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

Re: [ISO8601] Re: Space as time designator

Expand Messages
  • Michael Deckers
    John Hynes wrote about the separator T betweeeen the date ... Right: a datetime notation such as 1981-04-05 14:30-05 is composed of _two_ ISO8601 notations:
    Message 1 of 29 , Jul 10, 2006
      John Hynes wrote about the separator 'T' betweeeen the date
      and time part of an ISO notation for datetime:

      > According to Wikipedia at http://en.wikipedia.org/wiki/ISO_8601 :
      >
      > The standard allows the replacement of T with a space or underscore
      > if no misunderstanding arises. This is commonly done for human
      > communications. A date/time with timezone like 1981-04-05T14:30-05
      > would then be written as 1981-04-05 14:30-05.
      >
      > .......................... However, according to ISO 8601:2000 sec4.4:
      >
      > The space character shall not be used in the representations.
      >
      > This does not allow for a space or any other character to be used.
      >
      > Has this been changed in :2004?

      To which question piebaldconsult answered:

      > Nope.

      Right: a datetime notation such as
      1981-04-05 14:30-05
      is composed of _two_ ISO8601 notations: one for date, one for time
      of day. ISO8601 sets rules for each of these tokens; it does not
      set rules on how these tokens may be combined in text. This is
      usually done with intervening white space or punctuation.

      Michael Deckers
    • piebaldconsult
      ... OK, I just read it, it s horrible. Many incorrect statements, too many to correct. There is no substitute for reading the actual standard. Perhaps we
      Message 2 of 29 , Jul 11, 2006
        > The article says nothing of the sort.

        OK, I just read it, it's horrible. Many incorrect statements, too many
        to correct.

        There is no substitute for reading the actual standard.

        Perhaps we should add an "external link" to here? There isn't even a
        link to the 2000 version.
      • Vincent Lefevre
        ... Yep, but this is not specific to the ISO 8601 standard. If you want to know what a standard says, read the standard. For instance, this is particularly
        Message 3 of 29 , Jul 11, 2006
          On 2006-07-11 16:06:53 -0000, piebaldconsult wrote:
          > OK, I just read it, it's horrible. Many incorrect statements, too many
          > to correct.
          >
          > There is no substitute for reading the actual standard.

          Yep, but this is not specific to the ISO 8601 standard. If you want
          to know what a standard says, read the standard. For instance, this
          is particularly true for the ISO C standard, for which the K&R book
          (latest edition) is inaccurate or completely wrong in several places.

          --
          Vincent Lefèvre <vincent@...> - Web: <http://www.vinc17.org/>
          100% accessible validated (X)HTML - Blog: <http://www.vinc17.org/blog/>
          Work: CR INRIA - computer arithmetic / SPACES project at LORIA
        • John Hynes
          ... I disagree that there are too many to correct. They are all correctable. The article isn t even that long, as Wikipedia goes. Perhaps if you listed the
          Message 4 of 29 , Jul 12, 2006
            --- In ISO8601@yahoogroups.com, "piebaldconsult" <PIEBALDconsult@...>
            wrote:
            > OK, I just read it, it's horrible. Many incorrect statements, too many
            > to correct.

            I disagree that there are too many to correct. They are all
            correctable. The article isn't even that long, as Wikipedia goes.
            Perhaps if you listed the biggest ones you found... What caught my
            attention right away was the large-print example at the top which had
            a space between the date and time. (It now has a "T".)

            > There is no substitute for reading the actual standard.
            >
            > Perhaps we should add an "external link" to here? There isn't even a
            > link to the 2000 version.

            It would be good if the current standard was online to link to. There
            have been significant changes since 2000.

            John Hynes
          • piebaldconsult
            ... The thing shouldn t even be as large as it is, it should give just a brief summary and provide links to the actual document. There is also a box farther
            Message 5 of 29 , Jul 12, 2006
              > I disagree that there are too many to correct. They are all
              > correctable. The article isn't even that long, as Wikipedia goes.
              > Perhaps if you listed the biggest ones you found...

              The thing shouldn't even be as large as it is, it should give just a
              brief summary and provide links to the actual document.

              There is also a box farther down that shows "<date>T<time>". When I
              first looked it also had "<date> <time>", but I saw this little "edit"
              link and darn well went ahead and fixed it myself!

              I just made a couple other edits, but I don't want to rewrite it. If it
              came to that I'd write a new entry just on the 2004 version.

              > It would be good if the current standard was online to link to. There
              > have been significant changes since 2000.

              Which is why ISO shouldn't charge for the darn things! (But there _is_
              a link to the ISO store.)
            • John Hynes
              ... Why not rewrite it? It happens all the time. It appears that you already have rewritten much of the article. I don t think that a new entry would be
              Message 6 of 29 , Jul 13, 2006
                --- In ISO8601@yahoogroups.com, "piebaldconsult" <PIEBALDconsult@...>
                wrote:
                > I just made a couple other edits, but I don't want to rewrite it. If it
                > came to that I'd write a new entry just on the 2004 version.

                Why not rewrite it? It happens all the time. It appears that you
                already have rewritten much of the article.

                I don't think that a new entry would be appropriate; the main article
                on ISO 8601 should focus on the current version. A history section
                could be added, stating the changes between editions, or just a
                section stating changes between 2000 and 2004.

                John Hynes
              • John Hynes
                ... or April 6, 2020. Since two-digit years are not permitted, the latter two are no longer valid. I have removed them from the article. John Hynes
                Message 7 of 29 , Jul 13, 2006
                  I just noticed one you missed. After your edit, the article states:
                  > ...200406 could either mean the year 200406; June 2004; April 6, 20;
                  or April 6, 2020.

                  Since two-digit years are not permitted, the latter two are no longer
                  valid. I have removed them from the article.

                  John Hynes

                  --- In ISO8601@yahoogroups.com, "piebaldconsult" <PIEBALDconsult@...>
                  wrote:
                  >
                  > > I disagree that there are too many to correct. They are all
                  > > correctable. The article isn't even that long, as Wikipedia goes.
                  > > Perhaps if you listed the biggest ones you found...
                  >
                  > The thing shouldn't even be as large as it is, it should give just a
                  > brief summary and provide links to the actual document.
                  >
                  > There is also a box farther down that shows "<date>T<time>". When I
                  > first looked it also had "<date> <time>", but I saw this little "edit"
                  > link and darn well went ahead and fixed it myself!
                  >
                  > I just made a couple other edits, but I don't want to rewrite it. If it
                  > came to that I'd write a new entry just on the 2004 version.
                  >
                  > > It would be good if the current standard was online to link to. There
                  > > have been significant changes since 2000.
                  >
                  > Which is why ISO shouldn't charge for the darn things! (But there _is_
                  > a link to the ISO store.)
                  >
                • Michael Deckers
                  On 2006-07-13, John Hynes wrote, first ... Nor is the former. The ISO 8601 notation 200406 can mean the time of day 20:04:06, but it cannot mean any date.
                  Message 8 of 29 , Jul 13, 2006
                    On 2006-07-13, John Hynes wrote, first
                    quoting "piebaldconsult" <PIEBALDconsult@...>:

                    > > ...200406 could either mean the year 200406; June 2004; April 6, 20;
                    > > or April 6, 2020.
                    >
                    > Since two-digit years are not permitted, the latter two are no longer
                    > valid. I have removed them from the article.

                    Nor is the former. The ISO 8601 notation 200406 can mean
                    the time of day 20:04:06, but it cannot mean any date.
                    Years with more than 4 digits need a sign (+200406),
                    and 2004-06 for April of 2004 is in basic format,
                    in both ISO 8601:2004 and in ISO8601:2000; the hyphens
                    cannot be omitted.

                    Michael Deckers
                  • piebaldconsult
                    ... Good, thanks.
                    Message 9 of 29 , Jul 13, 2006
                      > I just noticed one you missed.

                      Good, thanks.
                    • piebaldconsult
                      ... I suggest others in this group take a look as well, especially you in Texas.
                      Message 10 of 29 , Jul 13, 2006
                        > I have removed them from the article.

                        I suggest others in this group take a look as well, especially you in
                        Texas.
                      • piebaldconsult
                        ... It would be poor style to put _two_ values in one field, but yes, if you do it s OK. Just remember that the two values with the intervening space or
                        Message 11 of 29 , Jul 14, 2006
                          > Right: a datetime notation such as
                          > 1981-04-05 14:30-05
                          > is composed of _two_ ISO8601 notations: one for date, one for time
                          > of day. ISO8601 sets rules for each of these tokens; it does not
                          > set rules on how these tokens may be combined in text. This is
                          > usually done with intervening white space or punctuation.

                          It would be poor style to put _two_ values in one field, but yes, if
                          you do it's OK. Just remember that the two values with the intervening
                          space or whatever _do_not_ form one valid ISO 8601 value.
                        • piebaldconsult
                          ... As the section of the entry concerned dates I didn t bother mentioning the possibility of it being a time value. As to whether or not the year 200406
                          Message 12 of 29 , Jul 14, 2006
                            > Nor is the former. The ISO 8601 notation 200406 can mean
                            > the time of day 20:04:06, but it cannot mean any date.
                            > Years with more than 4 digits need a sign (+200406),
                            > and 2004-06 for April of 2004 is in basic format,
                            > in both ISO 8601:2004 and in ISO8601:2000; the hyphens
                            > cannot be omitted.

                            As the section of the entry concerned dates I didn't bother
                            mentioning the possibility of it being a time value.

                            As to whether or not the year 200406 requires a plus sign is up to
                            interpetation (as mentioned in my document on ambiguities). I left
                            the statement as it was, because it _could_ be the year 200406.

                            To reiterate my point, ISO 8601:2004 states in section 3.4.2
                            (Characters used in place of digits or signs):

                            "
                            (plus/minus) represents a plus sign [+] if in combination with the
                            following element a positive value or zero needs to be represented
                            (in this case, unless explicitly stated otherwise, the plus sign
                            shall not be omitted), or a minus sign [-] if in combination with the
                            following element a negative value needs to be presented.
                            "

                            I agree that this pretty much means, "if the year needs to be
                            represented and is positive, it requires a plus sign", but I would
                            allow others to interpret it to mean, "if the year needs to be
                            represented as positive, the plus sign is required". I expect most
                            people _don't_ think of the year as positive, but as an ordinal
                            number and doesn't need to be represented as positive. Otherwise, why
                            not require the plus sign on four-digit years?

                            Besides, the ambiguity only exists when basic format is used. If you
                            stick to extended format (as you should), there is no problem.
                          • John Hynes
                            ... Unfortunately, this is not what the article stated. It specifically stated that the T designator may be replaced by a space or underscore in combined
                            Message 13 of 29 , Jul 15, 2006
                              --- In ISO8601@yahoogroups.com, "piebaldconsult" <PIEBALDconsult@...>
                              wrote:
                              > > Right: a datetime notation such as
                              > > 1981-04-05 14:30-05
                              > > is composed of _two_ ISO8601 notations: one for date, one for time
                              > > of day. ISO8601 sets rules for each of these tokens; it does not
                              > > set rules on how these tokens may be combined in text. This is
                              > > usually done with intervening white space or punctuation.
                              >
                              > It would be poor style to put _two_ values in one field, but yes, if
                              > you do it's OK. Just remember that the two values with the intervening
                              > space or whatever _do_not_ form one valid ISO 8601 value.

                              Unfortunately, this is not what the article stated. It specifically
                              stated that the T designator may be replaced by a space or underscore
                              in combined representations, and gave examples of combined
                              representations with spaced throughout. (I just found a couple more
                              and fixed them.)

                              We could put language stating that the date and time may be
                              represented in two separate fields, but it should be clear that this
                              is not a single combined representation, and examples with spaces
                              should be avoided in the rest of the article.

                              John Hynes
                              July 15, 2006-07-15T00:17:00Z
                            • John Hynes
                              ... I hit send accidently, before I finished editing the date string. I ll have to make a script. John Hynes 2006-07-16T01:10:32Z
                              Message 14 of 29 , Jul 15, 2006
                                --- In ISO8601@yahoogroups.com, "John Hynes" <john@...> wrote:
                                > John Hynes
                                > July 15, 2006-07-15T00:17:00Z

                                I hit send accidently, before I finished editing the date string.
                                I'll have to make a script.

                                John Hynes
                                2006-07-16T01:10:32Z
                              • Pete Forman
                                ... In general there should be three parts: date, time, and zone designator. Leave aside the case of local time where the zone designator is empty. The ZD
                                Message 15 of 29 , Jul 17, 2006
                                  At 2006-07-16 00:16 +0000, John Hynes wrote:
                                  >[....]
                                  >We could put language stating that the date and time may be
                                  >represented in two separate fields, but it should be clear that this
                                  >is not a single combined representation, and examples with spaces
                                  >should be avoided in the rest of the article.

                                  In general there should be three parts: date, time, and zone
                                  designator. Leave aside the case of local time where the zone
                                  designator is empty. The ZD applies to the date and time. The three
                                  parts cannot be treated as separate. A space between date and time
                                  can be treated as an extension to ISO 8601 representing a single
                                  datetime but should not be treated as two ISO 8601 representations.


                                  A related standard of interest is for SQL, ISO/IEC 9075-2 (editions in
                                  1989, 1992, 1999, 2003). The timestamp format is very similar to ISO
                                  8601 extended but it mandates space as the separator between date and
                                  time.

                                  --
                                  Pete Forman -./\.- Disclaimer: This post is originated
                                  WesternGeco -./\.- by myself and does not represent
                                  pete.forman@... -./\.- the opinion of Schlumberger or
                                  http://petef.port5.com -./\.- WesternGeco.
                                • piebaldconsult
                                  ... Actually only with the time, which is somewhat of a shortcoming of ISO 8601. ... No, it can t. ... I was just trying to get in touch with one of my
                                  Message 16 of 29 , Jul 17, 2006
                                    > The ZD applies to the date and time.

                                    Actually only with the time, which is somewhat of a shortcoming of ISO
                                    8601.

                                    > A space between date and time
                                    > can be treated as an extension to ISO 8601

                                    No, it can't.

                                    > A related standard of interest is for SQL, ISO/IEC 9075-2 (editions in
                                    > 1989, 1992, 1999, 2003). The timestamp format is very similar to ISO
                                    > 8601 extended but it mandates space as the separator between date and
                                    > time.

                                    I was just trying to get in touch with one of my colleagues to ask
                                    where he got the following quote:

                                    "The wonderful thing about standards is that there are so many of them."
                                  • John Hynes
                                    ... This does not seem to be supported by the wording of the standard. The the zone designator is used only with time of the day representations or combined
                                    Message 17 of 29 , Jul 17, 2006
                                      --- In ISO8601@yahoogroups.com, Pete Forman <pete.forman@...> wrote:
                                      > In general there should be three parts: date, time, and zone
                                      > designator. Leave aside the case of local time where the zone
                                      > designator is empty. The ZD applies to the date and time. The three
                                      > parts cannot be treated as separate. A space between date and time
                                      > can be treated as an extension to ISO 8601 representing a single
                                      > datetime but should not be treated as two ISO 8601 representations.

                                      This does not seem to be supported by the wording of the standard.
                                      The the zone designator is used only with time of the day
                                      representations or combined representations. There is no zone
                                      designator with date-only representations. Also, spaces are
                                      explicitly prohibited, so if the date and time are separated by a
                                      space, then they are separate representations. The standard is very
                                      explicit that there are date repesentations, time representations and
                                      combined representations, and that the zone designator is a part of
                                      the latter two, not a separate representation.

                                      > A related standard of interest is for SQL, ISO/IEC 9075-2 (editions in
                                      > 1989, 1992, 1999, 2003). The timestamp format is very similar to ISO
                                      > 8601 extended but it mandates space as the separator between date and
                                      > time.

                                      Which is a different standard, so what it states does not apply
                                      necessarily to ISO 8601, which clearly states, "The space character
                                      shall not be used in the representations."
                                    • Tex Texin
                                      I agree with John. tex _____ From: ISO8601@yahoogroups.com [mailto:ISO8601@yahoogroups.com] On Behalf Of John Hynes Sent: Monday, July 17, 2006 2:44 PM To:
                                      Message 18 of 29 , Jul 18, 2006
                                        I agree with John.
                                        tex


                                        From: ISO8601@yahoogroups.com [mailto:ISO8601@yahoogroups.com] On Behalf Of John Hynes
                                        Sent: Monday, July 17, 2006 2:44 PM
                                        To: ISO8601@yahoogroups.com
                                        Subject: [ISO8601] Re: Space as time designator

                                        --- In ISO8601@yahoogroups .com, Pete Forman <pete.forman@ ...> wrote:

                                        > In general there should be
                                        three parts: date, time, and zone
                                        > designator. Leave aside the case of
                                        local time where the zone
                                        > designator is empty. The ZD applies to the
                                        date and time. The three
                                        > parts cannot be treated as separate. A space
                                        between date and time
                                        > can be treated as an extension to ISO 8601
                                        representing a single
                                        > datetime but should not be treated as two ISO 8601
                                        representations.

                                        This does not seem to be supported by the wording of the standard.
                                        The the zone designator is used only with time of the day
                                        representations or combined representations. There is no zone
                                        designator with date-only representations. Also, spaces are
                                        explicitly prohibited, so if the date and time are separated by a
                                        space, then they are separate representations. The standard is very
                                        explicit that there are date repesentations, time representations and
                                        combined representations, and that the zone designator is a part of
                                        the latter two, not a separate representation.

                                        > A related
                                        standard of interest is for SQL, ISO/IEC 9075-2 (editions in
                                        > 1989, 1992,
                                        1999, 2003). The timestamp format is very similar to ISO
                                        > 8601 extended
                                        but it mandates space as the separator between date and
                                        >
                                        time.

                                        Which is a different standard, so what it states does not apply
                                        necessarily to ISO 8601, which clearly states, "The space character
                                        shall not be used in the representations. "

                                      • piebaldconsult
                                        ... in ... ISO ... and ... Ah, here s the quote, I found it on http://www.sysprog.net/quotes.html The nice thing about standards is that there are so many of
                                        Message 19 of 29 , Jul 19, 2006
                                          > > A related standard of interest is for SQL, ISO/IEC 9075-2 (editions
                                          in
                                          > > 1989, 1992, 1999, 2003). The timestamp format is very similar to
                                          ISO
                                          > > 8601 extended but it mandates space as the separator between date
                                          and
                                          > > time.

                                          Ah, here's the quote, I found it on http://www.sysprog.net/quotes.html

                                          "
                                          The nice thing about standards is that there are so many of them to
                                          choose from. (Andrew Tannenbaum)
                                          "
                                        • Fred Bone
                                          ... Yes, it s in his Computer Networks . It continues something like And if you don t like what s on offer this year, just wait and there ll be more along
                                          Message 20 of 29 , Jul 19, 2006
                                            piebaldconsult said:

                                            > Ah, here's the quote, I found it on http://www.sysprog.net/quotes.html
                                            >
                                            > "
                                            > The nice thing about standards is that there are so many of them to
                                            > choose from. (Andrew Tannenbaum)
                                            > "

                                            Yes, it's in his "Computer Networks". It continues something like "And if
                                            you don't like what's on offer this year, just wait and there'll be more
                                            along next year"; except I can't find it to be sure of the words ...
                                          • Fred Bone
                                            ... Found it, of course, next time I went to the book: The nice thing about standards is that you have so many to choose from; furthermore, if you do not like
                                            Message 21 of 29 , Jul 19, 2006
                                              I said:

                                              > piebaldconsult said:
                                              >
                                              > > Ah, here's the quote, I found it on http://www.sysprog.net/quotes.html
                                              > >
                                              > > "
                                              > > The nice thing about standards is that there are so many of them to
                                              > > choose from. (Andrew Tannenbaum) "
                                              >
                                              > Yes, it's in his "Computer Networks". It continues something like "And if
                                              > you don't like what's on offer this year, just wait and there'll be more
                                              > along next year"; except I can't find it to be sure of the words ...

                                              Found it, of course, next time I went to the book:
                                              "The nice thing about standards is that you have so many to choose from;
                                              furthermore, if you do not like any of them, you can just wait for next
                                              year's model." - Andrew S Tanenbaum, "Computer Networks", ch. 4.

                                              (Note the spelling of his name, btw).

                                              He was talking about the development (if that's the word) of SNA into
                                              ADCCP, HDLC, LAP and LAP-B.
                                            • piebaldconsult
                                              Good, thanks. The point is that if you don t follow ISO 8601 (or any other standard) exactly then you simply aren t following the standard. Almost compliant
                                              Message 22 of 29 , Jul 19, 2006
                                                Good, thanks.

                                                The point is that if you don't follow ISO 8601 (or any other standard)
                                                exactly then you simply aren't following the standard. "Almost
                                                compliant" just won't cut it.

                                                See what the W3C did, they created their own standard which
                                                is "inspired" by ISO 8601. (And avoids a lot of trouble in the process.)

                                                ISO 8601 is a good standard by which to be inspired, but maybe the ISO
                                                can now be inspired by the W3C.
                                              Your message has been successfully submitted and would be delivered to recipients shortly.