Specification: DTV v1.0

Title: 'date time' associations

Source: Ed Barkmeyer (for the RTF), NIST, edbark@nist gov

Figure 10.17 has the wrong association names on 'date time' has 'date' and 'date time' has 'time'.

Should be: 'date time combines calendar date' and 'date time combines time of day coordinate'. And 'time' should be 'time of day coordinate'.

The business term 'date' should be a synonym for 'calendar date'.

We already have "of" as a synonymous form for 'combines' in 10.6.3. But the SynonymousForm 'compound time coordinate of atomic time coordinate' should be 'compound time coordinate has atomic time coordinate'.

SBVR experts:

In the Date Time Vocabulary v1.0 specification, clause 16.5 contains the following terminological entry:

individual situation kind has occurrence interval

Definition: the occurrence interval is the time span of the individual situation kind

Necessity: The individual situation kind has exactly one occurrence

Note: The time span of an individual situation kind is exactly the occurrence interval of its only occurrence.

Example: The occurrence interval of the Great Fire of London was 2 September 1666 through 5 September 1666 (English old style calendar).

The technical aspects of this are not at issue. An ‘individual situation kind’ (state of affairs) can have at most one occurrence. An occurrence (actuality) has exactly one ‘occurrence interval’ – the time interval during which it is occurring. And any situation kind can have a ‘time span’ – the smallest time interval that includes all occurrences of the situation kind. (If it never occurs, it may have No time span.)

There are two SBVR “style” problems here.

First, the Definition of the verb concept is (appropriately) a sentence involving the placeholders, but the first character of the sentence is not capitalized. By convention in SBVR, the first character of a Definition is not capitalized, and this is also true of example sentences in SBVR v1.2, e.g., in clause 8.2.2. The question is: Which convention – English or SBVR – is appropriate here? Or do we even care?

Second, and more importantly, the intent of the above Necessity is: If an individual situation kind has an occurrence interval, the individual situation kind has exactly one occurrence. But the text above omits the antecedent. Instead, it assumes that ‘the individual situation kind’ refers to whatever plays the ‘individual situation kind’ role in an instance of the verb concept (wording). Is it the intent of SBVR that such an omission is valid in this context (the terminological entry)? Or is the antecedent required (to establish the context of an instance of the verb concept)?

The considered opinion of the SBVR RTF will dictate the nature of any related changes to this entry in the DTV.

No Data Available

In clause 9.0 of the Date-Time Vocabulary, the vocabulary entry for the "Duration Values Vocabulary" has "Included Vocabulary: Duration Values Vocabulary". Per diagram 7.3, it should be "Included Vocabulary: Time Infrastructure Vocabulary".

sduplicate of issue # 18950

REFERENCXE:
section 11.7, 1st Definition on p 141:
" In mathematical form, the definition
above is:
sd = 685 263 + (365 * y)
+ (y/4) - ((y/100)*2)
+ ((y/400) * 2)
where:
sd is the index of the starting day
y is the index of a Gregorian
year ­ 1601
y >= zero
/ is integer division "

PROBLEM:
The formula is incorrect. For instance,
for years 1700 and 1701 the formula gives
sd(1700) = 685 263 + (365*99)
+ (99/4) - ((99/100)*2)
+ ((99/400)*2)
= 685 263 + 365*99 + 24
= 721 422
sd(1701) = 685 263 + (365·100)
+ (100/4) - ((100/100)·2)
+ ((100/400)·2)
= 685 263 + 365*100 + 25 - 2
= 721 786
implying that the year 1700 had only
364 d, which is obviously incorrect.
The formula gives the wrong year
length of 364 d for all Gregorian
years Y wher Y mod 400 is one
of 100, 200, 300.

PROPOSED CORRECTION:
Omit the incorrect factor 2 twice, and
correct the corresponding note trying
to justify that factor. Many sources
give the correct formula, valid for all
integral year numbers 1601 + y:

sd(1601 + y)
= sd(1601) + (365 * y)
+ floor(y/4) - floor(y/100)
+ floor(y/400)
where:
y is the index of the Gregorian year
with number (1601 + y)
sd(1601 + y) is the index of the
first day of the Gregorian
year (1601 + y)
floor is the largest integer
that is <= x

The formula was incorrectly represented. It will be corrected as suggested. The changes below incorporate several changes from Issue 19280 that correct errors resulting from faulty calculation of year lengths and incorrect base values. Two Examples of coordinate computations are revised to use the corrected formula.
The formula for the index origin of the Gregorian months scale is also incorrectly written.

REFERENCES:
section 11.2, last Note on p 126:

[a] "The calendar reform instituted by
Pope Gregory XIII [Inter Gravissimas]
deleted 10 days from the
Gregorian calendar, starting on
5 October 1582."

[b] "The previous calendar day had
index 577 739 on the Julian calendar,
computed as 1581 years of 365 days
plus 395 leap days + 279 days
from 1 January 1852 to
5 October 1582."

[c] "The following day was
15 October 1582."

[d] "From that date to the Convention du
Mètre on 20 May 1875, there were
106 870 calendar days including
leap days."

section 11.7, in a Note on p 141:

[e] "685 263 is the index of
1 January 1601,
computed as 684 609 (index of
15 October 1582) plus 6654 days from
15 October 1582 through
1 January 1601."

section 11.7, in the 2nd Example on p 141:

[f] "The first calendar day of 2010
is Gregorian day 830 991"

PROBLEMS:
The assertion in [a] is misunderstandable:
no days were deleted from the Gregorian
calendar by any pope. In fact, the
proleptic Gregorian calendar without
any "deletions" is
used by ISO 8601:2004, by computer
software and by some historians.
What really is meant is:
'The calendar reform instituted by
Pope Gregory XIII and promulgated in
the bull [Inter Gravissimas] started
the use of the Gregorian calendar with
the date 15 October 1582, which is the
same as 05 October 1582 in the Julian
calendar.'

As for [b]: The "previous calendar day"
would be 1582 Oct 04 in the Julian
calendar, not 1582 Oct 05 as suggested in
[b]. Whichever of the two is meant, the
following computation in [b] is
incorrect: the number of days from
1582 Jan 01 until 1582 Oct 05 is 277 and
not 279, as can be read directly from
[table 11.3, p 146].

As for [c]: It appears as if the
"following day" means the day
after J1582-10-05. This
following day is
J1582-10-06 = G1582-10-16,
not G1582-10-15 as asserted.
(We use prefixes
J and G to distinguish Julian and
Gregorian calendars.)

The computation in [d] is incorrect: the
number of days from G1582-10-15 to
G1875-05-20 is 106 868, not 106 870
as asserted,
because G1582-10-15 = JD 2299 160.5
and G1875-05-20 = JD 2406 028.5.

Assertion [e] is clearly
self-contradictory since
685 263 - 684 609 is not 6654.

Inconsistencies between [b, e, f]: If
1 January 1601 is the day number 685 263
as asserted in [e] then
day 0 is JD 1620 550.5 = G-0276-10-25,
and if day 684 609 is G1582-10-15 as
also asserted in [e] then
day 0 is JD 1614 551.5 = G-0292-05-23.
Still another zero point follows from [f]:
day 0 is JD 1624 206.5 = G-0266-10-29
and finally, if [b] is meant to refer to
G1582-10-14 as day number 577 739, then
day 0 is JD 1721 420.5 = G0000-12-27.
That last date might indicate that day 0
was actually meant to be still another
date, viz J0001-01-01 = G0000-12-30,
but this is just a wild guess of mine.

ANALYSIS OF THE ISSUES:
The inconsistencies may come from
several sources:
• from typos (though I am not able to
figure out which);
• from the error in the formula for
the duration of successive Gregorian
years (eg in [d]);
• from the use of intervals on the
"time axis" instead of points on it.

The time axis is an affine space whose
translation space, formed by the
differences T - T' for points T, T' on
the time axis, is the vector space of
"duration values". Intervals of
lengths > 0 s, however, do not form an
affine space. So one has to take the
lower or the upper bounds of the involved
intervals consistently to arrive at valid
date arithmetic (eg, satisfying the rule
T - T" = (T - T') + (T'- T")).

An error in numerical examples for
non-trivial specifications is particularly
unfortunate because such examples are
often taken as the very first test cases
for an implementation. It is therefore
appropriate to check all the examples
before publishing. Many calendrical
calculators are available online for that
purpose; the one at [http://emr.cs.iit.edu
/home/reingold
/calendar-book/Calendrica.html]
is particularly useful.

The suggested text clarification is accepted. The erroneous numbers will be corrected.
Because the changes made necessary by this resolution overlap those made in the resolution of Issue 19281, this issue is merged with Issue 19281.
Disposition: See issue 19281 for disposition

I've noticed a number of problems with reference schemes in DTV:

1) In clause 8.3, the reference scheme for 'duration' has a forward reference to the verb concept 'precise atomic duration value quantifies duration', which is in clause 9.2.2.
2) In clause 8.5, the first reference scheme for 'time point', which reads 'an occurrence at the time point', has a forward reference to the synonymous form 'occurrence at time interval', which is in clause 16.2.
3) In clause 8.5, the second reference scheme for 'time point', which reads 'a time coordinate that indicates the time point', has a forward reference to the verb concept 'time coordinate indicates time point', which is in clause 10.5.1.
4) In clause 8.5, the third reference scheme for 'time point', which reads 'the time scale of the time point and the index of the time point', makes no sense, since a time point can belong to multiple time scales (e.g. 'time of day'). Suggested form: a time scale that has the time point and the
5) In clause 8.5, the fourth reference scheme for 'time point', which reads 'a time point kind and an index', has a forward reference to the general concept 'time point kinds', which is in clause 10.3. Moreover, the reference scheme is not defined properly; it should read "a time point kind and an index of the time point'.
6) In clause 8.5, the fifth reference scheme for 'time point', which reads 'the name of the time point', refers to a verb concept 'time point has name' that does not exist. Perhaps it is intended as 'a representation of the time point', but considering that a 'time coordinate' is a representation of a time point, it appears that this fifth reference scheme duplicates the second reference scheme, which read 'a time coordinate that indicates the time point'.
7) In clause 10.1, the reference scheme for 'calendar', which reads 'the time scales that are defined by a calendar', refers to a missing synonymous form of 'calendar defines time scale'. Proposed solution: define this synonymous form.
8) In clause 10.4, the reference scheme for 'local calendar', which reads "a time offset by which the local calendar's day of hours-scale difference from the day-of-hours scale of UTC", has multiple problems: (a) the words 'by which' should be verb-styled; (b) apparently this is trying to use a synonymous form of 'calendar1 differs from calendar2 by offset' but the reference scheme talks about scales of calendar, whereas the verb concept is about calendars; (c) there is no such synonymous form. Proposed solution: define synonymous form 'time offset of calendar1 from calendar2' of existing verb concept 'calendar1 differs from calendar2 by time offset', and reword the reference scheme as "the time offset of the local calendar from UTC".
9) In Annex D.3, the reference scheme for 'particular quantity', which reads 'A definite description of the particular quantity', depends upon a noun concept 'definite description' that is not listed in Clause 4. Proposed solution: Change the reference scheme to 'a definite description that represents the particular quantity'. Add 'definite description' to clause 4, per SBVR, as a kind of 'intensional definition', which should be updated to define it as a kind of 'definition'. And 'definition' should be added as a kind of 'representation'. Also, make sure that clause 4 defines 'concept' as a kind of 'meaning' and 'meaning' as a kind of 'thing'.

Entries 1, 2, 3: Forward references are made necessary because there is only one glossary entry for the concept. There are similar examples in SBVR itself.
4. Clause 8.6, in the entry for ‘time scale has time point’, makes a time point a part of exactly one time scale.; ‘time of day’ and ‘calendar day’ are time point kinds, not time points.
5. The cited reference scheme is invalid – a time point kind does not necessarily identify a scale, and thus the index is ambiguous. Consider ‘calendar day’ and 1.
6. The redundant reference scheme will be deleted. (It is a vestige of a draft in which time coordinates did not include terms for time points.)
7. The missing synonymous form will be defined.
8. The suggested synonymous (noun) form will be defined.
9. It is sufficient to adopt ‘definite description’ and ‘concept has definition’ from SBVR. (Note: intensional definition is already present)

In DTV clause 8.5, the second Necessity of 'time scale has time point' reads 'Each time point is of exactly one time scale.' This is clearly wrong. Several of the time points defined in clause 10.2 are on multiple time scales. For example, 'time of day' is defined as a 'time point that is on a time scale that has a granularity that is less than 1 day'. Examples include 'second of day' and 'second of hour'.

Also in clause 8.5, the verb concept 'time point has index' makes no sense. A time point has different indices, depending upon what scale is used. This should be a ternary verb concept: 'time point has index on time scale'.

Numerous Definitions and Necessities depend upon a verb concept 'time point is on time scale' that is not defined anywhere but could be a Synonymous Form of 'time scale has time point'. Examples include:

• clause 10.2, definition of 'time of day' (see above)
• clause 11.2, most definitions
• clause 13.2, most definitions

There are no time points in 10.2 (or any other clause) that are on multiple time scales. ‘time of day’ is not a time point; it is a ‘time point kind’ – a general category of time points. Each ‘second of day’ time point is only on the day of seconds time scale, and each ‘second of hour’ is only on the ‘hour of seconds time scale’. The referenced Necessity in 8.5 is valid, and is important. As a consequence of it, the concept ‘index of time point’ is well-defined.
The form ‘time point is on time scale’ is used as described, and it should be a synonymous form for ‘time scale has time point’.

Specification: DTV v1.0
Title: incorrect reference schemes for time point sequence

Summary:
In clause 8.7, the last two reference schemes for a time point sequence are incomplete, because they are misworded. They read:
Reference Scheme: The first time point of the time point sequence, if the time point sequence has no last time point.
Reference Scheme: The last time point of the time point sequence, if the time point sequence has no first time point.

The problem is that the first time point of the time point sequence ALONE is not sufficient, whether the time point sequence has a last time point or not. When it has no last point, the fact that it has no last point must be a characteristic that is included in the reference scheme.
The Reference Scheme should be worded:
The first time point of the time point sequence and the characteristic ' the time point sequence has no last time point'.
and similarly for the sequence that has no first time point.

The reference schemes will be reworded as suggested.

In DTV Clause 8.2.3, in the entry for 'time interval1 equals time interval2', the second Note says:
"SBVR uses the verb is for this relationship, but the equals relationship here is a specialization of 'thing is thing' for time intervals. Two time intervals are equal if they share particular properties of time interval, and the definition of equal does not involve properties that are suitable for a reference scheme."

This relationship is important to business usage of the vocabulary. The SBVR formal mechanism for stating the first sentence is:
General concept: 'thing1 is thing2'
which makes the specialization clear. If two time intervals are equal, they are identical.
One could also state it as a Necessity (with translations to CLIF and OCL):
A time interval1 equals a time interval2 if and only if time interval1 is time interval2.
which is actually a stronger (and correct) statement. It adds: if two time intervals are identical, they are necessarily equal.

The second sentence of the Note is at best confusing, since reference schemes have nothing to do with equality. It should be deleted.

Resolution:
This issue was discovered in an RTF discussion of another issue. The meeting agreed to the General concept and Necessity.

In clause 11.2, the definition of the Gregorian year scale is:

“the indefinite time scale of granularity ‘year’ and of ‘Gregorian year’ time points”

The definitions of the Gregorian months scale and the Gregorian days scale use the same pattern.

This pattern relies on two verb concepts: ‘time scale of granularity’ (aka ‘granularity has time scale’) and ‘time scale of time point’ (aka ‘time point has time scale’). Neither of these verb concepts exists in Clause 8. The intended concepts are ‘time scale has granularity’ and ‘time scale has time point’.

The Definitions should be reworded:

“the indefinite time scale that has granularity ‘year’ and that has ‘Gregorian year’ time points”

and similarly for the others.

Agreed. The intended verb concept wordings will be used.

Figure 10.2 includes ‘calendar week’, which is not defined in clause 10. Similarly, Figure 10.3 includes ‘week period’ and ‘hour period’ , which are not defined in Clause 10. None of these elements is in the Calendars package or in any package it imports. They are defined in clauses 12 and 13. These concepts should not appear in Clause 10, or at least not without a Note to the Figures.

These diagram elements were erroneously retained in the Figures when the DTV elements were re-packaged. They will be removed.

OMG Specification: Date Time Vocabulary

Version: 1.0

Title: The definitions of 'starts before' and 'finishes after' are too complex

Summary:

In clause 8.2.4, the definition of ‘time interval1 starts before time interval2’ enumerates several possible Allen relations, but what is intended can be stated more simply: There is a time interval3 that is part of time interval1 and precedes time interval2.

Similarly ‘time interval1 ends after time interval2’ is: There is a time interval3 that is part of time interval1 and follows time interval2.

Time interval1 begins time interval2 seems to be misdefined. The example says 2009 begins 2010 and that is false. The intent of this verb concept should be clarified. Similarly, in Time interval1 ends time interval2, the example that is mislabeled Definition is the only example that a business person would recognize as T1 ends T2 and that is an instance of the Allen relation: T1 finishes T2.

The Definitions of ‘time interval1 starts before time interval2’ and ‘time interval1 finishes after time interval2’ are incorrect. The definition of ‘starts before’ omits the case where time interval2 finishes time interval1, and similarly ‘finishes before’ omits the case where time interval2 starts time interval1. The definitions will be corrected as suggested.
The example “2009 begins 2010” is not valid and will be removed.

OMG Specification: Date Time Vocabulary

Version: 1.0

Summary:

In clause 8.3.3, Figure 8.12 has no relationship to the terms and concepts in the clause. (It is a duplicate of Figure 8.8)

Replace it with a proper diagram of the concepts in the clause.

Replace the Figure with the correct UML diagram. Issue 18253 modifies the text of this section and modifies the diagram accordingly. This issue is merged with Issue 18253.
Disposition: See issue 18253 for disposition

There are two problems with the Necessity in the entry for "time table" in clause 17.2:

1) The Necessity assumes a verb concept "time table has table entry" that does not exist.
2) The Necessity is misspelled as "Each time table must have at least one table entry." It should be "Each time table has at least one table entry. "

Note: the definition of "time table has table entry" should make clear the relationship of this concept to the "table entries" mentioned in the definition of "time table" as a "set of table entries".

The UML diagram (Figure 17.1) disagrees with the definition of time table, and introduces the missing verb concept (time table has table entry) as a specialization of ‘set has element’ (from SBVR). The UML model reflects the intent. The text will be revised to support the model.

The formula given for the possible lengths
for an integral number of successive
Gregorian years is incorrect. It implies
an average length of the year of 365.235 d
while the correct value is 365.2425 d.
For instance it is well known that any
successive 400 Gregorian years comprise
exactly 146 097 d, while the formula
gives 146 094 d.
For the record, with a little algebra,
the number of days between Gregorian
calendar(Y', Jan, 01) and Gregorian
calendar(Y, Jan, 01) is easily seen
to be
floor( Δ·365.25 )
+ floor(-3/4·floor(Δ/100))
+ |Δ mod 4 > (Y - 1)mod 4|

• Δ mod 100 > (Y - 1)mod 100

  + |Δ mod 400 > (Y - 1)mod 400|
  where Δ = Y - Y'.

The reference is to the formula in the entry for ‘year value specifies duration value set’. The formula includes two erroneous factors of 2, and does not account for year values greater than 400 properly. The formula and the supporting concepts will be corrected.

Clause 15.3 has numerous entries that are “Concept Type: unitary concept”, but “unitary concept” is not defined. This concept should be added to clause 4.

Agreed. The adopted SBVR concept will be added to clause 4.

OMG Specification: Date Time Vocabulary

Version: 1.0

Title: Errors in the axioms for ‘time interval1 is before time interval2’

Summary:

In clause 8.2.2, the axioms for ‘time interval1 is before time interval2’ have several problems.

1. In the first axiom, the second Corollary does not follow from the Axiom. if t1 overlaps t2, it is not before or after t2. So: if t1 is before or after t2, it does not overlap t2. But the so-called Corollary (the Axiom of totality) states the converse: if t1 does not overlap t2, then t1 is before or after t2.

2. The SBVR SE statement of irreflexivity demonstrates a disconnect between Structured English and English. Surely the Axiom can be more clearly stated: No time interval is before itself. If the bizarre formulation is required for SBVR, the English expression should be given first.

3. The current “totality” corollary to the Axiom of assymetry is actually just a partition that follows from the assymetry Axiom and the mislabeled axiom above.

Make the second Corollary to the first axiom an axiom in its own right. Add the English version of the irreflexivity axiom as the text of the Axiom proper. There is no problem with the “totality” corollary.

In Annex D.3.3, in the entry for "quantity value quantifies quantity", the three synonymous forms are captioned "Synonym:" instead of "Synonymous Form:".

The caption “Synonymous Form” was erroneously rendered as “Synonym” in a number of places in the specification. All of them are corrected.

In clause 10.6.3, in the entry for 'compound time coordinate combines atomic time coordinate', the Synonymous Form that reads 'atomic time coordinate of coordinate time coordinate' should be 'atomic time coordinate of compound time coordinate'. There is no general concept that is termed 'coordinate time coordinate'.

a stated

In clause 11.2, the glossary entries for "Gregorian year of months scale", "Gregorian year of days scale", and "Gregorian month of days scale" each include Necessities of the form "The first element of the ...", where "first" is unstyled. This is a problem because Necessities should never have unstyled text. More importantly, these Necessities could instead by written as "The first member of the ...." If not, the second Note under "index origin member" in Annex D.2.3 is wrong.
-----------------------------

The issue is correct and the recommended approach solves the problem.

In clause 11.2, "Gregorian Time Points", the definition of "Gregorian calendar day" is "Gregorian day or Gregorian day of year or Gregorian day of month or Gregorian day of week". The problem with this definition is that there is no term "Gregorian day of week" anywhere in DTV – probably because the week calendar and the Gregorian calendar are unrelated.

Recommended solution: drop "Gregorian day of week" from this definition

As described

In DTV Clause 17.1, in the entry for ‘regular time table’, the following Necessity appears:

If the index of some table entry2 of the time table is 1 greater than the index of some

table entry1 of the time table, then the duration from table entry1 to table entry2 is the

repeat interval of the time table.

This goes well beyond the definition, which says only that the time table has an ‘intensional definition’, i.e., a rule that determines the entries. This Necessity states one kind of scheduling rule, but it prevents the use of a rule that specifies table entries by events, or by event plus or minus duration. (Consider military time tables, which state schedules for preparatory actions relative to a planned event whose occurrence interval is not fixed. These time tables have clear intensional definitions, but they don’t have ‘repeat intervals’.)

If the Description (“A regular time table has time table entries that repeat...”) is what is intended, then the Definition is not even close to conveying that. That concept is a regular sequence of not necessarily contiguous time intervals, which are determined by a starting point and a repeat interval. If this is what is intended, this is what the Definition should say.

If the intent is as general as the definition leads one to believe, this Necessity should be deleted, or used in a Note as a pattern for a kind of intensional definition that is based on a fixed repeat interval.

The above Necessity also assumes that the time table has exactly one repeat interval, which is not itself stated as a Necessity. It should be a requirement that a regular time table has at most 1 repeat interval. If the Description is what is meant, it must have exactly one. Note also that the second “Definition” under ‘repeat interval’ should be “Possibility”.

If the definition is what is intended, there should also be a requirement that the time table has exactly 1 ‘intensional definition’, since the term is marked as an SBVR concept.

The Definition in the v1.0 specification says that the table entries “repeat according to the repeat interval”. That is the intent. A regular time table is a repeating sequence of time intervals with a fixed repeat interval. The Necessity that the issue refers to is correct. The Definition will be reworded to clarify this by eliminating ‘intensional definition’.
The term ‘repeat interval’ is confusing, because it refers to a duration, not a time interval. It will be replaced.

In clause 11.7, in the Definition of 'Gregorian date', the reference to "Gregorian year month date coordinate" should be spelled "Gregorian year month day coordinate". The reference to "year weekday coordinate" should be spelled "year week day coordinate". Also, the concept 'year week day coordinate' is defined in the Week Calendar Vocabulary, which (according to figure 7.3) is not included by the Gregorian Calendar Vocabulary – and cannot be included without creating a circular inclusion problem.

The erroneous term “Gregorian year month date coordinate” occurs in the entry for Gregorian year month day coordinate as well as in the entry for Gregorian day.
The reference to year week day coordinate in clause 11.7 is a consequence of the definition of Gregorian day. It is properly a “calendar date that refers to a Gregorian day”. The definitions of the other three concepts should refer to ‘Gregorian date’ as the more general type.

Here's some more typos:

9. In clause 9.5, in the entry for "duration value set equals duration", the third synonymous form has the same wording as the primary form of the verb concept.
10. In clause 16.4, in the primary verb form for "occurrence1 is between occurrence2 and occurrence3", the word "and" should use verb style, rather than keyword style. In the first synonymous form, the word "and" should use verb style, rather than keyword style. The second and third synonymous forms are identical; one should be removed.
11. In Annex D.2.3, in the entry for "sequence has index origin position", the first Necessity should read "Each sequence has at most one index origin position.", rather than "Each sequence has at most one index origin value.".
-----------------------------
Mark H. Linehan
STSM, IBM Research
----- Forwarded by Mark H Linehan/Watson/IBM on 01/02/2014 04:04 PM -----

From: Mark H Linehan/Watson/IBM
To: dtv-rtf@omg.org,
Date: 12/30/2013 05:06 PM
Subject: Fw: DTV typos

--------------------------------------------------------------------------------

Here's another typo:

8. In clause 16.4, in the entry for "individual situation kind has occurrence interval", the Necessity "The individual situation kind has exactly one occurrence" should be "Each individual situation kind has exactly one occurrence interval".
-----------------------------
Mark H. Linehan
STSM, IBM Research
----- Forwarded by Mark H Linehan/Watson/IBM on 12/30/2013 05:00 PM -----

From: Mark H Linehan/Watson/IBM@IBMUS
To: dtv-rtf@omg.org, issues@omg.org,
Date: 12/30/2013 11:29 AM
Subject: DTV typos

--------------------------------------------------------------------------------

I've noticed several more typos in the DTV spec (formal-13-08-01.pdf):

1. In clause 11.2, in the entry for "Gregorian year of days scale", the first Necessity should be a Note.
2. In clause 11.2, in the entry for "Gregorian month of days scale", the first Necessity should be a Note.
3. In clause 16.7, in the entry for "time interval1 to occurrence specifies time interval2", "occurrence to time interval1 specifies time interval2", "occurrence1 to occurrence2 specifies time interval", "time interval1through individual situation kind specifies time interval2", the word "The" in the Necessity caption should have keyword style.
4. In clause 17.1, in the entry for "time table", the use of the keyword "must " in the first Necessity is not consistent with SBVR-SE style because modal keywords are not used in Necessities and Possibilities. The Necessity should read "Each time table has at least one table entry.".
5. In clause 17.2, in the entry for "schedule is for general situation kind", the first Necessity should read "Each schedule is for..." instead of "A schedule is for...".
6. In Annex D.2.1, the modal keyword "may" is used in the Possibilities for "sequence has first position" and "sequence has last position". In SBVR-SE style, modal keywords are implicit, not explicit in Possibility and Necessity statements.
7. In Annex D.2.3, the Necessity for "sequence has index origin position" should read "Each sequence has at most one index origin position." instead of "Each sequence has at most one index origin value.".
-----------------------------
Mark H. Linehan
STSM, IBM Research

All identified typographical errors will be corrected. Errors in SBVR entry subtitles will also be corrected.
This issue resolution incorporates other changes that only repair typographical errors.

The second Necessity in the glossary entry for "time point sequence" in clause 8.7 reads "Each time point sequence includes at least one time point.". The Necessity depends upon a verb concept "time point sequence includestime point" that does not exist.

Suggested solution: add a glossary entry for "time point sequence includestime point".

The verb ‘includes’ should be a reference to ‘sequence has member’. No new entry is needed. The wording will be corrected.

Source: Mark H. Linehan, IBM Research, mlinehan@us.ibm.com
Summary:
The verb concept 'time interval1 precedes time interval2', defined in clause 8.1.4, appears to have the same semantics as 'time interval1 is before time interval2' in clause 8.1.2. Also, figure 8.5 fails to show 'time interval1 precedes time interval2'.

‘Time interval1 precedes time interval2’ is redundant. It and its synonymous forms should be made synonymous forms for ‘time interval1 is before time interval2’ in clause 8.2.2. The entry can be kept in 8.2.4 with a See reference to the entry in 8.2.2. The UML model (which only captures primary terms) need not be modified.
Note: In 8.2.2, one of the synonymous forms for ‘is before’ is incorrect. It is also corrected by the changes below.

Title: Clause 8.3.2 dependency upon clause 10.2
Source: Mark H. Linehan, IBM Research, mlinehan@us.ibm.com
Summary:
The definitions of several standard time units that are defined in clause 8.3.2 are dependent upon "period" concepts that are defined in clause 10.2. Specifically:

The Definition of day in 8.3.2 references calendar day, which is in 10.2
The Definition of year in 8.3.2 references calendar year, in 10.2
The Definition of month in 8.3.2 references calendar month, in 10.2
The Definition of week in 8.3.2 references calendar week, in 10.2

This violates the Vocabulary structure shown in figure 7.3

It is not necessary for any of these units to make definitive references to the concepts in clause 10.2. Two of these units are taken to be precise multiples of ‘day’. The other two are nominal units that approximate the time intervals of astronomical events. Those same intervals are cited in the definitions in clause 10.2. They can be used in both places, so that (v1.0) clause 8.4.2 does not definitively reference clause 10.

In clause xx, the definition of 'time point kind' is "concept that is a specialization of the concept 'time point'", where 'specialization' is styled as a term. However, 'specialization' is not defined anywhere in either DTV or SBVR.

The definition should read "concept that specializesthe concept 'time point'"
-----------------------------

as described

In clause 12.3 of the Date-Time Vocabulary, in the Definition of 'weeks scale', the term 'indefinite time scale' should be styled as a term, rather than as an individual concept. Also, the definition should probably start "the indefinite time scale that ...."

as stated

In clause 9.5, the first caption under "duration value set2 = duration – duration value set1", should be a Synonymous Form, not a Definition. And the second caption should be a Definition rather than a Synonymous Form.

Also (following on a previous request), the primary term for several glossary entries in this section should use English words rather than algebraic symbols, as in "duration value set2 equals duration minus duration value set1".

The second paragraph above is not about a typo. It requires correcting the primary verb terms for three of the entries in 9.5, which is the subject of Issue 19016.
Disposition: See issue 19016 for disposition

Algebraic symbols (<, <=, =) are incorporated in the terms of various concepts, such as "duration1 < duration2". This causes problems when these concepts are mapped to UML, CLIF, or OWL because these languages restrict the character set of names to alphanumeric characters. You can see this in figure 8.10, where this concept is mapped to the "is less than" method name – probably because OCL does not support the "<" character.

Suggestion: use primary terms that are only alphanumeric, and restrict algebraic symbols to Synonymous Forms. For example, as is already done for "time interval1 is before time interval2".

The RTF agrees that the mathematical symbols should not be the primary terms, in order to simplify mappings to implementation languages that restrict the characters permitted in terms. In each case, the mathematical symbol will be retained as a synonymous form for the verb concept. In addition, the specification of the ‘noun forms’ for the arithmetic verbs is inconsistent; both the sign form and the word form will be specified for both plus and minus.
The revised text also corrects a paragraph labeling error (Issue 19032), and removes a redundant copy of the axioms for the duration ordering.

In clause 8.1.3, the definition of time interval1 meets time interval2:

“the time interval1 is before the time interval2 and the time interval1 is not before a time interval3 that is before the time interval2”

is inaccurate at best. The CLIF and OCL definitions are correct.

As stated, the definition says that there is one time interval3 that is before time interval2 and that time interval1 is not before, but it does not say that there is no other time interval3 is before time interval2 and that time interval1 is not before. The definition should read:

“time interval1 is before time interval2 and there is NO time interval3 that is after time interval1 that is before time interval2”

This revised statement matches the CLIF and OCL definitions.

There may be other such misstatements in 8.1.3.

Correct the text as recommended

In DTV Beta-2,Clause 8.2.3, there are two verb concepts:
time interval1 is duration before time interval2
time interval1 is duration after time interval2

From the alternative form: "duration before/after time interval2", it seems clear that the intent of these verb concepts is to allow a time interval to be defined by a reference time interval and a duration, e.g., the two weeks before the jump-off date, the day after the meeting (day). Each of these denotes exactly one time interval.

But the Definitions mean that the verb concepts simply state the duration between two time intervals. This may be useful when the intent is to state the duration between two events, but it is not the meaning of 'duration before time interval', and it cannot be used to define a time interval. Either these verb concepts should be defined to be the ones intended by the alternative forms, or the alternative forms should be separate verb concepts.

The intent of these two verb concepts is to define time intervals, but there are two ways to define time intervals in terms of durations. ‘the two weeks before the meeting’ refers to a time period of two weeks, ending with the meeting, as the issue describes. ‘two weeks before the meeting’ refers to a time interval (a day) that is separated from the meeting by two weeks, which is the intent of the existing text. The concepts in the issue statement are additional concepts, not replacement concepts. They are added.
The RTF also notes that the most common bases for duration before and after are used with events rather than time intervals. This simplifies the business usage by avoiding the circumlocution ‘the time interval when ...’. The corresponding verbs are added to clause 16. These are simplifications of concepts that use the verb concepts at issue.
Clause 16.7 was already very large. The edit also formally creates two subsections at the obvious boundary.

DTV clause 4 has a list of concepts that are extracted from the SBVR specification. This clause includes "representation uses expression", which does not in fact appear in SBVR. The correct concept is "representation has expression".

Also, DTV clause 4 includes "meaning has representation", which is a Synonymous Form, not a primary term, of "representation represents meaning" in the SBVR specification. Both the primary term and the two Synonymous Forms should be copied from the SBVR specification.

The references to the SBVR vocabulary entries will be corrected.

The DTV index has an entry 'd 71' that appears nowhere in the text. It should be removed.

as stated

The individual concept 'second' is defined in clause 8.4.2 as a 'precise time unit', which is a kind of 'measurement unit'. 'Second' should also be defined as a kind of 'base measurement unit' (annex D.3.2) because 'second' is one in the SI system.

Since all the other precise time units are derived from ‘second’, it is important to say that this is the base measurement unit. Add a second Definition.

In annex D.2, Sequences, the entry for "first member" has "General Concept: role" and "Possibility: thing". It should be "Concept Type: role" and "General Concept: thing".

These errors were corrected in DTV v1.0.
Disposition: Closed No Change

In DTV clause 8.6, the Definition of ‘time point1 to time point2 specifies time period’ reads:

Definition: time point1 is the first time point of a time point sequence and some time point3 is the

last time point of the time point sequence and time point2 is just before time point3 in

the time point sequence and time point1 through time point2 specifies the time period

The subscripts on time point2 and time point3 are reversed. It should read:

Definition: time point1 is the first time point of a time point sequence and time point2 is the

last time point of the time point sequence and there is a time point3 that is just before time point2 in

the time point sequence and time point1 through time point3 specifies the time period

Correct the text as recommended.

In clause 9.0 of the Date-Time Vocabulary, the vocabulary entry for the "Duration Values Vocabulary" has "Included Vocabulary: Duration Values Vocabulary". Per diagram 7.3, it should be "Included Vocabulary: Time Infrastructure Vocabulary".
An extension of the same issue:

In clause 10.0, the vocabulary entry for the "Calendars Vocabulary" has "Included Vocabulary: Calendars Vocabulary". It should have "Included Vocabulary: Time Infrastructure Vocabulary".

In clause 11.0, the vocabulary entry for the "Gregorian Calendars Vocabulary" has "Included Vocabulary: Gregorian Calendars Vocabulary". It should have "Included Vocabulary: Calendars Vocabulary".

as described. These are editorial errors. The changes to clause 11.1 are incorporated in the resolution to Issue 18961, which changes the “included vocabularies”.

The Date-Time Vocabulary 1.0 entry for "index" reads "Definition: integer". It should be "General Concept: integer".

as stated

Clause 11, Gregorian Calendar, defines concepts 'year value', 'years duration value set', 'month value', and 'months duration value set' – all of which are dependent upon concepts defined in clause 9, 'Duration Values'. But (1) the 'Gregorian Calendar Vocabulary' does not include the 'Duration Values Vocabulary', and (2) figure 7.3 fails to show the dependency. Both of these should be fixed.

As described. The included calendar references in the Gregorian Calendar Vocabulary are corrected. Introductory text is also added to the beginning of Clause 11.

(See red circle in attached diagram). Figure 8.11, Duration Operations, shows that each instance of the class 'duration' participates in exactly one instance of the class 'duration1 equals number times duration2' as role 'duration1'. Clearly this is false. For example, the duration "4 hours" is the result of both "2 times 2 hours" and "4 times 1 hour".

Note that the UML diagrams show that instances of each role can participate in any number of instances of the reified classes. For example, see "duration1 equals duration2 minus duration3".

The erroneous multiplicity in the diagram is corrected to 0..*.

Summary:
These two ‘year of weeks scale’ and ‘year of weekdays scale’ are defined using the ‘time scale subdivides time point’ verb concept. That verb concept assumes that the start of the time point coincides with the start of the scale, but this is not true for these two scales because calendar weeks are not coherent with calendar years. A special verb concept needs to be used to relate weeks and weekdays to calendar years.

The issue is correct: subdivides does not characterize the year of weeks or the year of weekdays scales. These time scales are finite sequences that map directly to the weeks calendar and year of days calendar respectively