Issue 17428: basic time coordinate concepts are badly described (date-time-ftf) Source: NIST (Mr. Edward J. Barkmeyer, edbark(at)nist.gov) Nature: Uncategorized Issue Severity: Summary: Figure 12.1 shows 'time coordinate' as a specialization of SBVR 'representation', which is said to be the relationship between an expression and the meaning it denotes. No UML diagram shows 'time expression' (from 12.1.2), which one would expect to be the corresponding specialization of 'expression'. In 12.1, the definition of 'time coordinate' is: "representation of a time point by an atomic time coordinate or compound time coordinate that indicates the time point". This is both circular and incorrect. A time coordinate cannot represent a time point by a subtype of itself, it must represent a time point by an expression. The definition should read: "representation of a time point by a time expression" (to exclude representation by a definite description). One would expect a 'time expression' to be "the expression of a time coordinate" (from SBVR 'representation has expression', mislabeled 'expression uses representation' in the diagram), but it isn't. It is said to be an expression involving an index (integer) and a time scale, which means it is not the expression of a compound time coordinate. In 12.1.2, the definition of 'atomic time coordinate indicates time point' requires the time expression to contain an index, which means that the "February" example is invalid. The expression (string) "February" includes neither an index nor a time scale. The time coordinate, as a 'representation', is the association of "February" with the time point that is month of year 2. The atomic time coordinate thus acquires the time scale and index properties from the time point it indicates. But the expression "February" does not have those properties. "February" is associated with that time point via 'concept has designation'. If DTV is to explain how a time expression is associated with a time point, it has to distinguish the properties of the expression from the properties of the association (that SBVR calls a 'representation'). A 'simple time expression' is either the expression of an 'index', which would be some expression of an integer, OR a 'signifier' for a time point (from SBVR 'concept has designation' and 'designation has signifier (expression)'). The signifier could be a given name, like "March", or a constructed term involving the scale and the index, like "Gregorian month 3". (What 12.1.2 describes is only the last case.) A 'compound time expression' is some syntax that combines two or more simple time expressions. In SBVR style, then, an atomic time coordinate is a time coordinate whose expression is a simple time expression, and a compound time coordinate is a time coordinate whose expression is a compound time expression. And the rules for determining what a simple time expression indicates, i.e., how the link that is the atomic time coordinate is constructed, depend on the nature of the time expression. In particular, the context of a compound time expression may make the intent of an index expression clear, as in "3/31/2012", where the "3" is only recognized as a month of year index because of its placement. Resolution: Revised Text: Actions taken: June 14, 2012: received issue Discussion: End of Annotations:===== te: Thu, 14 Jun 2012 19:46:07 -0400 From: Ed Barkmeyer Reply-To: Organization: NIST User-Agent: Thunderbird 2.0.0.23 (Windows/20090812) To: "issues@omg.org" Subject: DTV Issue: basic time coordinate concepts are badly described X-NISTMEL-MailScanner-Information: Please contact postmaster@mel.nist.gov for more information X-NISTMEL-MailScanner-ID: q5ENkCBL025752 X-NISTMEL-MailScanner: Found to be clean X-NISTMEL-MailScanner-SpamCheck: X-NISTMEL-MailScanner-From: edbark@nist.gov X-NISTMEL-MailScanner-Watermark: 1340322373.68056@Wx5s0TFF8haLv35bx0xPCg X-Spam-Status: No Specification: Date Time Vocabulary Version: Beta-1 Title: basic time coordinate concepts are badly described Source: Ed Barkmeyer, NIST, edbark@nist.gov Summary: Figure 12.1 shows 'time coordinate' as a specialization of SBVR 'representation', which is said to be the relationship between an expression and the meaning it denotes. No UML diagram shows 'time expression' (from 12.1.2), which one would expect to be the corresponding specialization of 'expression'. In 12.1, the definition of 'time coordinate' is: "representation of a time point by an atomic time coordinate or compound time coordinate that indicates the time point". This is both circular and incorrect. A time coordinate cannot represent a time point by a subtype of itself, it must represent a time point by an expression. The definition should read: "representation of a time point by a time expression" (to exclude representation by a definite description). One would expect a 'time expression' to be "the expression of a time coordinate" (from SBVR 'representation has expression', mislabeled 'expression uses representation' in the diagram), but it isn't. It is said to be an expression involving an index (integer) and a time scale, which means it is not the expression of a compound time coordinate. In 12.1.2, the definition of 'atomic time coordinate indicates time point' requires the time expression to contain an index, which means that the "February" example is invalid. The expression (string) "February" includes neither an index nor a time scale. The time coordinate, as a 'representation', is the association of "February" with the time point that is month of year 2. The atomic time coordinate thus acquires the time scale and index properties from the time point it indicates. But the expression "February" does not have those properties. "February" is associated with that time point via 'concept has designation'. If DTV is to explain how a time expression is associated with a time point, it has to distinguish the properties of the expression from the properties of the association (that SBVR calls a 'representation'). A 'simple time expression' is either the expression of an 'index', which would be some expression of an integer, OR a 'signifier' for a time point (from SBVR 'concept has designation' and 'designation has signifier (expression)'). The signifier could be a given name, like "March", or a constructed term involving the scale and the index, like "Gregorian month 3". (What 12.1.2 describes is only the last case.) A 'compound time expression' is some syntax that combines two or more simple time expressions. In SBVR style, then, an atomic time coordinate is a time coordinate whose expression is a simple time expression, and a compound time coordinate is a time coordinate whose expression is a compound time expression. And the rules for determining what a simple time expression indicates, i.e., how the link that is the atomic time coordinate is constructed, depend on the nature of the time expression. In particular, the context of a compound time expression may make the intent of an index expression clear, as in "3/31/2012", where the "3" is only recognized as a month of year index because of its placement. To: date-time-ftf@omg.org Subject: Date-Time Issue 17428 - basic time coordinate concepts are badly described X-KeepSent: 03116EA6:B3976650-85257A58:00826232; type=4; name=$KeepSent X-Mailer: Lotus Notes Release 8.5.3 September 15, 2011 From: Mark H Linehan Date: Sun, 12 Aug 2012 19:45:38 -0400 X-MIMETrack: Serialize by Router on D01MC604/01/M/IBM(Release 8.5.3 ZX853HP5|January 12, 2012) at 08/12/2012 19:45:39 X-Content-Scanned: Fidelis XPS MAILER x-cbid: 12081223-6078-0000-0000-00000E2E021E This one has some embedded questions that I want to discuss. -------------------------------- Mark H. Linehan STSM, IBM Research Date-Time Issue 17428 - basic time coordinate concepts are badly described.doc Disposition: ??? OMG Issue No: 17428 Title: basic time coordinate concepts are badly described Source: Ed Barkmeyer, NIST, edbark@nist.gov Summary: Figure 12.1 shows 'time coordinate' as a specialization of SBVR 'representation', which is said to be the relationship between an expression and the meaning it denotes. No UML diagram shows 'time expression' (from 12.1.2), which one would expect to be the corresponding specialization of 'expression'. In 12.1, the definition of 'time coordinate' is: "representation of a time point by an atomic time coordinate or compound time coordinate that indicates the time point". This is both circular and incorrect. A time coordinate cannot represent a time point by a subtype of itself, it must represent a time point by an expression. The definition should read: "representation of a time point by a time expression" (to exclude representation by a definite description). One would expect a 'time expression' to be "the expression of a time coordinate" (from SBVR 'representation has expression', mislabeled 'expression uses representation' in the diagram), but it isn't. It is said to be an expression involving an index (integer) and a time scale, which means it is not the expression of a compound time coordinate. In 12.1.2, the definition of 'atomic time coordinate indicates time point' requires the time expression to contain an index, which means that the "February" example is invalid. The expression (string) "February" includes neither an index nor a time scale. The time coordinate, as a 'representation', is the association of "February" with the time point that is month of year 2. The atomic time coordinate thus acquires the time scale and index properties from the time point it indicates. But the expression "February" does not have those properties. "February" is associated with that time point via 'concept has designation'. If DTV is to explain how a time expression is associated with a time point, it has to distinguish the properties of the expression from the properties of the association (that SBVR calls a 'representation'). A 'simple time expression' is either the expression of an 'index', which would be some expression of an integer, OR a 'signifier' for a time point (from SBVR 'concept has designation' and 'designation has signifier (expression)'). The signifier could be a given name, like "March", or a constructed term involving the scale and the index, like "Gregorian month 3". (What 12.1.2 describes is only the last case.) A 'compound time expression' is some syntax that combines two or more simple time expressions. In SBVR style, then, an atomic time coordinate is a time coordinate whose expression is a simple time expression, and a compound time coordinate is a time coordinate whose expression is a compound time expression. And the rules for determining what a simple time expression indicates, i.e., how the link that is the atomic time coordinate is constructed, depend on the nature of the time expression. In particular, the context of a compound time expression may make the intent of an index expression clear, as in "3/31/2012", where the "3" is only recognized as a month of year index because of its placement. Resolution: The relationships among .time coordinate., .atomic time coordinate., and .compound time coordinate. are disentangled to say that: . .time coordinate. is an .atomic time coordinate. or a .compound time coordinate. . .atomic time coordinate. is a designation of a time point, where the signifier is either a time expression or a time point name . .time point name. is the signifier of an individual concept of a time point . .compound time coordinate. is a .designation that represents a time point by more than one atomic time coordinate. . the definition of .atomic time coordinate indicates time point. is rewritten to explain how it works in the two cases where the signifier of the atomic time coordinate is a time point name versus when it is a time expression Other miscellaneous changes: . The text at the start of clause 10.5.1 is instead made the introductory text for all of 10.5. . References to the clauses where time coordinates are defined are fixed. . A verb concept .time coordinate has time scale. is added to support the definitions of .absolute time coordinate. and .relative time coordinate. . The definitions of .absolute time coordinate. and .relative time coordinate. are corrected The issue summary made the comment that .the context of a compound time expression may make the intent of an index expression clear, as in "3/31/2012", where the "3" is only recognized as a month of year index because of its placement.. This specification does not describe the the internal structure of a time coordinate, NOT the external representation of one. The interpretation of a date such as .3/31/2012. is a function of a tool, not of this specification. Open questions: . Given that DTV does not specify the external representation of time coordinates, is it misleading to define them as designations that have signifiers? . The updates given in this issue rework most of the key time coordinate concepts, but are necessarily presented out of context. The beta-2 document, clause 10.5 is rather chopped-up: concepts are not introduced in the optimal sequence. In particular, .compound time coordinate combines atomic time coordinate. should be earlier in the section. Would it make sense to replace the entire sub-clause in this resolution in order to clean it up. Revised Text: All changes described here are with respect to the beta-2 document. In clause 4, add the following entries in the appropriate alphabetical positions: concept has designation designation In clause 10, move the heading .10.5.1 General. to just before figure 10.6, so that the text between heading 10.5.1 and figure 10.6 becomes text directly under the clause 10 heading. In existing clause 10.5.1, replace the third sentence of paragraph 1, which reads: This clause specifies which combinations of atomic time coordinates form legitimate compound time coordinates. . with: Clauses 11, 12, and 13 specify which combinations of atomic time coordinates form legitimate compound time coordinates. In clause 10.5.1, replace figure 10.6 to match the new definition of .time coordinate. (below). In clause 10.5.1, replace the Definition and Reference of .time coordinate., which read: Definition: representation of a time point by an atomic time coordinate or compound time coordinate that indicates the time point Reference Scheme: the indices and time scales of the time coordinate . with: Definition: atomic time coordinate or compound time coordinate Reference Scheme: the signifier of the time coordinate In clause 10.5.1, replace the final Note under .time coordinate., which reads: Note: Particular kinds of time coordinates are defined below. . with: Note: Particular kinds of time coordinates are defined in clauses 11, 12, and 13. In clause 10.5.1, replace the definition of .time coordinate indicates time point., which reads: Definition: time point is the scale point of the time scales of the time coordinate that has the indices of the time coordinate . with a Synonymous Form, General Concept, and new Definition: Synonymous Form: time point is indicated by time coordinate General Concept: expression represents meaning Definition: the time coordinate represents the time point In clause 10.5.1, add the following after the glossary entry for .time coordinate indicates time point.: time coordinate has time scale Definition: the time scale is the time scale of the time point that is indicated by the time coordinate Example: The time scale of .Tuesday. is the week of days scale. Example: The time scale of .2012. is the Gregorian years scale. Example: The time scale of .12 August 2012. is the Gregorian days scale. In clause 10.5.2, replace the definition of .absolute time coordinate., which reads: Definition: time coordinate that has an indefinite time scale and refers to exactly one time interval . with: Definition: time coordinate that has an indefinite time scale and that indicates a time point that corresponds to exactly one time interval In clause 10.5.2, replace the definition of .relative time coordinate., which reads: Definition: time coordinate that does not have an indefinite time scale and refers to more than one time interval . with: Definition: time coordinate that has a finite time scale and that indicates a time point that corresponds to more than one time interval In clause 10.5.3, replace figure 10.8 with this version: In clause 10.5.3, replace the definition of .time expression., which reads: Definition: expression that is of exactly one index on exactly one time scale . with: Definition: signifier that is of an atomic time coordinate, and that includes exactly one index on exactly one time scale In clause 10.5.3, delete the Note under .time expression., which reads: Note: A time expression may be specified as an index and a time scale (e.g. "month 1") or as an individual concept (e.g. "January") that is defined in terms of an integer and a time scale. The time scale may be implicit (e.g. "2010" used as a time coordinate implies the Gregorian years scale). In clause 10.5.3, add a new glossary entry immediately before .atomic time coordinate.: time point name Definition: name of an individual concept that is a time point that is on exactly one time scale and has exactly one index Example: .January., .Wednesday. In clause 10.5.3, replace the Definition of .atomic time coordinate., which reads: Definition: time coordinate that represents a time point by exactly one time expression . with: Definition: designation that is of a time point and that has a signifier that is a time expression or a time point name In clause 10.5.3, delete the Reference Scheme for .atomic time coordinate.: Reference Scheme: The index of the time expression and the time scale of the time expression. In clause 10.5.3, replace the Definition of .atomic time coordinate indicates time point., which reads: Definition: the index of the time expression of the atomic time coordinate is the index of the time point on the time scale of the time point . with a General Concept and a new Definition: General Concept: time coordinate indicates time point Definition: if the signifier of the atomic time coordinate is a name, then the name indicates the time point, and if the signifier of the atomic time point is a time expression, then the index of the time expression of the atomic time coordinate is the index of the time point on the time scale of the time point In clause 10.5.3, replace the Definition of .compound time coordinate., which reads: Definition: time coordinate that represents a time point by more than one time expression . with: Definition: designation that represents a time point by more than one atomic time coordinate In clause 10.5.3, delete the first Example under .compound time coordinate., which reads: Example: "2010" is a precise atomic time coordinate indicating a Gregorian year on the Gregorian years scale. In clause 10.5.3, delete the glossary entry for .compound time coordinate has atomic time coordinate. because it duplicates the entry for .compound time coordinate combines atomic time coordinate.. Disposition: ??? To: date-time-ftf@omg.org Subject: DTV Issue 17428 - Time point subdivision is out of place twice X-KeepSent: 2E9A5577:9227B942-85257A69:00688A2D; type=4; name=$KeepSent X-Mailer: Lotus Notes Release 8.5.3 September 15, 2011 From: Mark H Linehan Date: Wed, 29 Aug 2012 15:03:49 -0400 X-MIMETrack: Serialize by Router on D01MC604/01/M/IBM(Release 8.5.3FP2IF1|July 25, 2012) at 08/29/2012 15:03:49 X-Content-Scanned: Fidelis XPS MAILER x-cbid: 12082919-9360-0000-0000-000009F9FE82 Here is an updated resolution from our conversation this morning plus a few more changes -- all of which are highlighted. -------------------------------- Mark H. Linehan STSM, IBM Research Date-Time Issue 16951- time point subdivision is out of place twice.docx Disposition: Resolved OMG Issue No: 16951 Title: Time point subdivision is out of place twice Source: Ed Barkmeyer, NIST, edbark@nist.gov Summary: At the end of section 9.3 the fact type 'finite time scale subdivides time point' is defined and depicted in Figure 9-6. This fact type is a commonly used property of finite time scales, and is not specific to individual time points. That is, the finite time scale day-of-hours subdivides every day-of-month, for example. So the text and diagram should be at the end of 9.2, where finite time scale is defined. Also, diagram 9-6 does not show the operations on time point and finite time scale that are associated with this fact type and shown in diagram 12-12. Finally, diagram 12-12 (section 12.4) shows this fact type as well, but section 12.4 never uses it, and diagram 12-12 shows a <> relationship that is never discussed anywhere in the text. If the relationship is important, it should be discussed. Otherwise, the subdivision association is out of place in diagram 12-12 and the "specialization" dependency should be deleted from the UML model. Recommendation: Move the diagram and text for 'finite time scale subdivides time point' to the end of 9.2, include the operations on the diagram. Delete the fact type from diagram 12-12, and delete the "specialization" dependency altogether. Resolution: The FTF agrees that the fact-type 'finite time scale subdivides time point' should be moved, along with Figure 9-6, but it depends on concepts defined in 9.3 and 9.4. It is an important concept that should have its own subsection. This concept is used only in defining finite time scales within Calendars. So the new section goes in clause 10 (formerly 9.5) Calendars. The verb concept 'finite time scale subdivides time point' is not quite the way it is used. All of the existing usages have the form: 'finite time scale subdivides time point into number of time point kind.' But the finite time scale determines the time point kind, so that is redundant. Also, it is not clear whether the intent is to specify the cardinality of the finite time scale or the cardinality of the time point sequence that is the subdivision of individual time points, such as months of the year. The text is revised to clarify that. distinguish subdivides as a characteristic of finite time scales from the specification of the actual time point sequences used to subdivide individual time points. The term exactly subdivides is used when all the time points have the same subdivision; and the special cases use the (now defined) verb 'time point has number of time point kind' that was already present in the entries for the special cases. Note: This clarification was not applied to the 'year of weeks' and 'year of weekdays' time scales, whose nature is addressed by a different issue. Simply stated, the year of weeks does not subdivide a calendar year. The concept in diagram 10.14 (formerly 12.12) . time point converts to time point sequence on time scale . is only distantly related. The nature of the time point sequences involved is different, and the purposes are different. The purpose of time point subdivision is to be part of the definition of finite time scales. Diagram 10.14 (formerly 12.12) no longer shows the "specialization"; it was removed by resolution of another issue. Revised Text: 1. At the end of 8.5 (was 9.3), DELETE Figure 8.18 (was 9.6) and the following text: (Figure 8.18) Figure 8.18: Time Scale Subdivision finite time scale subdivides time point Definition: the finite time scale divides the time point into a number of time intervals that each have a duration, and the number is the cardinality of the finite time scale, and the duration is the granularity of the finite time scale Example: The day of hours scale subdivides the Gregorian date 3 January 2010 into 24 time intervals, each of duration "hour" 2. Immediately before section 10.3 (was 9.5), create a new subsection numbered 10.3, and renumber the subsequent sections: 10.3 Time Point Subdivision The purpose of finite time scales is to provide finer-grained resolution of time intervals within the time intervals that are instances of time points with coarser granularities. In this specification, the relationship between a finite time scale and a coarser time point is called .time point subdivision.. Many finite time scales are defined by the category of time point they subdivide and the granularity of the time points they contain. Figure 10.x: Time Point Subdivision time point kind Definition: concept that is a specialization of the concept 'time point' Concept Type: categorization type Example: .Gregorian month of year. is a time point kind because it specializes .calendar month. which itself specializes .time point.. .January. is not a time point kind because it is an instance (not a specialization) of .Gregorian month of year.. finite time scale subdivides time point kind Definition: for each time point that is an instance of time point kind, there is a set of time point sequences that are on the finite time scale, and the first time point of each time point sequence that is in the set is the index origin member of the finite time scale, and each time interval that is an instance of the time point is an instance of some time point sequence that is in the set Definition: each instance of each time point that is an instance of the time point kind is an instance of some time point sequence that is on the finite time scale, and the first time point of the time point sequence is the index origin member of the finite time scale Note: This verb concept describes the purpose of the finite time scale: each time point of the finite time scale corresponds to time intervals according to their position relative to the start of a time interval that is an instance of some time point of the time point kind. The first time point of the finite time scale corresponds to time intervals that start the larger time intervals and have a duration equal to the granularity of the finite time scale. Necessity: For most time point subdivisions, the "set of time point sequences" has exactly one member . one time point sequence corresponds to all instances of the time point. The sole exception is the time point February. Necessity: The granularity of each finite time scale is less than the granularity of each time point that is an instance of the time point kind that the finite time scale subdivides. Necessity: The granularity of each time point kind that is subdivided by a finite time scale is greater than the granularity of the finite time scale. Note: The time point sequence may correspond to many more time intervals than the instances of such time points. Note: The finite time scale may subdivide all time points of a kind into the same time point sequence or it may subdivide different time points The same time point sequence may be the subdivision of all time points the time point kind, or different time points may be subdivided into time point sequences of different lengths. Note:Example: The day of hours scale subdivides Gregorian calendar day. Every time point that is a Gregorian calendar day is subdivided into 24 hour of day time points, and each corresponding time interval is divided into 24 time intervals, each of which is an instance of one hour- of- day. Note: A time point kind may be subdivided by finite time scales of different cardinalities. Example: The Gregorian month of days scale subdivides month of year. Every time point that is a Gregorian month of year is subdivided into some number of day of month time points, and the time point sequences all begin with day of month 1, but the length cardinality of the time point sequence depends on which month time point is subdivided. time point has number of time point kind Note: there is a time point sequence that is on the finite time scale that each instance of the time point kind is a member of, and the first time point of the time point sequence is the index origin member of the finite time scale, and the cardinality of the time point sequence is number, and the time point sequence corresponds to each time interval that is an instance of time point Definition: each instance of the time point kind is a member of some finite time scale and some time point sequence is on the finite time scale and the first time point of the time point sequence is the index origin member of the finite time scale and the number is the cardinality of the time point sequence and the time point sequence corresponds to each instance of time point Note: This verb concept describes the relationship between the finite time scale and an individual time point of the kind that the finite time scale subdivides. That is, it completely specifies the subdividing time point sequence. Example: If a time point has some number of time point kind1, the finite time scale that each instance of time point kind1 is a member of subdivides the time point kind2 that the time point is an instance of. Necessity: Each time point that has a number of some time point kind is an instance of some time point kind2 and each instance of the time point kind is a member of some finite time scale and the finite time scale subdivides the time point kind2. Example: Gregorian day 3 January 2010 is subdivided by the day of hours scale intohas 24 'hour of day' time points. The time interval corresponding to Gregorian date 3 January 2010 is implicitly subdivided into 24 time intervals, each of which is an instance of one hour- of- day. But that same 24-hour time point sequence is the subdivision of every Gregorian day, and it corresponds to every time interval that is an instance of a Gregorian day time interval. finite time scale exactly subdivides time point kind General Concept: finite time scale subdivides time point kind Definition: the finite time scale is a time point sequence that corresponds to each time interval that is an instance of each time point that is an instance of the time point kind Necessity: for each time point that is an instance of time point kind, the time point sequence that is the finite time scale corresponds to each time interval that is an instance of the time point Necessity: Each finite time scale that exactly subdivides a time point kind subdivides the time point kind. Necessity: If a finite time scale exactly subdivides time point kind1, and each time point of the finite time scale is an instance of time point kind2, then the number of time point kind2 that each time point that is an instance of time point kind1 has is the cardinality of the finite time scale. Necessity: The cardinality of each finite time scale that exactly subdivides a time point kind times the granularity of the finite time scale equals the duration of each instance of each time point that is an instance of the time point kind. Example: The day of hours scale exactly subdivides Gregorian day of month. Every Gregorian day of month therefore has 24 of 'hour of day', because 24 is the cardinality of the 'day of hours scale'. 3. In clause 11.1 (was 9.5.5), in the entry for 'Gregorian year of months scale', REPLACE the Definition and the first Necessity: Definition: the finite time scale that subdivides .Gregorian year. into 12 of .Gregorian month of year. Necessity: The granularity of the Gregorian year of months scale is .month.. with: Definition: the finite time scale that has granularity 1 month and that has cardinality 12 and that exactly subdivides 'Gregorian year' Necessity: Each Gregorian year is subdivided into a the same sequence of has 12 of Gregorian month of year time points. 4. In clause 11.1 (was 9.5.5), in the entry for 'Gregorian year of days scale', REPLACE the Definition and the first Necessity: Definition: the finite time scale that subdivides .Gregorian year. into 366 of .Gregorian day of year. Necessity: The granularity of the Gregorian year of days scale is .day.. with: Definition: the finite time scale that has granularity 1 day and that has cardinality 366 and that subdivides 'Gregorian year' Note: Each leap year is subdivided into 366 Gregorian day of year time points. Each common year is subdivided into 365 Gregorian day of year time points. 5. In clause 11.1 (was 9.5.5), in the entry for 'Gregorian month of days scale', REPLACE the Definition and the first Necessity: Definition: the finite time scale that subdivides .Gregorian month. into 31 of .Gregorian day of month. Necessity: The granularity of the Gregorian month of days scale is .day.. with: Definition: the finite time scale that has granularity 1 day and that has cardinality 31 and that subdivides 'Gregorian month of year' Note: Each Gregorian month of year is subdivided into a specific number of Gregorian day of month time points. The subdivision of February is a set of two time sequences. 6. In clause 11.2, in the entry for 'common year', after the Definition paragraph, ADD: Necessity: Each common year has 365 of 'Gregorian day of year'. 7. In clause 11.2, in the entry for 'leap year', after the Definition paragraph, ADD: Necessity: Each leap year has 366 of 'Gregorian day of year'. 8. In clause 11.2, before the entry for Gregorian day, INSERT a new entry: Gregorian calendar month Definition: Gregorian month or Gregorian month of year Concept Type: concept type 9. In clause 11.2, at the end, ADD a new entry: Gregorian calendar day Definition: Gregorian day or Gregorian day of year or Gregorian day of month or Gregorian day of week Concept Type: concept type 10. [Note to Editor: There is a pattern to these changes: For all but February, insert "of '" between the number and the term that follows it, and follow the term with a closing single quote ('). 10a. In clause 11.3, in the entry for January, REPLACE the Necessity that reads: Necessity: January has 31 Gregorian days of month. with Necessity: January has 31 of 'Gregorian day of month'. 10b. In clause 11.3, in the entry for February, REPLACE the Necessities that read: Necessity: If the Gregorian year is a common year, then February has 28 Gregorian days of month. Necessity: If the Gregorian year is a leap year, then February has 29 Gregorian days of month. with: Necessity: The number of 'Gregorian day of month' that February has is 28 or 29. Necessity: The time point sequence that is Gregorian day of month 1 through Gregorian day of month 28 corresponds to each February that is during a common year. Necessity: The time point sequence that is Gregorian day of month 1 through Gregorian day of month 29 corresponds to each February that is during a leap year. Note: The set of these two time point sequences is how Gregorian month of days subdivides February. 10c. In clause 11.3, in the entry for March, REPLACE the Necessity that reads: Necessity: March has 31 Gregorian days of month. with Necessity: March has 31 of 'Gregorian day of month'. 10d. In clause 11.3, in the entry for April, REPLACE the Necessity that reads: Necessity: April has 30 Gregorian days of month. with Necessity: April has 30 of 'Gregorian day of month'. 10e. In clause 11.3, in the entry for May, REPLACE the Necessity that reads: Necessity: May has 31 Gregorian days of month. with Necessity: May has 31 of 'Gregorian day of month'. 10f. In clause 11.3, in the entry for June, REPLACE the Necessity that reads: Necessity: June has 30 Gregorian days of month. with Necessity: June has 30 of 'Gregorian day of month'. 10g. In clause 11.3, in the entry for July, REPLACE the Necessity that reads: Necessity: July has 31 Gregorian days of month. with Necessity: July has 31 of 'Gregorian day of month'. 10h. In clause 11.3, in the entry for August, REPLACE the Necessity that reads: Necessity: August has 31 Gregorian days of month. with Necessity: August has 31 of 'Gregorian day of month'. 10i. In clause 11.3, in the entry for September, REPLACE the Necessity that reads: Necessity: September has 30 Gregorian days of month. with Necessity: September has 30 of 'Gregorian day of month'. 10j. In clause 11.3, in the entry for October, REPLACE the Necessity that reads: Necessity: October has 31 Gregorian days of month. with Necessity: October has 31 of 'Gregorian day of month'. 10k. In clause 11.3, in the entry for November, REPLACE the Necessity that reads: Necessity: November has 30 Gregorian days of month. with Necessity: November has 30 of 'Gregorian day of month'. 10l. In clause 11.3, in the entry for December, REPLACE the Necessity that reads: Necessity: December has 31 Gregorian days of month. with Necessity: December has 31 of 'Gregorian day of month'. 11. In clause 12.2, in the entry for week of days scale, REPLACE the Definition and the first Necessity: Definition: the finite time scale that subdivides .calendar week. into 7 of .calendar day. Necessity: The granularity of the week of days scale is .day.. with: Definition: the finite time scale that has granularity 1 day and that has cardinality 7 and that exactly subdivides 'calendar week' Necessity: Each calendar week has 7 of 'day of week'. 12. In clause 13.1, in the entry for day of hours scale, REPLACE the Definition and the first Necessity: Definition: the finite time scale of 24 .hour of day. Necessity: The granularity of the day of hours scale is .hour.. with: Definition: the finite time scale that has granularity 1 hour and that has cardinality 24 and that exactly subdivides 'Gregorian calendar day' Necessity: Each calendar day has 24 of 'hour of day'. 13. In clause 13.1, in the entry for day of minutes scale, REPLACE the Definition and the first Necessity: Definition: the finite time scale that subdivides each calendar day into 1440 .minute of day. Necessity: The granularity of the day of minutes scale is .minute.. with: Definition: the finite time scale that has granularity 1 minute and that has cardinality 1440 and that exactly subdivides 'Gregorian calendar day' Necessity: Each calendar day has 1440 of 'minute of day'. 14. In clause 13.1, in the entry for day of seconds scale, REPLACE the Definition, the following Note and the first Necessity: Definition: the finite time scale that subdivides each calendar day into at most 86401 .second of day. Note: The second of day with index 86400 corresponds only to leap seconds. Because of leap seconds, a calendar day can have a duration of 86399, 86400, or 86401 seconds. Necessity: The granularity of the day of seconds scale is .second.. with: Definition: the finite time scale that has granularity 1 second and that has cardinality 86400 and that exactly subdivides 'Gregorian calendar day' Necessity: Each calendar day has 86400 of 'second of day'. 15. In clause 13.1, in the entry for hour of minutes scale, REPLACE the Definition and the first Necessity: Definition: the finite time scale that subdivides each minute of hour into 60 .minute of hour. Necessity: The granularity of the hour of minutes scale is .minute.. with: Definition: the finite time scale that has granularity 1 minute and that has cardinality 60 and that exactly subdivides 'minute of hour' Necessity: Each hour of day has 60 of 'minute of hour'. 16. In clause 13.1, in the entry for minute of seconds scale, REPLACE the Definition, the following Note and the first Necessity: Definition: the finite time scale that subdivides each minute of day into at most 61 .second of minute. Note: The second of minute with index 60 corresponds only to leap seconds. Because of leap seconds, a minute of hour 59 can have a duration of 59, 60, or 61 seconds. Necessity: The granularity of the minute of seconds scale is .second.. with: Definition: the finite time scale that has granularity 1 second and that has cardinality 60 and that exactly subdivides 'minute of hour' Necessity: Each minute of hour has 60 of 'second of minute'. Disposition: Resolved To: date-time-ftf@omg.org Subject: Date-Time Issue 17428: basic time coordinate concepts are badly described X-KeepSent: 8AE3893A:3BDBAA69-85257A6D:006594D5; type=4; name=$KeepSent X-Mailer: Lotus Notes Release 8.5.3 September 15, 2011 From: Mark H Linehan Date: Sun, 2 Sep 2012 20:01:25 -0400 X-MIMETrack: Serialize by Router on D01MC604/01/M/IBM(Release 8.5.3FP2IF1|July 25, 2012) at 09/02/2012 20:01:26 X-Content-Scanned: Fidelis XPS MAILER x-cbid: 12090300-6078-0000-0000-00000EF5ACCF Here is an updated resolution proposal: Ed - this needs new versions of figures 10.6 and 10.8. -------------------------------- Mark H. Linehan STSM, IBM Research Date-Time Issue 17428 - basic time coordinate concepts are badly described1.doc Disposition: Resolved OMG Issue No: 17428 Title: basic time coordinate concepts are badly described Source: Ed Barkmeyer, NIST, edbark@nist.gov Summary: Figure 12.1 shows 'time coordinate' as a specialization of SBVR 'representation', which is said to be the relationship between an expression and the meaning it denotes. No UML diagram shows 'time expression' (from 12.1.2), which one would expect to be the corresponding specialization of 'expression'. In 12.1, the definition of 'time coordinate' is: "representation of a time point by an atomic time coordinate or compound time coordinate that indicates the time point". This is both circular and incorrect. A time coordinate cannot represent a time point by a subtype of itself, it must represent a time point by an expression. The definition should read: "representation of a time point by a time expression" (to exclude representation by a definite description). One would expect a 'time expression' to be "the expression of a time coordinate" (from SBVR 'representation has expression', mislabeled 'expression uses representation' in the diagram), but it isn't. It is said to be an expression involving an index (integer) and a time scale, which means it is not the expression of a compound time coordinate. In 12.1.2, the definition of 'atomic time coordinate indicates time point' requires the time expression to contain an index, which means that the "February" example is invalid. The expression (string) "February" includes neither an index nor a time scale. The time coordinate, as a 'representation', is the association of "February" with the time point that is month of year 2. The atomic time coordinate thus acquires the time scale and index properties from the time point it indicates. But the expression "February" does not have those properties. "February" is associated with that time point via 'concept has designation'. If DTV is to explain how a time expression is associated with a time point, it has to distinguish the properties of the expression from the properties of the association (that SBVR calls a 'representation'). A 'simple time expression' is either the expression of an 'index', which would be some expression of an integer, OR a 'signifier' for a time point (from SBVR 'concept has designation' and 'designation has signifier (expression)'). The signifier could be a given name, like "March", or a constructed term involving the scale and the index, like "Gregorian month 3". (What 12.1.2 describes is only the last case.) A 'compound time expression' is some syntax that combines two or more simple time expressions. In SBVR style, then, an atomic time coordinate is a time coordinate whose expression is a simple time expression, and a compound time coordinate is a time coordinate whose expression is a compound time expression. And the rules for determining what a simple time expression indicates, i.e., how the link that is the atomic time coordinate is constructed, depend on the nature of the time expression. In particular, the context of a compound time expression may make the intent of an index expression clear, as in "3/31/2012", where the "3" is only recognized as a month of year index because of its placement. Resolution: The relationships among .time coordinate., .atomic time coordinate., and .compound time coordinate. are disentangled to say that: . .time coordinate., .time expression., and related concepts are not .expressions. in the SBVR sense . .time coordinate. is an .atomic time coordinate. or a .compound time coordinate. . .atomic time coordinate. is either a time point formation or an individual concept of a time point . .time point expression. is renamed .time point formation. to avoid the idea that it is an SBVR .expression. . .compound time coordinate. is redefined as a .time coordinate that combines more than one atomic time coordinate. . the definition of .atomic time coordinate indicates time point. is rewritten to explain how it works in the two cases where the atomic time coordinate is a time point formation versus an individual concept of a time point kind Other miscellaneous changes: . The text at the start of clause 10.5.1 is instead made the introductory text for all of 10.5. . References to the clauses where time coordinates are defined are fixed. . A verb concept .time coordinate has time scale. is added to support the definitions of .absolute time coordinate. and .relative time coordinate. . A verb concept .time coordinate has expression. is added to make it clear that a time coordinate is NOT an expression, and in fact that there may be multiple expressions of a time coordinate . The definitions of .absolute time coordinate. and .relative time coordinate. are corrected . Missing SBVR concepts are added to clause 4 The issue summary made the comment that .the context of a compound time expression may make the intent of an index expression clear, as in "3/31/2012", where the "3" is only recognized as a month of year index because of its placement.. This specification does not describe the internal structure of a time coordinate, NOT the external representation of one. The interpretation of a date such as .3/31/2012. is a function of a tool, not of this specification. Revised Text: All changes described here are with respect to the beta-2 document. In clause 4, add these glossary entries in the appropriate alphabetical positions: expression expression represents meaning In clause 4, replace this glossary: representation uses expression . with representation has expression In clause 10, move the heading .10.5.1 General. to just before figure 10.6, so that the text between heading 10.5.1 and figure 10.6 becomes text directly under the clause 10 heading. In existing clause 10.5.1, replace the third sentence of paragraph 1, which reads: This clause specifies which combinations of atomic time coordinates form legitimate compound time coordinates. . with: Clauses 11, 12, and 13 specify which combinations of atomic time coordinates form legitimate compound time coordinates. In clause 10.5.1, replace figure 10.6 to match the new definition of .time coordinate. (below). In clause 10.5.1, replace the Definition and Reference of .time coordinate., which read: Definition: representation of a time point by an atomic time coordinate or compound time coordinate that indicates the time point Reference Scheme: the indices and time scales of the time coordinate . with: Definition: atomic time coordinate or compound time coordinate Reference Scheme: an expression of the time coordinate Delete the following Note in the glossary entry for .time coordinate.: Note: Named time coordinates, such as .Tuesday,. are defined in terms of an index and a time scale. Per SBVR, named time coordinates are individual concepts because their extension contains one thing: a particular time point on a particular time scale. In clause 10.5.1, replace the final Note under .time coordinate., which reads: Note: Particular kinds of time coordinates are defined below. . with: Note: Particular kinds of time coordinates are defined in clauses 11, 12, and 13. In clause 10.5.1, add the following glossary entry immediately after .time coordinate.: time coordinate has expression Definition: the expression communicates the time coordinate Possibility: A time coordinate has more than one expression. Example: .February 2012. and .2012 February. and .Féiere 2012. are three different expressions for the same time coordinate. In clause 10.5.1, replace the definition of .time coordinate indicates time point., which reads: Definition: time point is the scale point of the time scales of the time coordinate that has the indices of the time coordinate . with: Definition: the time coordinate identifies the time point In clause 10.5.1, add the following after the glossary entry for .time coordinate indicates time point.: time coordinate has time scale Definition: the time scale is the time scale of the time point that is indicated by the time coordinate Example: The time scale of .Tuesday. is the week of days scale. Example: The time scale of .2012. is the Gregorian years scale. Example: The time scale of .12 August 2012. is the Gregorian days scale. In clause 10.5.2, replace the definition of .absolute time coordinate., which reads: Definition: time coordinate that has an indefinite time scale and refers to exactly one time interval . with: Definition: time coordinate that has an indefinite time scale and that indicates a time point that corresponds to exactly one time interval In clause 10.5.2, replace the definition of .relative time coordinate., which reads: Definition: time coordinate that does not have an indefinite time scale and refers to more than one time interval . with: Definition: time coordinate that has a finite time scale and that indicates a time point that corresponds to more than one time interval In clause 10.5.3, replace figure 10.8 with this version: In clause 10.5.3, replace the glossary entries for .time expression., .time expression has index. and .time expression has time scale. with: time formation Definition: specification of a time point by an index and a time scale Example: .month 1. specifies the time point that has index 1 on the Gregorian year of months scale time formation has index Definition: the index indexes a sequence position in a time scale Note: Each time formation has exactly one index. Note: .Index. is defined in Annex D.2 as a role of an integer. Necessity: The assignment of an index to a time point depends upon the index origin value of the time scale. Hours of day, minutes of hour, and seconds of minute are conventionally numbered from 0, while calendar months, calendar weeks, and calendar days are numbered from 1. Calendar years are numbered from 1 in the Gregorian calendar, but are numbered from 0 by [ISO 8601]. Example: The index .10. of the time formation .day 10.. time formation has time scale Definition: the time formation specifies a time point that is on the time scale Necessity: Each time formation has exactly one time scale. Note: The expression of the time formation may imply, rather than explicitly specify, the time scale. Example: The time scale .Gregorian years scale. that is meant by the atomic time coordinate .2010.. In clause 10.5.3, replace the Definition of .atomic time coordinate., which reads: Definition: time coordinate that represents a time point by exactly one time expression . with: Definition: time formation, or individual concept of a time point In clause 10.5.3, replace the Reference Scheme for .atomic time coordinate.: Reference Scheme: The index of the time expression and the time scale of the time expression. . with: Reference Scheme: An expression of the atomic time coordinate. In clause 10.5.3, replace the Definition of .atomic time coordinate indicates time point., which reads: Definition: the index of the time expression of the atomic time coordinate is the index of the time point on the time scale of the time point . with a General Concept and a new Definition: General Concept: time coordinate indicates time point Definition: if the atomic time coordinate is a time formation, then the time point has the index of the time formation on the time scale of the time formation, and if the atomic time coordinate is an individual concept, then the individual concept corresponds to the time point In clause 10.5.3, replace the Definition of .compound time coordinate., which reads: Definition: time coordinate that represents a time point by more than one time expression . with: Definition: time coordinate that combines more than one atomic time coordinate In clause 10.5.3, delete the first Example under .compound time coordinate., which reads: Example: "2010" is a precise atomic time coordinate indicating a Gregorian year on the Gregorian years scale. In clause 10.5.3, delete the glossary entry for .compound time coordinate has atomic time coordinate. because it duplicates the entry for .compound time coordinate combines atomic time coordinate.. Disposition: Resolved To: date-time-ftf@omg.org Subject: DTV Issue 17428 - Basic time coordinate concepts are badly described X-KeepSent: D1FC1853:746E329D-85257A8F:007D2645; type=4; name=$KeepSent X-Mailer: Lotus Notes Release 8.5.3 September 15, 2011 From: Mark H Linehan Date: Sat, 6 Oct 2012 18:47:55 -0400 X-MIMETrack: Serialize by Router on D01MC604/01/M/IBM(Release 8.5.3FP2IF1|July 25, 2012) at 10/06/2012 18:47:56 x-cbid: 12100622-7182-0000-0000-000002C2481E Here's another attempt for this issue, based on our last discussion: ----------------------------- Mark H. Linehan STSM, IBM Research Date-Time Issue 17428 - basic time coordinate concepts are badly described2.doc Disposition: Resolved OMG Issue No: 17428 Title: basic time coordinate concepts are badly described Source: Ed Barkmeyer, NIST, edbark@nist.gov Summary: Figure 12.1 shows 'time coordinate' as a specialization of SBVR 'representation', which is said to be the relationship between an expression and the meaning it denotes. No UML diagram shows 'time expression' (from 12.1.2), which one would expect to be the corresponding specialization of 'expression'. In 12.1, the definition of 'time coordinate' is: "representation of a time point by an atomic time coordinate or compound time coordinate that indicates the time point". This is both circular and incorrect. A time coordinate cannot represent a time point by a subtype of itself, it must represent a time point by an expression. The definition should read: "representation of a time point by a time expression" (to exclude representation by a definite description). One would expect a 'time expression' to be "the expression of a time coordinate" (from SBVR 'representation has expression', mislabeled 'expression uses representation' in the diagram), but it isn't. It is said to be an expression involving an index (integer) and a time scale, which means it is not the expression of a compound time coordinate. In 12.1.2, the definition of 'atomic time coordinate indicates time point' requires the time expression to contain an index, which means that the "February" example is invalid. The expression (string) "February" includes neither an index nor a time scale. The time coordinate, as a 'representation', is the association of "February" with the time point that is month of year 2. The atomic time coordinate thus acquires the time scale and index properties from the time point it indicates. But the expression "February" does not have those properties. "February" is associated with that time point via 'concept has designation'. If DTV is to explain how a time expression is associated with a time point, it has to distinguish the properties of the expression from the properties of the association (that SBVR calls a 'representation'). A 'simple time expression' is either the expression of an 'index', which would be some expression of an integer, OR a 'signifier' for a time point (from SBVR 'concept has designation' and 'designation has signifier (expression)'). The signifier could be a given name, like "March", or a constructed term involving the scale and the index, like "Gregorian month 3". (What 12.1.2 describes is only the last case.) A 'compound time expression' is some syntax that combines two or more simple time expressions. In SBVR style, then, an atomic time coordinate is a time coordinate whose expression is a simple time expression, and a compound time coordinate is a time coordinate whose expression is a compound time expression. And the rules for determining what a simple time expression indicates, i.e., how the link that is the atomic time coordinate is constructed, depend on the nature of the time expression. In particular, the context of a compound time expression may make the intent of an index expression clear, as in "3/31/2012", where the "3" is only recognized as a month of year index because of its placement. Resolution: The relationships among .time coordinate., .atomic time coordinate., and .compound time coordinate. are disentangled to say that: . .time coordinate., .time expression., and related concepts are not .expressions. in the SBVR sense . .time coordinate. is an .atomic time coordinate. or a .compound time coordinate. . .atomic time coordinate. is either a time point formation or an individual concept of a time point . .time point expression. is renamed .time point formation. to avoid the idea that it is an SBVR .expression. . .compound time coordinate. is redefined as a .time coordinate that combines more than one atomic time coordinate. . the definition of .atomic time coordinate indicates time point. is rewritten to explain how it works in the two cases where the atomic time coordinate is a time point formation versus an individual concept of a time point kind Other miscellaneous changes: . The text at the start of clause 10.5.1 is instead made the introductory text for all of 10.5. . References to the clauses where time coordinates are defined are fixed. . A verb concept .time coordinate has time scale. is added to support the definitions of .absolute time coordinate. and .relative time coordinate. . A verb concept .time coordinate has expression. is added to make it clear that a time coordinate is NOT an expression, and in fact that there may be multiple expressions of a time coordinate . The definitions of .absolute time coordinate. and .relative time coordinate. are corrected . Missing SBVR concepts are added to clause 4 The issue summary made the comment that .the context of a compound time expression may make the intent of an index expression clear, as in "3/31/2012", where the "3" is only recognized as a month of year index because of its placement.. This specification does not describe the internal structure of a time coordinate, NOT the external representation of one. The interpretation of a date such as .3/31/2012. is a function of a tool, not of this specification. Note: this issue is dependent upon 16951, which defines 'time point kind'. Revised Text: All changes described here are with respect to the beta-2 document. 1. In clause 4, add these glossary entries in the appropriate alphabetical positions: expression expression represents meaning 2. In clause 4, replace this glossary entry: representation uses expression . with representation has expression 3. In clause 8.5, add three additional Reference Schemes after the existing Reference Schemes in the Glossary Entry for 'time point': Reference Scheme: the time scale of the time point and the index of the time point Reference Scheme: a time point kind and an index Reference Scheme: the name of a time point 4. In clause 10, move the heading .10.5.1 General. to just before figure 10.6, so that the text between heading 10.5.1 and figure 10.6 becomes text directly under the clause 10 heading. 5. In existing clause 10.5.1, replace the third sentence of paragraph 1, which reads: This clause specifies which combinations of atomic time coordinates form legitimate compound time coordinates. . with: Clauses 11, 12, and 13 specify which combinations of atomic time coordinates form legitimate compound time coordinates. 6. In clause 10.5.1, replace figure 10.6 with: <> 7. In clause 10.5.1, replace the glossary entry for .time coordinate., which reads: time coordinate Synonym: time stamp Definition: representation of a time point by an atomic time coordinate or compound time coordinate that indicates the time point Reference Scheme: the indices and time scales of the time coordinate Example: January 2009, 2009 month 1 Note: Time coordinates may be either atomic or compound (see clause .12.1.2). Note: Time coordinates may be given by an index and an implied time scale (e.g."month 1"), or by a name (e.g. "January"). Note: Named time coordinates, such as 'Tuesday', are defined in terms of an index and a time scale. Per SBVR, named time coordinates are individual concepts because their extension contains one thing: a particular time point on a particular time scale. Note: Particular kinds of time coordinates are defined below, in clauses .12.3.1, .12.3.2, and .12.3.3. . with: time coordinate Synonym: time stamp Definition: representation of a time point Reference Scheme: an expression of the time coordinate Example: January 2009, 2009 month 1, 2009 Note: Time coordinates may be either absolute or relative (see clause 10.5.2). Note: Time coordinates may be either atomic or compound (see clause 10.5.3). Note: Particular kinds of time coordinates are defined below, in clauses 11, 12, and 13. 8. In clause 10.5.1, replace the glossary entry for .time coordinate indicates time point., which reads: time coordinate indicates time point Definition: time point is the scale point of the time scales of the time coordinate that has the indices of the time coordinate Necessity: Each time point is indicated by at least one time coordinate. Necessity: Each time coordinate indicates exactly one time point. Possibility: A time point is indicated by more than one time coordinate. Note: See 'atomic time coordinate indicates time point' and 'compound time coordinate indicates time point' for definitions of exactly how a time coordinate indicates a time point. . with: time coordinate indicates time point See: representation represents meaning Note: See 'compound time coordinate indicates time point' for definitions of exactly how a compound time coordinate indicates a time point. 9. In clause 10.5.2, replace the definition of .absolute time coordinate., which reads: Definition: time coordinate that has an indefinite time scale and refers to exactly one time interval . with: Definition: time coordinate that has an indefinite time scale Necessity: Each absolute time coordinate corresponds to exactly one time interval. 10. In clause 10.5.2, replace the definition of .relative time coordinate., which reads: Definition: time coordinate that does not have an indefinite time scale and refers to more than one time interval . with: Definition: time coordinate that has a finite time scale Necessity: Each relative time coordinate corresponds to more than one time interval. 11. In clause 10.5.3, replace figure 10.8 with this version: <> 12. In clause 10.5.3, delete the glossary entries for .time expression., .time expression has index. and .time expression has time scale.. 13. In clause 10.5.3, replace the glossary entry for .atomic time coordinate., which reads: atomic time coordinate Definition: time coordinate that represents a time point by exactly one time expression Reference Scheme: The index of the time expression and the time scale of the time expression. Example: 2010, day 53, Tuesday . with: atomic time coordinate Definition: time coordinate that has an expression that is the name of a time point, or that is a time point kind and an index Note: The two possible forms for the expression of an atomic time coordinate are based on two of the reference schemes for a time point. Expressions of these forms directly represent time points. Example: 2010, day 53, Tuesday 14. In clause 10.5.3, delete the glossary entry for .atomic time coordinate indicates time point.. This concept is not needed because an atomic coordinate directly represents a time point. atomic time coordinate indicates time point Synonymous Form: time point indicated by atomic time coordinate Definition: the index of the time expression of the atomic time coordinate is the index of the time point on the time scale of the time point Note: Each kind of atomic time coordinate defines the kind of time point that it indicates. Example: "week 15", indicating the 15th week of year, which is on the year of weeks scale. Note: Some time points are given by individual concepts that are defined in terms of indices and time points. Example: "February" indicates the time point that is Gregorian month of year 2 on the Gregorian year of months scale 15. In clause 10.5.3, replace the Definition of .compound time coordinate., which reads: Definition: time coordinate that represents a time point by more than one time expression . with: Definition: time coordinate that combines more than one atomic time coordinate 16. In clause 10.5.3, delete the first Example under .compound time coordinate., which reads: Example: "2010" is a precise atomic time coordinate indicating a Gregorian year on the Gregorian years scale. 17. In clause 10.5.3, delete the glossary entry for .compound time coordinate has atomic time coordinate. because it duplicates the entry for .compound time coordinate combines atomic time coordinate.. Disposition: Resolved To: date-time-ftf@omg.org Subject: DTV Issue 17428 - Basic time coordinate concepts are badly described X-KeepSent: D1FC1853:746E329D-85257A8F:007D2645; type=4; name=$KeepSent X-Mailer: Lotus Notes Release 8.5.3 September 15, 2011 From: Mark H Linehan Date: Sat, 6 Oct 2012 18:47:55 -0400 X-MIMETrack: Serialize by Router on D01MC604/01/M/IBM(Release 8.5.3FP2IF1|July 25, 2012) at 10/06/2012 18:47:56 x-cbid: 12100622-7182-0000-0000-000002C2481E Here's another attempt for this issue, based on our last discussion: ----------------------------- Mark H. Linehan STSM, IBM Research Date-Time Issue 17428 - basic time coordinate concepts are badly described2.doc To: date-time-ftf@omg.org Subject: DTV Issue 17428 - basic time coordinate concepts are badly described X-KeepSent: 821B03E8:954BE109-85257A91:0073079B; type=4; name=$KeepSent X-Mailer: Lotus Notes Release 8.5.3 September 15, 2011 From: Mark H Linehan Date: Mon, 8 Oct 2012 16:57:26 -0400 X-MIMETrack: Serialize by Router on D01MC604/01/M/IBM(Release 8.5.3FP2IF1|July 25, 2012) at 10/08/2012 16:57:28 x-cbid: 12100820-9360-0000-0000-00000B7D0068 This is a minor update that corrects a couple of examples in the version I sent out over the weekend. It still needs new figures. ----------------------------- Mark H. Linehan STSM, IBM Research Date-Time Issue 17428 - basic time coordinate concepts are badly described3.doc Disposition: Resolved OMG Issue No: 17428 Title: basic time coordinate concepts are badly described Source: Ed Barkmeyer, NIST, edbark@nist.gov Summary: Figure 12.1 shows 'time coordinate' as a specialization of SBVR 'representation', which is said to be the relationship between an expression and the meaning it denotes. No UML diagram shows 'time expression' (from 12.1.2), which one would expect to be the corresponding specialization of 'expression'. In 12.1, the definition of 'time coordinate' is: "representation of a time point by an atomic time coordinate or compound time coordinate that indicates the time point". This is both circular and incorrect. A time coordinate cannot represent a time point by a subtype of itself, it must represent a time point by an expression. The definition should read: "representation of a time point by a time expression" (to exclude representation by a definite description). One would expect a 'time expression' to be "the expression of a time coordinate" (from SBVR 'representation has expression', mislabeled 'expression uses representation' in the diagram), but it isn't. It is said to be an expression involving an index (integer) and a time scale, which means it is not the expression of a compound time coordinate. In 12.1.2, the definition of 'atomic time coordinate indicates time point' requires the time expression to contain an index, which means that the "February" example is invalid. The expression (string) "February" includes neither an index nor a time scale. The time coordinate, as a 'representation', is the association of "February" with the time point that is month of year 2. The atomic time coordinate thus acquires the time scale and index properties from the time point it indicates. But the expression "February" does not have those properties. "February" is associated with that time point via 'concept has designation'. If DTV is to explain how a time expression is associated with a time point, it has to distinguish the properties of the expression from the properties of the association (that SBVR calls a 'representation'). A 'simple time expression' is either the expression of an 'index', which would be some expression of an integer, OR a 'signifier' for a time point (from SBVR 'concept has designation' and 'designation has signifier (expression)'). The signifier could be a given name, like "March", or a constructed term involving the scale and the index, like "Gregorian month 3". (What 12.1.2 describes is only the last case.) A 'compound time expression' is some syntax that combines two or more simple time expressions. In SBVR style, then, an atomic time coordinate is a time coordinate whose expression is a simple time expression, and a compound time coordinate is a time coordinate whose expression is a compound time expression. And the rules for determining what a simple time expression indicates, i.e., how the link that is the atomic time coordinate is constructed, depend on the nature of the time expression. In particular, the context of a compound time expression may make the intent of an index expression clear, as in "3/31/2012", where the "3" is only recognized as a month of year index because of its placement. Resolution: The definition of .time coordinate. is simplified to say just that it is a representation of a time point. To support this: . .time expression. and related concepts are deleted . .atomic time coordinate. is a time coordinate that has an expression that is either the name of a time point or a time point kind and an index . .compound time coordinate. is redefined as a .time coordinate that combines more than one atomic time coordinate. . .atomic time coordinate indicates time point. is identified as a synonymous form of SBVR's 'representation represents meaning' . 'compound time coordinate indicates time point' is redefined in terms of the existing verb concept 'compound time coordinate combines atomic time point' Other miscellaneous changes: . Several new reference schemes are added to 'time point' . The text at the start of clause 10.5.1 is instead made the introductory text for all of 10.5. . References to the clauses where time coordinates are defined are fixed. . The definitions of .absolute time coordinate. and .relative time coordinate. are corrected . Missing SBVR concepts are added to clause 4 The issue summary made the comment that .the context of a compound time expression may make the intent of an index expression clear, as in "3/31/2012", where the "3" is only recognized as a month of year index because of its placement.. This specification does not describe the internal structure of a time coordinate, NOT the external representation of one. The interpretation of a date such as .3/31/2012. is a function of a tool, not of this specification. Note: this issue is dependent upon 16951, which defines 'time point kind'. Revised Text: All changes described here are with respect to the beta-2 document. 1. In clause 4, add these glossary entries in the appropriate alphabetical positions: expression expression represents meaning 2. In clause 4, replace this glossary entry: representation uses expression . with representation has expression 3. In clause 8.5, add three additional Reference Schemes after the existing Reference Schemes in the Glossary Entry for 'time point': Reference Scheme: the time scale of the time point and the index of the time point Reference Scheme: a time point kind and an index Reference Scheme: the name of a time point 4. In clause 10, move the heading .10.5.1 General. to just before figure 10.6, so that the text between heading 10.5.1 and figure 10.6 becomes text directly under the clause 10 heading. 5. In existing clause 10.5.1, replace the third sentence of paragraph 1, which reads: This clause specifies which combinations of atomic time coordinates form legitimate compound time coordinates. . with: Clauses 11, 12, and 13 specify which combinations of atomic time coordinates form legitimate compound time coordinates. 6. In clause 10.5.1, replace figure 10.6 with: <> 7. In clause 10.5.1, replace the glossary entry for .time coordinate., which reads: time coordinate Synonym: time stamp Definition: representation of a time point by an atomic time coordinate or compound time coordinate that indicates the time point Reference Scheme: the indices and time scales of the time coordinate Example: January 2009, 2009 month 1 Note: Time coordinates may be either atomic or compound (see clause .12.1.2). Note: Time coordinates may be given by an index and an implied time scale (e.g."month 1"), or by a name (e.g. "January"). Note: Named time coordinates, such as 'Tuesday', are defined in terms of an index and a time scale. Per SBVR, named time coordinates are individual concepts because their extension contains one thing: a particular time point on a particular time scale. Note: Particular kinds of time coordinates are defined below, in clauses .12.3.1, .12.3.2, and .12.3.3. . with: time coordinate Synonym: time stamp Definition: representation of a time point Reference Scheme: an expression of the time coordinate Example: January 2009, 2009 month 1, 2009 Note: Time coordinates may be either absolute or relative (see clause 10.5.2). Note: Time coordinates may be either atomic or compound (see clause 10.5.3). Note: Particular kinds of time coordinates are defined below, in clauses 11, 12, and 13. 8. In clause 10.5.1, replace the glossary entry for .time coordinate indicates time point., which reads: time coordinate indicates time point Definition: time point is the scale point of the time scales of the time coordinate that has the indices of the time coordinate Necessity: Each time point is indicated by at least one time coordinate. Necessity: Each time coordinate indicates exactly one time point. Possibility: A time point is indicated by more than one time coordinate. Note: See 'atomic time coordinate indicates time point' and 'compound time coordinate indicates time point' for definitions of exactly how a time coordinate indicates a time point. . with: time coordinate indicates time point See: representation represents meaning Note: See 'compound time coordinate indicates time point' for definitions of exactly how a compound time coordinate indicates a time point. 9. In clause 10.5.2, replace the definition of .absolute time coordinate., which reads: Definition: time coordinate that has an indefinite time scale and refers to exactly one time interval . with: Definition: time coordinate that has an indefinite time scale Necessity: Each absolute time coordinate corresponds to exactly one time interval. 10. In clause 10.5.2, replace the definition of .relative time coordinate., which reads: Definition: time coordinate that does not have an indefinite time scale and refers to more than one time interval . with: Definition: time coordinate that has a finite time scale Necessity: Each relative time coordinate corresponds to more than one time interval. 11. In clause 10.5.3, replace figure 10.8 with this version: <> 12. In clause 10.5.3, delete the glossary entries for .time expression., .time expression has index. and .time expression has time scale.. 13. In clause 10.5.3, replace the glossary entry for .atomic time coordinate., which reads: atomic time coordinate Definition: time coordinate that represents a time point by exactly one time expression Reference Scheme: The index of the time expression and the time scale of the time expression. Example: 2010, day 53, Tuesday . with: atomic time coordinate Definition: time coordinate that has an expression that is the name of a time point, or that is a time point kind and an index Note: The two possible forms for the expression of an atomic time coordinate are based on two of the reference schemes for a time point. Expressions of these forms directly represent time points. Example: Tuesday, calendar week 53, 2010 (understood as a reference to the time point kind Gregorian year and index '2010') 14. In clause 10.5.3, delete the glossary entry for .atomic time coordinate indicates time point.. This concept is not needed because an atomic coordinate directly represents a time point. atomic time coordinate indicates time point Synonymous Form: time point indicated by atomic time coordinate Definition: the index of the time expression of the atomic time coordinate is the index of the time point on the time scale of the time point Note: Each kind of atomic time coordinate defines the kind of time point that it indicates. Example: "week 15", indicating the 15th week of year, which is on the year of weeks scale. Note: Some time points are given by individual concepts that are defined in terms of indices and time points. Example: "February" is an expression of an atomic time coordinate that indicates the time point that is Gregorian month of year 2 on the Gregorian year of months scale 15. In clause 10.5.3, replace the Definition of .compound time coordinate., which reads: Definition: time coordinate that represents a time point by more than one time expression . with: Definition: time coordinate that combines more than one atomic time coordinate 16. In clause 10.5.3, delete the first Example under .compound time coordinate., which reads: Example: "2010" is a precise atomic time coordinate indicating a Gregorian year on the Gregorian years scale. 17. In clause 10.5.3, delete the glossary entry for .compound time coordinate has atomic time coordinate. because it duplicates the entry for .compound time coordinate combines atomic time coordinate.. Disposition: Resolved