[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Question: Week Number Semantics



Frank,

This is a complex issue.  For a inter-calendaring system operation I believe
that it will not be possible to use the week number as a scheduling value.
As you point out the definition of a week varies considerably, and may in
some instances have more than seven days.  At a manufacturing company that I
worked for some twenty years ago we used a 50 week year for planning
purposes.  The factory was shut for two weeks in the summer, these calendar
weeks were not included in the calendar, or rather the work week prior to
that week was extended to last three weeks, only five of which were working
days, more interestingly though the two weeks over Easter were combined into
an eight working day week.  As this was a UK based company the Good Friday
and Easter Monday were public Holidays.
If you use the week number for scheduling outside of the calendar system
that you are running then the structure of the weeks in the year will also
have to be communicated, this will be complex to say the least.  A simpler
solution is probably to require the originating system to convert the week
number to a specific date and then schedule normally.  This way there is no
confusion as to what I mean by the Monday of the 23rd week.
The concept of what is a week becomes even more interesting when you start
to deal with individuals working on different week concepts. Working with
the middle east you find that Thursday and Friday are the weekend, or
non-working days for Muslim countries and Friday and Saturday for the Jewish
world.  Are we to expect that their working week occupies two of our weeks?
In the process industries the Shift rotation period is more important than
the calendar week.  Shift rotation periods are usually consistent in length,
but rarely 7 days long, some are shorter and some are longer and may have
more than two non-working days.
To handle these concepts properly we need to look at scheduling periods,
which are usually of a fixed length, may have exceptions, and are locally
understood.  In this scheme the 'Week' is simply a well understood
scheduling period of seven days with possibly one or more non-working days.
Now what about the third Thursday of the second month, whose calendar are we
working with?, Gregorian, Lunar or Chinese.  The days of expecting every one
to be working in terms of the Western Calendar are rapidly coming to an end
with the rapid adoption of computer systems out-side of the Western world.
The challenge is to permit the user to work in the calendar-space that they
are used to and to inter-operate with any one else any where on the globe.

 Chris.
 
 *****************
 Christopher A Lowde
 Texaco
 4800 Fournace Pl.
 Bellaire, TX 77401-2324
 Tel: +1(713)432-2901
 Fax: +1(713)838-4529
 mailto:lowdeca@xxxxxxxxxx
 


-----Original Message-----
From: Frank Dawson [mailto:fdawson@xxxxxxxxxxxxx]
Sent: Tuesday, December 15, 1998 12:21
To: IETF Calendaring & Scheduling Mail List
Cc: vCalendar Mailing List
Subject: Question: Week Number Semantics


Folks:

RFC 2445 references the concept of "Week Number" in the "BYWEEKNO"
recurrence grammar. In addition, ISO 8601 references the concept of "Week
Number". It looks like the views of ISO 8601 are a subset of those in RFC
2445. Does everyone agree with the implications of the extended view taken
by RFC 2445 as it relates to how a CUA should display "Week Number" in a
"Weekly View" UI?

RFC 2445 indicates that:

   The BYWEEKNO rule part specifies a COMMA character (US-ASCII decimal
   44) separated list of ordinals specifying weeks of the year. Valid
   values are 1 to 53 or -53 to -1. This corresponds to weeks according
   to week numbering as defined in [ISO 8601]. A week is defined as a
   seven day period, starting on the day of the week defined to be the
   week start (see WKST). Week number one of the calendar year is the
   first week which contains at least four (4) days in that calendar
   year. This rule part is only valid for YEARLY rules. For example, 3
   represents the third week of the year.

ISO 8601 indicates that:

   3.17  week, calendar: A seven day period within a calendar year,
   starting on a Monday and identified by its ordinal number within
   the year; the first calendar week of the year is the one that
   includes the first Thursday of that year. In the Gregorian
   calendar, this is equivalent to the week which includes 4 January.

Taken together, it appears that ISO 8601 assumes that for the purpose of
calculating the "week number", the week _always_ starts on a Monday. On the
other hand, RFC 2445 clearly allows for a week to start on any day of the
week. Thus, a RFC 2445 "week number" might vary by "start of the week".

The problem that I see is that a CUA's UI may wish to display "Week Number"
in a "Week View" of the calendar. Is "Week Number" a canonical calendar
concept that _always_ must assume Monday as the start of the week? For
example, as is often used across multinational manufacturing corporations or
equivalent "teaming" relationships; where a corporation's factory calendar
is computed, based on weeks of the year. Or is this a calendar concept that
is sensitive to the CUA's local definition of "start of the week" and can
yield different presentations?

It would appear that if CUAs from different vendors are to be iCalendar
"standards" based they should behave the same. Is this right?

If so, then should we view "Week Number" as a non-permutable concept? Or
should the display of "Week Number" vary as the "start of the week" varies?

So end-user feedback on this point would be helpful.

-- Frank Dawson