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

148ISO8601 2000 and future revisions

Expand Messages
  • Thorvaldur Gunnlaugsson
    Feb 21, 2001
    • 0 Attachment
      In my attempts to present and explain the ISO8601 standard
      here in Iceland I have found most annoying the lack of a
      letter for weekday.
      I have had the response from this group that the weekday can
      be spelled out, but as English is not yet the universal
      standard language, this is not a solution for us.
      As W is used for the double digit week number, V could
      be used to precede the single digit weekday.
      The letter V is the initial for week in some languages.
      I have considered other possibilities e.g. S(semana) from
      Latin languages, U(uge) Danish and J or G for day in French
      and Italian, but think these are more confusing.
      It would ease the acceptance of the standard if we had a
      proposal of how to resolve this soon.

      I also have some other minor complaints about the publicly
      available version of the standard:
      1 - The ambiguity of using M for both Month and Minute
      certainly complicates parsing. If this was resolved
      by using upper/lower case letters or a different letter
      for month the T would not be needed in duration.
      2 - The use of W for both preceding the Week number and
      in duration to indicate a period of weeks also complicates
      parsing (most importantly for the human eye).
      If V was used for weekday the W could be phased
      out in the date, e.g. 2001-10V7 in stead of 2001W10-7.
      The W would then be used only in duration.
      The P preceding duration would then never be needed
      though it might be judged desireable to keep it.
      (there have been comments here about excluding redundancy)
      3 - When hour or minute is omitted the colon is replaced by
      a hyphen thereby causing possible confusion with negative
      local time designation. To me T:10 is clearer and looks
      better than T-10 for the tenth minute, then
      2001-01-01T-10 would be unambiguous as local time.
      Why would anyone want to specify day with only the local time
      deviation? For instance, if it is to be followed with
      a listing of several timings for that day.

      Some additional thoughts:

      On 2001-01-23 Ian said:
      >Repetitive events are usually just listed by their dates.
      Is SPACE/TAB/NL the prefered separator in a list of times?
      I assume that in a list of times the ommited components default
      to the latest specified?

      I have seen in air travel schedules when times are given
      for a given date but then a flight exceeds 24:00, then the time
      is given with a + appended which means the day after, e.g.
      - take off 22:30 - landing 01:44+
      This seems to be rather common and convenient but may be
      get confusing if there is also a local time deviation.

      The letter for local time is J. It looks to me logical
      to replace the T with Z or J accordingly:
      e.g. 2001-01-11Z12:48 = 2001-01-11J10:48-2

      I had a look at the revised 2000 version of the standard.
      The newly added part about repeated events did not look
      good at first sight, but it may be that I misunderstood
      something. The words recurrent and recursive seemd to be
      used interchangeably. The standard both has a R3/duration
      and duration#3 to say something will be repeated three times
      but I did not see anything about interval.
      The most common thing in a schedule is something which
      has a duration an then gets repeated at regular intervals
      To specify that an event starts on 2001-06-22T12:30 and
      lasts 1H and gets repeated every other week three times
      I would like to see something like:
      2001-06-22T12:30/1H3R2W or

      Does the groop have any feelings about this?
      Does it matter at all what is written here?

      Th.G. - thg@...
    • Show all 5 messages in this topic