was interested in the TAI 64 format. However, I
      was really trying to start a discussion of leap second handling, and
      TAI64 was a specific example format of one approach. I'll make this
      more properly general and call it "LEAP SECONDS." Leap seconds were
      apparently last discussed here for a few messages beginning with Msg.
      #468, 2002-07-01. As background, UTC uses the SI (or TAI) second as
      its basic unit; however, over time, it gradually loses
      synchronization with the rotation of the earth, which is erratic, but
      generally slowing down. This is "fixed" by adding leap seconds to the
      time stream. Although negative leap seconds are possible, all to date
      have been positive and take the form of changing the normal time
      count to 23:59:59 -> 23:59:60 -> 00:00:00 (and the date advances).

      Although some operating systems can represent the instance of leap
      second, the conversion formulas from date-time to internal count
      assume a uniform 86,400 s/day and forget the leap second afterward.
      Time spans measured across a leap second are generally in error by a
      second. Since many computer clocks, especially in the past, haven't
      been very accurate anyway, this hasn't been a very big deal.
      However, as more applications depend on accurate time from NTP
      servers, it may be. In many applications, continuity and "reasonable
      accuracy" across the leap second are more important than absolute
      accuracy. The clock may be deliberately run at the wrong rate for a
      short period before and after the leap second to "absorb" the leap
      second and get resynchronized to UTC. The other approach is to
      explicitly use TAI for internal reprsentation and use a leap second
      lookup table to convert between TAI and UTC. The TAI64 format I
      mentioned in the other thread is only one such implementation.

      At least three concerns around leap seconds seem to absorb a lot of
      energy from the discussion of using TAI instead of UTC:
      1) Up to date leap second information
      2) Qualifying time entry if it is the instance of leap second (does
      software accept second "60.")
      3) Date-times more than 6-12 months in the future do not have an
      absolute TAI representation
      because the number of leap seconds is uncertain.
      I think I've had a couple of insights that help this, but I haven't
      implemented anything yet. I'd like to see what others think of the
      ideas. Basically, they are discussed under two themes in more detail
      below.
      1) Leap second tables need explicit "horizons" included
      2) A unique, flagged representation is needed for UTC datetimestamps
      that can not yet be resolved to TAI.

      Leap Second Horizons

      Leap seconds are announced by IERS in Bulletin C about 5.5 months
      ahead. Leap seconds can only be introduced at UTC time 23:59:59 on
      preferred dates June 30 and Dec. 31 or secondary dates March 31 and
      Sep. 30. The secondary dates have never been used. Bulletin C also
      announces when a leap second is not going to happen. Besides the
      specific date it defines, it also assures a horizon of six months
      less a day that no leap seconds will happen. I believe that
      specifically including this horizon in leap second tables solves a
      couple of problems.

      Leap second tables are just lists of dates when they were added,
      usually represented by the second after leap second and the value of
      TAI - UTC that begins at that instant. At least two sources are
      available on the web for past leap seconds:
      ftp://maia.usno.navy.mil/ser7/tai-utc.dat
      http://inms-ienm.cnrc.gc.ca/timefreq/IERS.html
      Both could be machine parsable, but it is not clear whether a leap
      second is added soon after announcement via Bulletin C or only after
      it has occurred. Neither includes the "Horizon" entry I am
      recommending.


      I am recommending that leap second tables be updated every six months
      in response to Bulletin C and include as the last line the end of the
      horizon for which they assure no "surprise" leap seconds. The
      latest
      (2004-01-15) Bulletin C announced no change in TAI-UTC at 2004-07-01,
      and none can occur before 2005-01-01. At a high level, my proposal
      would look like:
      Leap Second Table (Updated 2004-01-15)
      YYYY-MM-DD TAI-UTC
      1972-01-01 10
      . . . . . . (intermediate entries omitted for brevity)
      1997-07-01 31
      1999-01-01 32
      2004-12-31 32
      The last line and the update date would not normally appear. The last
      line must be carefully interpreted, it only assures no leap second
      prior to 23:59:59 and is silent on the possibility of 23:59:60 on
      that date. However, the next Bulletin C around 2004-07-15 will
      resolve that. This horizon date allows a specific test of whether the
      table is up to date. Since Bulletin C is published about 5.5 months
      ahead, if there is less than 150 days to the end date, it should
      raise a warning flag to seek an update, If the date is past the end
      date, it should raise an error or fatal error. This would make
      maintenance a planned event, not an afterthought.. It would also
      resolve a user input as to second "60" and whether it is a
      leap
      second or user error as the table contains the latest data from
      Bulletin C. If the leap second isn't announced yet, then it may
      not
      exist and entry shouldn't be allowed.

      Flagged UTC Representation
      In the example above, any date in 2005 or beyond can not be converted
      from UTC to TAI because the number of leap seconds is unknown. If
      dates are normally converted to an "internal count" format,
      say TAI
      seconds from an epoch, this is a problem. If a flag field is
      available, it can simply be converted as a UTC internal count with
      flag set, and leap second correction applied later. Provided that
      approximate date manipulation can still be performed, about 150 days
      before the specified time, it can be converted to a proper TAI count
      representation and the flag cleared. The TAI64 format, which I
      mentioned before, reserves the 2^63 bit for future use. The format
      does not sanction this use, but one use would be to use it as the UTC
      flag and mark internal counts that are not yet corrected for TAI-UTC.
      In math, the flag could be tested, masked, and dates subtracted to
      see if they are in the 150 days (12,960,000 s) range, where
      resolvability should be guaranteed. The flag can also be used as a
      testable warning that a UTC date is implied but can not be properly
      expressed in TAI format. Other ways of implementing this could also
      be devised.
    • piebaldconsult
      Zzzzz... huh? wuh? Oh you stopped talking? I fail to see how that has anything to do with ISO 8601. ISO 8601 specifies formats for time data interchange; as
      Message 2 of 18 , Mar 17, 2004
      • 0 Attachment
        Zzzzz... huh? wuh? Oh you stopped talking?

        I fail to see how that has anything to do with ISO 8601.
        ISO 8601 specifies formats for time data interchange; as long
        as "23:59:60" is valid, then everything's fine.

        Taking leap seconds into account or not for timespans must be decided
        by the data interchanging parties "by mutual agreement".
      • John M. Steele
        Actually, it s invalid about 99.8% of the time, and it may be useful to know the few times it s valid. However, it is more concerned with internal
        Message 3 of 18 , Mar 17, 2004
        • 0 Attachment
          Actually, it's invalid about 99.8% of the time, and it may be useful to know the few times it's valid.
           
          However, it is more concerned with internal representation than external ISO8601 transmission of time info.
          -----Original Message-----
          From: piebaldconsult [mailto:PIEBALDconsult@...]
          Sent: Wednesday, March 17, 2004 12:06 PM
          To: ISO8601@yahoogroups.com
          Subject: [ISO8601] Re: Leap Seconds

           > as long   as "23:59:60" is valid, then everything's fine.
        • Tex Texin
          ... John is right that it is nevertheless a problem for software that allowing 61 seconds (0-60) because a minute may sometimes have a 61st, allows erroneous
          Message 4 of 18 , Mar 17, 2004
          • 0 Attachment
            piebaldconsult is skeptical:
            >
            > Zzzzz... huh? wuh? Oh you stopped talking?
            >
            > I fail to see how that has anything to do with ISO 8601.
            > ISO 8601 specifies formats for time data interchange; as long
            > as "23:59:60" is valid, then everything's fine.
            >
            > Taking leap seconds into account or not for timespans must be decided
            > by the data interchanging parties "by mutual agreement".

            John is right that it is nevertheless a problem for software that allowing 61
            seconds (0-60) because a minute may sometimes have a 61st, allows erroneous
            data to be created for the vast majority of minutes that shouldn't have leap
            seconds. A table of values or other approaches are needed to insure the
            "agreement" is followed.

            And since the agreement is changing, (leap seconds are added over time)
            mechanisms for authorizing and incorporating changes need to be supported.

            It's a difficult problem.
            Perhaps uninteresting to piebaldconsult, but it is in scope for date
            interchange discussion.
            (ie. Better buy more "z"s... ;-) )

            --
            -------------------------------------------------------------
            Tex Texin cell: +1 781 789 1898 mailto:Tex@...
            Xen Master http://www.i18nGuy.com

            XenCraft http://www.XenCraft.com
            Making e-Business Work Around the World
            -------------------------------------------------------------
          • piebaldconsult
            Oh, so you don t trust your business partners, I see. Well what would you do with a 23:59:60 that doesn t appear on your list? Send it back for confirmation?
            Message 5 of 18 , Mar 18, 2004
            • 0 Attachment
              Oh, so you don't trust your business partners, I see.

              Well what would you do with a "23:59:60" that doesn't appear on your
              list? Send it back for confirmation? You can't just say "Well they
              must have meant..." . And will you hold up your own production while
              you wait for confirmation?

              My agreement with a partner would specify that a "23:59:60" isn't
              questioned, just processed. If it's later found to be invalid, well
              shame on them. And if there's a possiblilty of abuse, then a review
              process would need to be established.
            • Tex Texin
              It s not so much a case of trusting people, as trusting their code... And yes, if data isn t valid it can be a reason to bounce a request for service of some
              Message 6 of 18 , Mar 18, 2004
              • 0 Attachment
                It's not so much a case of trusting people, as trusting their code...

                And yes, if data isn't valid it can be a reason to bounce a request for service
                of some kind. It doesn't necessarily hold up my production.

                Just curious, but what do you if other fields are invalid? 22:90:59...

                But it's just curiousity. With respect to dates and partner agreements, I hope
                we can just stipulate that applications take different approaches with invalid
                data, and how invalid leap seconds are processed might be covered by the
                agreement as well.
                tex

                piebaldconsult wrote:
                >
                > Oh, so you don't trust your business partners, I see.
                >
                > Well what would you do with a "23:59:60" that doesn't appear on your
                > list? Send it back for confirmation? You can't just say "Well they
                > must have meant..." . And will you hold up your own production while
                > you wait for confirmation?
                >
                > My agreement with a partner would specify that a "23:59:60" isn't
                > questioned, just processed. If it's later found to be invalid, well
                > shame on them. And if there's a possiblilty of abuse, then a review
                > process would need to be established.
                >
                >
                > Yahoo! Groups Links
                >
                >
                >
                >

                --
                -------------------------------------------------------------
                Tex Texin cell: +1 781 789 1898 mailto:Tex@...
                Xen Master http://www.i18nGuy.com

                XenCraft http://www.XenCraft.com
                Making e-Business Work Around the World
                -------------------------------------------------------------
              • BUDAI Andrew
                A note from a language stickler: see the second paragraph below. Budai, Andrew ... From: Tex Texin To: ISO8601@yahoogroups.com Sent: 2004 03 19, (Friday)
                Message 7 of 18 , Mar 18, 2004
                • 0 Attachment
                   A note from a language stickler: see the second paragraph below.  Budai, Andrew
                  ----- Original Message -----
                  From: Tex Texin
                  Sent: 2004 03 19, (Friday)  05:49
                  Subject: Re: [ISO8601] Re: Leap Seconds

                  It's not so much a case of trusting people, as trusting their code...

                  And yes, if data isn't valid 
                  (should read "if data aren't valid" as the singular of data is datum)
                  it can be a reason to bounce a request for service
                  of some kind. It doesn't necessarily hold up my production.

                  Just curious, but what do you if other fields are invalid? 22:90:59...

                  But it's just curiousity. With respect to dates and partner agreements, I hope
                  we can just stipulate that applications take different approaches with invalid
                  data, and how invalid leap seconds are processed might be covered by the
                  agreement as well.
                  tex

                  piebaldconsult wrote:
                  >
                  > Oh, so you don't trust your business partners, I see.
                  >
                  > Well what would you do with a "23:59:60" that doesn't appear on your
                  > list? Send it back for confirmation? You can't just say "Well they
                  > must have meant..." . And will you hold up your own production while
                  > you wait for confirmation?
                  >
                  > My agreement with a partner would specify that a "23:59:60" isn't
                  > questioned, just processed. If it's later found to be invalid, well
                  > shame on them. And if there's a possiblilty of abuse, then a review
                  > process would need to be established.
                  >
                  >
                  > Yahoo! Groups Links
                  >
                  >
                  >
                  >

                  --
                  -------------------------------------------------------------
                  Tex Texin   cell: +1 781 789 1898   mailto:Tex@...
                  Xen Master                          http://www.i18nGuy.com
                                          
                  XenCraft             http://www.XenCraft.com
                  Making e-Business Work Around the World
                  -------------------------------------------------------------
                • NGUYEN Adam
                  Good that you pointed that out, as I didn t know wether data is a singular or plural word.
                  Message 8 of 18 , Mar 18, 2004
                  • 0 Attachment
                             Good that you pointed that out, as I didn't know wether data is a singular or plural word.

                    At 2004-03-19 06:58 (UTC+0800), you wrote:

                     A note from a language stickler: see the second paragraph below.  Budai, Andrew

                    ----- Original Message -----
                    From: Tex Texin
                    To: ISO8601@yahoogroups.com
                    Sent: 2004 03 19, (Friday)  05:49
                    Subject: Re: [ISO8601] Re: Leap Seconds

                    It's not so much a case of trusting people, as trusting their code...

                    And yes, if data isn't valid 
                    (should read "if data aren't valid" as the singular of data is datum)
                    it can be a reason to bounce a request for service
                    of some kind. It doesn't necessarily hold up my production.

                    Just curious, but what do you if other fields are invalid? 22:90:59...

                    But it's just curiousity. With respect to dates and partner agreements, I hope
                    we can just stipulate that applications take different approaches with invalid
                    data, and how invalid leap seconds are processed might be covered by the
                    agreement as well.
                    tex
                  • Tex Texin
                    Andrew, thanks. If applications validated my English, I would never get any requests served... ;-) -- ... Tex Texin cell: +1 781 789 1898
                    Message 9 of 18 , Mar 18, 2004
                    • 0 Attachment
                      Andrew, thanks. If applications validated my English, I would never get any
                      requests served... ;-)



                      --
                      -------------------------------------------------------------
                      Tex Texin cell: +1 781 789 1898 mailto:Tex@...
                      Xen Master http://www.i18nGuy.com

                      XenCraft http://www.XenCraft.com
                      Making e-Business Work Around the World
                      -------------------------------------------------------------
                    • Tex Texin
                      Adam, you meant to use whether . A wether is a castrated ram. ;-) ... -- ... Tex Texin cell: +1 781 789 1898 mailto:Tex@XenCraft.com Xen Master
                      Message 10 of 18 , Mar 18, 2004
                      • 0 Attachment
                        Adam, you meant to use "whether". A "wether" is a castrated ram.

                        ;-)



                        NGUYEN Adam wrote:
                        >
                        > Good that you pointed that out, as I didn't know wether data is a
                        > singular or plural word.
                        >
                        > At 2004-03-19 06:58 (UTC+0800), you wrote:
                        >
                        > > A note from a language stickler: see the second paragraph below. Budai,
                        > > Andrew
                        > > ----------------------------------------------------------------------------
                        > >
                        > > ----- Original Message -----
                        > > From: Tex Texin
                        > > To: ISO8601@yahoogroups.com
                        > > Sent: 2004 03 19, (Friday) 05:49
                        > > Subject: Re: [ISO8601] Re: Leap Seconds
                        > >
                        > > It's not so much a case of trusting people, as trusting their code...
                        > >
                        > > And yes, if data isn't valid
                        > > (should read "if data aren't valid" as the singular of data is datum)
                        > > it can be a reason to bounce a request for service
                        > > of some kind. It doesn't necessarily hold up my production.
                        > >
                        > > Just curious, but what do you if other fields are invalid? 22:90:59...
                        > >
                        > > But it's just curiousity. With respect to dates and partner
                        > > agreements, I hope
                        > > we can just stipulate that applications take different approaches with
                        > > invalid
                        > > data, and how invalid leap seconds are processed might be covered by
                        > > the
                        > > agreement as well.
                        > > tex
                        > >
                        >
                        > ---------------------------------------------------------------
                        > Yahoo! Groups Links
                        >
                        > * To visit your group on the web, go to:
                        > http://groups.yahoo.com/group/ISO8601/
                        >
                        > * To unsubscribe from this group, send an email to:
                        > ISO8601-unsubscribe@yahoogroups.com
                        >
                        > * Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

                        --
                        -------------------------------------------------------------
                        Tex Texin cell: +1 781 789 1898 mailto:Tex@...
                        Xen Master http://www.i18nGuy.com

                        XenCraft http://www.XenCraft.com
                        Making e-Business Work Around the World
                        -------------------------------------------------------------
                      • John M. Steele
                        But as a standards group, we should point out that there are standards for language. For American English, it is Webster s dictionary. On data, mine say
                        Message 11 of 18 , Mar 18, 2004
                        • 0 Attachment
                          But as a standards group, we should point out that there are standards for language. For American English, it is Webster's dictionary.
                          On data, mine say "plural, but singular or plural in construction." (Actually, it says it all in abbreviations.) But either "data is" or "data are" is correct per Webster. The standard reference for British or International English is the Oxford, and it may not agree.
                          -----Original Message-----
                          From: BUDAI Andrew [mailto:bandi@...]
                          Sent: Thursday, March 18, 2004 05:58 PM
                          To: ISO8601@yahoogroups.com
                          Subject: Re: [ISO8601] Re: Leap Seconds

                           A note from a language stickler: see the second paragraph below.  Budai, Andrew
                          ----- Original Message -----
                          From: Tex Texin
                          Sent: 2004 03 19, (Friday)  05:49
                          Subject: Re: [ISO8601] Re: Leap Seconds

                          It's not so much a case of trusting people, as trusting their code...

                          And yes, if data isn't valid 
                          (should read "if data aren't valid" as the singular of data is datum)
                          it can be a reason to bounce a request for service
                          of some kind.
                        • NGUYEN Adam
                          ... Well, me and BUDAI Andrew, as well as countless others use British (International) English; while American English is restricted to a particular country,
                          Message 12 of 18 , Mar 18, 2004
                          • 0 Attachment
                            At 2004-03-18 19:57 (UTC-0500), you wrote:

                            But as a standards group, we should point out that there are standards for language. For American English, it is Webster's dictionary.
                            On data, mine say "plural, but singular or plural in construction." (Actually, it says it all in abbreviations.) But either "data is" or "data are" is correct per Webster. The standard reference for British or International English is the Oxford, and it may not agree.

                                     Well, me and BUDAI Andrew, as well as countless others use British (International) English; while American English is restricted to a particular country, and/or industry. It's roots are British English, however. I thought the word data has always been used in a plural form (like "This is good data."). I don't agree in American English's way of quoting, where sentence puncutation that's not a part of the actual content being quoted is put inside the quotation marks ("My dog is big," he said in American English and "My dog is big", he said in British (International) English), rather than where they should be: on the outside.

                                     I don't want to turn this to a flame war so, I'm just stating my personal opinion, which can differ, depending on someone else reading it.
                          • John M. Steele
                            We re a little off topic, and this should really go to the other group. The issue you raise is not unique to English. Differences exist between Brasilian and
                            Message 13 of 18 , Mar 18, 2004
                            • 0 Attachment
                              We're a little off topic, and this should really go to the other group. The issue you raise is not unique to English. Differences exist between Brasilian and Continental Portuguese, Spanish as spoken in Spain vs. Latin America, and versions of French. There is some tendency for Latin America, Japan, and perhaps some other Asian countries to use American rather than British English.
                               
                              While we're being picky, even in British English, wouldn't that be "BUDAI Andrew and I?" On American vs.. British English, it seems necessary to regard both as correct, to be aware of the differences, and for each party to consistently use the version they learned. Although I am familiar with most of the differences, I feel it would be an affectation for me to try to use British English.
                               
                              "How about those leapseconds?"
                              -----Original Message-----
                              From: NGUYEN Adam [mailto:adam917@...]
                              Sent: Thursday, March 18, 2004 08:59 PM
                              To: ISO8601@yahoogroups.com
                              Subject: RE: [ISO8601] Re: Leap Seconds

                              At 2004-03-18 19:57 (UTC-0500), you wrote:

                              But as a standards group, we should point out that there are standards for language. For American English, it is Webster's dictionary.
                              On data, mine say "plural, but singular or plural in construction." (Actually, it says it all in abbreviations.) But either "data is" or "data are" is correct per Webster. The standard reference for British or International English is the Oxford, and it may not agree.

                                       Well, me and BUDAI Andrew, as well as countless others use British (International) English; while American English is restricted to a particular country, and/or industry. It's roots are British English, however. I thought the word data has always been used in a plural form (like "This is good data."). I don't agree in American English's way of quoting, where sentence puncutation that's not a part of the actual content being quoted is put inside the quotation marks ("My dog is big," he said in American English and "My dog is big", he said in British (International) English), rather than where they should be: on the outside.

                                      
                            • Tex Texin
                              Sure. If a duration is a billion leap seconds, how long is it in the US and how long is it in the UK? (hint, they are not the same.) ;-) ... -- ... Tex Texin
                              Message 14 of 18 , Mar 18, 2004
                              • 0 Attachment
                                Sure. If a duration is a billion leap seconds, how long is it in the US and how
                                long is it in the UK?

                                (hint, they are not the same.)
                                ;-)

                                > "John M. Steele" wrote:
                                >
                                > We're a little off topic, and this should really go to the other group. The
                                > issue you raise is not unique to English. Differences exist between Brasilian
                                > and Continental Portuguese, Spanish as spoken in Spain vs. Latin America, and
                                > versions of French. There is some tendency for Latin America, Japan, and
                                > perhaps some other Asian countries to use American rather than British
                                > English.
                                >
                                > While we're being picky, even in British English, wouldn't that be "BUDAI
                                > Andrew and I?" On American vs.. British English, it seems necessary to regard
                                > both as correct, to be aware of the differences, and for each party to
                                > consistently use the version they learned. Although I am familiar with most
                                > of the differences, I feel it would be an affectation for me to try to use
                                > British English.
                                >
                                > "How about those leapseconds?"
                                >
                                > -----Original Message-----
                                > From: NGUYEN Adam [mailto:adam917@...]
                                > Sent: Thursday, March 18, 2004 08:59 PM
                                > To: ISO8601@yahoogroups.com
                                > Subject: RE: [ISO8601] Re: Leap Seconds
                                >
                                > At 2004-03-18 19:57 (UTC-0500), you wrote:
                                >
                                > > But as a standards group, we should point out that there are
                                > > standards for language. For American English, it is Webster's
                                > > dictionary.
                                > > On data, mine say "plural, but singular or plural in
                                > > construction." (Actually, it says it all in abbreviations.) But
                                > > either "data is" or "data are" is correct per Webster. The
                                > > standard reference for British or International English is the
                                > > Oxford, and it may not agree.
                                >
                                > Well, me and BUDAI Andrew, as well as countless others use
                                > British (International) English; while American English is
                                > restricted to a particular country, and/or industry. It's roots are
                                > British English, however. I thought the word data has always been
                                > used in a plural form (like "This is good data."). I don't agree in
                                > American English's way of quoting, where sentence puncutation
                                > that's not a part of the actual content being quoted is put inside
                                > the quotation marks ("My dog is big," he said in American English
                                > and "My dog is big", he said in British (International) English),
                                > rather than where they should be: on the outside.
                                >
                                >
                                >
                                > Yahoo! Groups Sponsor
                                > ADVERTISEMENT
                                > [click here]
                                >
                                >
                                > ---------------------------------------------------------------
                                > Yahoo! Groups Links
                                >
                                > * To visit your group on the web, go to:
                                > http://groups.yahoo.com/group/ISO8601/
                                >
                                > * To unsubscribe from this group, send an email to:
                                > ISO8601-unsubscribe@yahoogroups.com
                                >
                                > * Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

                                --
                                -------------------------------------------------------------
                                Tex Texin cell: +1 781 789 1898 mailto:Tex@...
                                Xen Master http://www.i18nGuy.com

                                XenCraft http://www.XenCraft.com
                                Making e-Business Work Around the World
                                -------------------------------------------------------------
                              • John M. Steele
                                10^9 (US), 10^12 (UK). What does it mean to table the topic? ... From: Tex Texin [mailto:tex@xencraft.com] Sent: Thursday, March 18, 2004 09:25 PM To:
                                Message 15 of 18 , Mar 18, 2004
                                • 0 Attachment
                                  10^9 (US), 10^12 (UK). What does it mean to table the topic?
                                  -----Original Message-----
                                  From: Tex Texin [mailto:tex@...]
                                  Sent: Thursday, March 18, 2004 09:25 PM
                                  To: ISO8601@yahoogroups.com
                                  Subject: Re: [ISO8601] Re: Leap Seconds

                                  Sure. If a duration is a billion leap seconds, how long is it in the US and how
                                  long is it in the UK?

                                  (hint, they are not the same.)
                                  ;-)

                                  > "John M. Steele" wrote:
                                  >
                                  > We're a little off topic, and this should really go to the other group. The
                                  > issue you raise is not unique to English. Differences exist between Brasilian
                                  > and Continental Portuguese, Spanish as spoken in Spain vs. Latin America, and
                                  > versions of French. There is some tendency for Latin America, Japan, and
                                  > perhaps some other Asian countries to use American rather than British
                                  > English.
                                  >
                                  > While we're being picky, even in British English, wouldn't that be "BUDAI
                                  > Andrew and I?" On American vs.. British English, it seems necessary to regard
                                  > both as correct, to be aware of the differences, and for each party to
                                  > consistently use the version they learned. Although I am familiar with most
                                  > of the differences, I feel it would be an affectation for me to try to use
                                  > British English.
                                  >
                                  > "How about those leapseconds?"
                                  >
                                  >      -----Original Message-----
                                  >      From: NGUYEN Adam [mailto:adam917@...]
                                  >      Sent: Thursday, March 18, 2004 08:59 PM
                                  >      To: ISO8601@yahoogroups.com
                                  >      Subject: RE: [ISO8601] Re: Leap Seconds
                                  >
                                  >      At 2004-03-18 19:57 (UTC-0500), you wrote:
                                  >
                                  >     >  But as a standards group, we should point out that there are
                                  >     >  standards for language. For American English, it is Webster's
                                  >     >  dictionary.
                                  >     >  On data, mine say "plural, but singular or plural in
                                  >     >  construction." (Actually, it says it all in abbreviations.) But
                                  >     >  either "data is" or "data are" is correct per Webster. The
                                  >     >  standard reference for British or International English is the
                                  >     >  Oxford, and it may not agree.
                                  >
                                  >              Well, me and BUDAI Andrew, as well as countless others use
                                  >      British (International) English; while American English is
                                  >      restricted to a particular country, and/or industry. It's roots are
                                  >      British English, however. I thought the word data has always been
                                  >      used in a plural form (like "This is good data."). I don't agree in
                                  >      American English's way of quoting, where sentence puncutation
                                  >      that's not a part of the actual content being quoted is put inside
                                  >      the quotation marks ("My dog is big," he said in American English
                                  >      and "My dog is big", he said in British (International) English),
                                  >      rather than where they should be: on the outside.
                                  >
                                  >
                                  >
                                  >                    Yahoo! Groups Sponsor
                                  >                        ADVERTISEMENT
                                  >                         [click here]
                                  >
                                  >
                                  >        ---------------------------------------------------------------
                                  > Yahoo! Groups Links
                                  >
                                  >    * To visit your group on the web, go to:
                                  >      http://groups.yahoo.com/group/ISO8601/
                                  >
                                  >    * To unsubscribe from this group, send an email to:
                                  >      ISO8601-unsubscribe@yahoogroups.com
                                  >
                                  >    * Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

                                  --
                                  -------------------------------------------------------------
                                  Tex Texin   cell: +1 781 789 1898   mailto:Tex@...
                                  Xen Master                          http://www.i18nGuy.com
                                                          
                                  XenCraft                        http://www.XenCraft.com
                                  Making e-Business Work Around the World
                                  -------------------------------------------------------------

                                • Tex Texin
                                  In the US table means postpone it for discussion some other time. in the UK discuss it now. But what does table have to do with leap seconds? ;-) Speaking of
                                  Message 16 of 18 , Mar 18, 2004
                                  • 0 Attachment
                                    In the US table means postpone it for discussion some other time. in the UK
                                    discuss it now.

                                    But what does table have to do with leap seconds? ;-)

                                    Speaking of leaping:

                                    I am wearing a jumper. What part(s) of me are covered?
                                    (Not too graphic a description and please no images.)




                                    > "John M. Steele" wrote:
                                    >
                                    > 10^9 (US), 10^12 (UK). What does it mean to table the topic?
                                    >
                                    > -----Original Message-----
                                    > From: Tex Texin [mailto:tex@...]
                                    > Sent: Thursday, March 18, 2004 09:25 PM
                                    > To: ISO8601@yahoogroups.com
                                    > Subject: Re: [ISO8601] Re: Leap Seconds
                                    >
                                    > Sure. If a duration is a billion leap seconds, how long is it in
                                    > the US and how
                                    > long is it in the UK?
                                    >
                                    > (hint, they are not the same.)
                                    > ;-)
                                    >
                                    > > "John M. Steele" wrote:
                                    > >
                                    > > We're a little off topic, and this should really go to the other
                                    > group. The
                                    > > issue you raise is not unique to English. Differences exist
                                    > between Brasilian
                                    > > and Continental Portuguese, Spanish as spoken in Spain vs. Latin
                                    > America, and
                                    > > versions of French. There is some tendency for Latin America,
                                    > Japan, and
                                    > > perhaps some other Asian countries to use American rather than
                                    > British
                                    > > English.
                                    > >
                                    > > While we're being picky, even in British English, wouldn't that
                                    > be "BUDAI
                                    > > Andrew and I?" On American vs.. British English, it seems
                                    > necessary to regard
                                    > > both as correct, to be aware of the differences, and for each
                                    > party to
                                    > > consistently use the version they learned. Although I am familiar
                                    > with most
                                    > > of the differences, I feel it would be an affectation for me to
                                    > try to use
                                    > > British English.
                                    > >
                                    > > "How about those leapseconds?"
                                    > >
                                    > > -----Original Message-----
                                    > > From: NGUYEN Adam [mailto:adam917@...]
                                    > > Sent: Thursday, March 18, 2004 08:59 PM
                                    > > To: ISO8601@yahoogroups.com
                                    > > Subject: RE: [ISO8601] Re: Leap Seconds
                                    > >
                                    > > At 2004-03-18 19:57 (UTC-0500), you wrote:
                                    > >
                                    > > > But as a standards group, we should point out that there
                                    > are
                                    > > > standards for language. For American English, it is
                                    > Webster's
                                    > > > dictionary.
                                    > > > On data, mine say "plural, but singular or plural in
                                    > > > construction." (Actually, it says it all in
                                    > abbreviations.) But
                                    > > > either "data is" or "data are" is correct per Webster. The
                                    > > > standard reference for British or International English is
                                    > the
                                    > > > Oxford, and it may not agree.
                                    > >
                                    > > Well, me and BUDAI Andrew, as well as countless
                                    > others use
                                    > > British (International) English; while American English is
                                    > > restricted to a particular country, and/or industry. It's
                                    > roots are
                                    > > British English, however. I thought the word data has always
                                    > been
                                    > > used in a plural form (like "This is good data."). I don't
                                    > agree in
                                    > > American English's way of quoting, where sentence
                                    > puncutation
                                    > > that's not a part of the actual content being quoted is put
                                    > inside
                                    > > the quotation marks ("My dog is big," he said in American
                                    > English
                                    > > and "My dog is big", he said in British (International)
                                    > English),
                                    > > rather than where they should be: on the outside.
                                    > >
                                    > >
                                    > >
                                    > > Yahoo! Groups Sponsor
                                    > > ADVERTISEMENT
                                    > > [click here]
                                    > >
                                    > >
                                    > >
                                    > ---------------------------------------------------------------
                                    > > Yahoo! Groups Links
                                    > >
                                    > > * To visit your group on the web, go to:
                                    > > http://groups.yahoo.com/group/ISO8601/
                                    > >
                                    > > * To unsubscribe from this group, send an email to:
                                    > > ISO8601-unsubscribe@yahoogroups.com
                                    > >
                                    > > * Your use of Yahoo! Groups is subject to the Yahoo! Terms of
                                    > Service.
                                    >
                                    > --
                                    > -------------------------------------------------------------
                                    > Tex Texin cell: +1 781 789 1898 mailto:Tex@...
                                    > Xen Master http://www.i18nGuy.com
                                    >
                                    > XenCraft http://www.XenCraft.com
                                    > Making e-Business Work Around the World
                                    > -------------------------------------------------------------
                                    >
                                    > ---------------------------------------------------------------
                                    > Yahoo! Groups Links
                                    >
                                    > * To visit your group on the web, go to:
                                    > http://groups.yahoo.com/group/ISO8601/
                                    >
                                    > * To unsubscribe from this group, send an email to:
                                    > ISO8601-unsubscribe@yahoogroups.com
                                    >
                                    > * Your use of Yahoo! Groups is subject to the Yahoo! Terms of
                                    > Service.

                                    --
                                    -------------------------------------------------------------
                                    Tex Texin cell: +1 781 789 1898 mailto:Tex@...
                                    Xen Master http://www.i18nGuy.com

                                    XenCraft http://www.XenCraft.com
                                    Making e-Business Work Around the World
                                    -------------------------------------------------------------
                                  • NGUYEN Adam
                                    ... Ooooh... How about those? How about 86400 of them? What would a time be if it had that many leap seconds in that day?
                                    Message 17 of 18 , Mar 18, 2004
                                    • 0 Attachment
                                      At 2004-03-18 21:16 (UTC-0500), you wrote:
                                      "How about those leapseconds?"

                                      Ooooh... How about those? How about 86400 of them? What would a time be if it had that many leap seconds in that day?
                                    • Tex Texin
                                      not sure about the time, but its a pretty good guess the date is feb 29. ;-)
                                      Message 18 of 18 , Mar 18, 2004
                                      • 0 Attachment
                                        not sure about the time, but its a pretty good guess the date is feb 29.
                                        ;-)

                                        NGUYEN Adam wrote:
                                        > > "How about those leapseconds?"
                                        >
                                        > Ooooh... How about those? How about 86400 of them? What would a time be if it
                                        > had that many leap seconds in that day?
                                      Your message has been successfully submitted and would be delivered to recipients shortly.