Issue 17425: rename 'calendar date' (date-time-ftf) Source: NIST (Mr. Edward J. Barkmeyer, edbark(at)nist.gov) Nature: Uncategorized Issue Severity: Summary: DTV section 12.3.4 defines 'calendar date' as: "Gregorian year month date coordinate or Gregorian day of year coordinate or year weekday coordinate". This is clearly a definition of 'Gregorian calendar date' or 'Gregorian date'. The general concept 'calendar date' is "absolute time coordinate that indicates a time point that corresponds to exactly one calendar day". The list given in the definition is just the ones defined for Gregorian calendars. The general concept 'calendar date' is useful, and the term should not be assigned to Gregorian dates only. As defined, a 'date time' (date and time) also requires a Gregorian date, but it is not clear that there is a useful generalization. Recommendation: Rename 'calendar date' to 'Gregorian date' or something the like, and define the general concept 'calendar date' as well. Resolution: Revised Text: Actions taken: June 13, 2012: received issue Discussion: End of Annotations:===== te: Wed, 13 Jun 2012 16:16:19 -0400 From: Ed Barkmeyer Reply-To: Organization: NIST User-Agent: Thunderbird 2.0.0.23 (Windows/20090812) To: "issues@omg.org" Subject: DTV Issue: rename 'calendar date' X-NISTMEL-MailScanner-Information: Please contact postmaster@mel.nist.gov for more information X-NISTMEL-MailScanner-ID: q5DKGPkQ031058 X-NISTMEL-MailScanner: Found to be clean X-NISTMEL-MailScanner-SpamCheck: X-NISTMEL-MailScanner-From: edbark@nist.gov X-NISTMEL-MailScanner-Watermark: 1340223386.10699@bBdQHeKecGyDdo/7ZLLIUA X-Spam-Status: No Specification: Date Time Vocabulary Version: Beta-1 Title: rename 'calendar date' Source: Ed Barkmeyer, NIST, edbark@nist.gov Summary: DTV section 12.3.4 defines 'calendar date' as: "Gregorian year month date coordinate or Gregorian day of year coordinate or year weekday coordinate". This is clearly a definition of 'Gregorian calendar date' or 'Gregorian date'. The general concept 'calendar date' is "absolute time coordinate that indicates a time point that corresponds to exactly one calendar day". The list given in the definition is just the ones defined for Gregorian calendars. The general concept 'calendar date' is useful, and the term should not be assigned to Gregorian dates only. As defined, a 'date time' (date and time) also requires a Gregorian date, but it is not clear that there is a useful generalization. Recommendation: Rename 'calendar date' to 'Gregorian date' or something the like, and define the general concept 'calendar date' as well. -- Edward J. Barkmeyer Email: edbark@nist.gov National Institute of Standards & Technology Manufacturing Systems Integration Division 100 Bureau Drive, Stop 8263 Tel: +1 301-975-3528 Gaithersburg, MD 20899-8263 Cel: +1 240-672-5800 Date: Mon, 2 Jul 2012 19:53:19 -0400 From: Ed Barkmeyer Reply-To: Organization: NIST User-Agent: Thunderbird 2.0.0.23 (Windows/20090812) To: OMG DateTimeVoc FTF Subject: Issue 17425 and 17429 calendar date X-NISTMEL-MailScanner-Information: Please contact postmaster@mel.nist.gov for more information X-NISTMEL-MailScanner-ID: q62NrOe9013493 X-NISTMEL-MailScanner: Found to be clean X-NISTMEL-MailScanner-SpamCheck: X-NISTMEL-MailScanner-From: edbark@nist.gov X-NISTMEL-MailScanner-Watermark: 1341878005.85376@wSAzq0eKlTZWo0x/fbVS8A X-Spam-Status: No I attach a draft resolution, reflecting discussions on June 25 and July 2. This is a rewrite of 12.3.4 in anticipation of its being moved bodily into the new Calendars section. There are two comments in this writeup. One is a synchronization issue. The other is an ugly solution to a problem in expressing the definition of date time. There is also a missing comment: It seems we need absolute time scales of hours, minutes and seconds for date time coordinates to have a time point to indicate. -Ed -- Edward J. Barkmeyer Email: edbark@nist.gov National Institute of Standards & Technology Manufacturing Systems Integration Division 100 Bureau Drive, Stop 8263 Tel: +1 301-975-3528 Gaithersburg, MD 20899-8263 Cel: +1 240-672-5800 Issue 17429-definition of calendar date1.docx To: Cc: OMG DateTimeVoc FTF Subject: Re: Issue 17425 and 17429 calendar date X-KeepSent: 686ED73A:37AD1196-85257A30:00432F74; type=4; name=$KeepSent X-Mailer: Lotus Notes Release 8.5.3 September 15, 2011 From: Mark H Linehan Date: Tue, 3 Jul 2012 14:01:38 -0400 X-MIMETrack: Serialize by Router on D01MC604/01/M/IBM(Release 8.5.3 ZX853HP5|January 12, 2012) at 07/03/2012 14:01:42, Serialize complete at 07/03/2012 14:01:42 X-Content-Scanned: Fidelis XPS MAILER x-cbid: 12070318-7182-0000-0000-000001E94053 Regarding 17425: * The definition of 'Gregorian date' should drop '... or year weekday coordinate'. * We should be consistent about stylizing in text paragraphs, notes, etc. * Should the Note under 'time of day' really be a Necessity or a Definition? * We do not currently have 'time of day time point'. Suggest we add it at the end of 9.5.2 as 'hour of day or minute of day or second of day'. * The 'day of hours scale' (etc.) should 'subdivide calendar day -- but they don't. Suggest you change this in the same issue. * We do not have any absolute time scales of hours, minutes, or seconds. Reason (1): times of day are always computed with respect to midnight of a given day, not on any absolute time scale; (2) if we have such absolute time scales then we need to work out the conversion of date-times to those scales, and that's really ugly considering leap days. Net: lets not go there. * Instead, I suggest that 'date time' should be defined as indicating the time of day time point indicated by the time coordinate 'within' the calendar day indicated by the date coordinate. With this solution we need a 'time interval corresponds to time of day time point within calendar day' verb concept which we could define as 'the time interval that corresponds to the time of day time point starts some duration after midnight on the calendar day and the duration is the index of the time of day time point * the granularity of the time scale of the time of day time point' * Suggest having notes for 'time of day' and 'calendar date' that identify the subtype time coordinates corresponding to each one. -------------------------------- Mark H. Linehan STSM, IBM Research From: Ed Barkmeyer To: OMG DateTimeVoc FTF , Date: 07/02/2012 07:55 PM Subject: Issue 17425 and 17429 calendar date -------------------------------------------------------------------------------- I attach a draft resolution, reflecting discussions on June 25 and July 2. This is a rewrite of 12.3.4 in anticipation of its being moved bodily into the new Calendars section. There are two comments in this writeup. One is a synchronization issue. The other is an ugly solution to a problem in expressing the definition of date time. There is also a missing comment: It seems we need absolute time scales of hours, minutes and seconds for date time coordinates to have a time point to indicate. -Ed -- Edward J. Barkmeyer Email: edbark@nist.gov National Institute of Standards & Technology Manufacturing Systems Integration Division 100 Bureau Drive, Stop 8263 Tel: +1 301-975-3528 Gaithersburg, MD 20899-8263 Cel: +1 240-672-5800 [attachment "Issue 17429-definition of calendar date.docx" deleted by Mark H Linehan/Watson/IBM] [attachment "Issue 17425-calendar-date.docx" deleted by Mark H Linehan/Watson/IBM] Date: Tue, 3 Jul 2012 17:26:14 -0400 From: Edward Barkmeyer Reply-To: Organization: NIST User-Agent: Thunderbird 2.0.0.23 (Windows/20090812) To: Mark H Linehan CC: OMG DateTimeVoc FTF Subject: Re: Issue 17425 and 17429 calendar date Mark H Linehan wrote: Regarding 17425: * The definition of 'Gregorian date' should drop '... or year weekday coordinate'. But then, can the Weeks calendar add it back? Or do we just ignore it? I assume ISO 8601 permits it, but I have never seen a date so identified. * We should be consistent about stylizing in text paragraphs, notes, etc. OK. You will want to proofread the next draft. * Should the Note under 'time of day' really be a Necessity or a Definition? * We do not currently have 'time of day time point'. Suggest we add it at the end of 9.5.2 as 'hour of day or minute of day or second of day'. OK. That eliminates the need for the Note, which just defines the undefined term. * The 'day of hours scale' (etc.) should 'subdivide calendar day -- but they don't. Suggest you change this in the same issue. I think that is addressed in a separate issue resolution. Maybe we should combine them, especially if they result in changes to the same diagrams. * We do not have any absolute time scales of hours, minutes, or seconds. Reason (1): times of day are always computed with respect to midnight of a given day, not on any absolute time scale; (2) if we have such absolute time scales then we need to work out the conversion of date-times to those scales, and that's really ugly considering leap days. Net: lets not go there. I understand, but then a 'date time' is not clearly an 'absolute time coordinate'. It refers to exactly one time interval, but it does not indicate a time point on any identified time scale. This arose from trying to say what it indicates. It is, as the draft shows, easier to say what time interval a date time refers to. * Instead, I suggest that 'date time' should be defined as indicating the time of day time point indicated by the time coordinate 'within' the calendar day indicated by the date coordinate. With this solution we need a 'time interval corresponds to time of day time point within calendar day' verb concept which we could define as 'the time interval that corresponds to the time of day time point starts some duration after midnight on the calendar day and the duration is the index of the time of day time point * the granularity of the time scale of the time of day time point' I understand, but there are no time of day time points 'within' a calendar day. The time periods are within/during the day period. What we are really talking about is constructing a time point on an absolute scale by taking each calendar day and subdividing it into time points of some time-of-day granularity G. By doing this to every calendar day, you get a sequence of time points on an indefinite time scale with granularity G. * Suggest having notes for 'time of day' and 'calendar date' that identify the subtype time coordinates corresponding to each one. Good point. They can be forward pointers to the Gregorian and Standard time sections. -Ed -------------------------------- Mark H. Linehan STSM, IBM Research From: Ed Barkmeyer To: OMG DateTimeVoc FTF , Date: 07/02/2012 07:55 PM Subject: Issue 17425 and 17429 calendar date ------------------------------------------------------------------------ I attach a draft resolution, reflecting discussions on June 25 and July 2. This is a rewrite of 12.3.4 in anticipation of its being moved bodily into the new Calendars section. There are two comments in this writeup. One is a synchronization issue. The other is an ugly solution to a problem in expressing the definition of date time. There is also a missing comment: It seems we need absolute time scales of hours, minutes and seconds for date time coordinates to have a time point to indicate. -Ed -- Edward J. Barkmeyer Email: edbark@nist.gov National Institute of Standards & Technology Manufacturing Systems Integration Division 100 Bureau Drive, Stop 8263 Tel: +1 301-975-3528 Gaithersburg, MD 20899-8263 Cel: +1 240-672-5800 [attachment "Issue 17429-definition of calendar date.docx" deleted by Mark H Linehan/Watson/IBM] [attachment "Issue 17425-calendar-date.docx" deleted by Mark H Linehan/Watson/IBM] -- Edward J. Barkmeyer Email: edbark@nist.gov National Institute of Standards & Technology Manufacturing Systems Integration Division 100 Bureau Drive, Stop 8263 Tel: +1 301-975-3528 Gaithersburg, MD 20899-8263 Cel: +1 240-672-5800 "The opinions expressed above do not reflect consensus of NIST, and have not been reviewed by any Government authority." To: Cc: OMG DateTimeVoc FTF Subject: Re: Issue 17425 and 17429 calendar date X-KeepSent: 285F177E:0B0748B9-85257A31:00534A22; type=4; name=$KeepSent X-Mailer: Lotus Notes Release 8.5.3 September 15, 2011 From: Mark H Linehan Date: Wed, 4 Jul 2012 11:20:42 -0400 X-MIMETrack: Serialize by Router on D01MC604/01/M/IBM(Release 8.5.3 ZX853HP5|January 12, 2012) at 07/04/2012 11:20:45, Serialize complete at 07/04/2012 11:20:45 X-Content-Scanned: Fidelis XPS MAILER x-cbid: 12070415-7606-0000-0000-000001BBB421 Responses like this. Maybe we need to go forward with ballot 1 without 17425. -------------------------------- Mark H. Linehan STSM, IBM Research From: Edward Barkmeyer To: Mark H Linehan/Watson/IBM@IBMUS, Cc: OMG DateTimeVoc FTF Date: 07/03/2012 05:27 PM Subject: Re: Issue 17425 and 17429 calendar date -------------------------------------------------------------------------------- Mark H Linehan wrote: > Regarding 17425: > > * The definition of 'Gregorian date' should drop '... or year weekday > coordinate'. But then, can the Weeks calendar add it back? Or do we just ignore it? I assume ISO 8601 permits it, but I have never seen a date so identified. I believe I got that from 8601. And I believe I have seen it in real business applications. It doesn't have to be a "Gregorian Date" since it's a "calendar date" anyway. > * We should be consistent about stylizing in text paragraphs, notes, etc. OK. You will want to proofread the next draft. > * Should the Note under 'time of day' really be a Necessity or a > Definition? > * We do not currently have 'time of day time point'. Suggest we add > it at the end of 9.5.2 as 'hour of day or minute of day or second of > day'. OK. That eliminates the need for the Note, which just defines the undefined term. Good. > * The 'day of hours scale' (etc.) should 'subdivide calendar day -- > but they don't. Suggest you change this in the same issue. I think that is addressed in a separate issue resolution. Maybe we should combine them, especially if they result in changes to the same diagrams. Yes, certainly. Which issue? > * We do not have any absolute time scales of hours, minutes, or > seconds. Reason (1): times of day are always computed with respect to > midnight of a given day, not on any absolute time scale; (2) if we > have such absolute time scales then we need to work out the conversion > of date-times to those scales, and that's really ugly considering leap > days. Net: lets not go there. I understand, but then a 'date time' is not clearly an 'absolute time coordinate'. It refers to exactly one time interval, but it does not indicate a time point on any identified time scale. This arose from trying to say what it indicates. It is, as the draft shows, easier to say what time interval a date time refers to. Maybe we should describe it as a "time coordinate" rather than an "absolute" one ..? Or eliminate the part of the definition of 'absolute time coordinate' that talks about an indefinite time scale, so that it is defined as "time coordinate that refers to exactly one time interval" ..? > * Instead, I suggest that 'date time' should be defined as indicating > the time of day time point indicated by the time coordinate 'within' > the calendar day indicated by the date coordinate. With this solution > we need a 'time interval corresponds to time of day time point within > calendar day' verb concept which we could define as 'the time interval > that corresponds to the time of day time point starts some duration > after midnight on the calendar day and the duration is the index of > the time of day time point * the granularity of the time scale of the > time of day time point' I understand, but there are no time of day time points 'within' a calendar day. The time periods are within/during the day period. What we are really talking about is constructing a time point on an absolute scale by taking each calendar day and subdividing it into time points of some time-of-day granularity G. By doing this to every calendar day, you get a sequence of time points on an indefinite time scale with granularity G. I really don't want to go to define such indefinite time scales. You'd have to have 3 of them (1 for seconds, minutes, hours) and would have to figure out the origin number when the Convention du Metre was signed. And then we're back into what second of the day the Convention was signed -- which they did not specify as far as I can tell. Net: this really opens up a can of worms. In everyday language, we certainly talk about times in/of/on a day. For example, 3pm on Tuesday. Why should build our model to match that. I agree that we should be "taking each calendar day and subdividing it into time points of some time-of-day granularity G". But I don't see why we need to construct an indefinite time scale. > * Suggest having notes for 'time of day' and 'calendar date' that > identify the subtype time coordinates corresponding to each one. Good point. They can be forward pointers to the Gregorian and Standard time sections. Yes. -Ed > -------------------------------- > Mark H. Linehan > STSM, IBM Research > > > > From: Ed Barkmeyer > To: OMG DateTimeVoc FTF , > Date: 07/02/2012 07:55 PM > Subject: Issue 17425 and 17429 calendar date > ------------------------------------------------------------------------ > > > > I attach a draft resolution, reflecting discussions on June 25 and July 2. > > This is a rewrite of 12.3.4 in anticipation of its being moved bodily > into the new Calendars section. > > There are two comments in this writeup. One is a synchronization issue. > The other is an ugly solution to a problem in expressing the definition > of date time. > There is also a missing comment: It seems we need absolute time scales > of hours, minutes and seconds for date time coordinates to have a time > point to indicate. > > -Ed > > > -- > Edward J. Barkmeyer Email: edbark@nist.gov > National Institute of Standards & Technology > Manufacturing Systems Integration Division > 100 Bureau Drive, Stop 8263 Tel: +1 301-975-3528 > Gaithersburg, MD 20899-8263 Cel: +1 240-672-5800 > > > > [attachment "Issue 17429-definition of calendar date.docx" deleted by > Mark H Linehan/Watson/IBM] [attachment "Issue > 17425-calendar-date.docx" deleted by Mark H Linehan/Watson/IBM] -- Edward J. Barkmeyer Email: edbark@nist.gov National Institute of Standards & Technology Manufacturing Systems Integration Division 100 Bureau Drive, Stop 8263 Tel: +1 301-975-3528 Gaithersburg, MD 20899-8263 Cel: +1 240-672-5800 "The opinions expressed above do not reflect consensus of NIST, and have not been reviewed by any Government authority."