[osis-core] div type="date"

[Patrick Durusau]

Chris,

Chris Little wrote:

>Good morning everyone,
>
>On day one of the meetings, Steve asked me to look into what would be 
>needed to encode a div whose type is a essentially temporal in nature (for 
>daily devotionals, lectionaries, timelines, etc.).  I sketched out some 
>samples of what would reasonably be needed, and essentially came up with 
>something like ISO date format crammed into an osisID.  Steve seemed to 
>think it looked okay, so I would propose it as a best practice.
>  
>
Looks good to me. Actually, the use of dates for the osisID actually 
corresponds to the implied reference system for such works.

>I don't know what the best type would be for these elements, or whether
>they should necessarily even be the same type.  Perhaps we should have
>divs with type "dailyDevotional", "lectionaryEntry", "timelinePoint", etc.
>(just examples--I'm really not married to these names).  The essential
>issue is that these divs encode things coded to time/date, so the osisIDs
>should have a consistent format.
>  
>
Don't forget that within a particular div, such as 01.01.AM, you might 
have separate divs for meditations, prayers, etc., so you might have a 
type for the overall div and then other types for the portions under 
that div.

>A simple 365-day, single entry per day devotional would have osisIDs for 
>each day's div of a format MM.DD.  Spurgeon's Mornings & Evenings would 
>have MM.DD.[AP]M to capture morning vs. evening readings.  (AM/PM isn't 
>part of ISO or anything that I know of, but it's a natural extension in 
>this situation.)
>  
>
AM/PM works across Christian traditions, might want to have alternatives 
available for before or after sunrise/sunset for Jewish traditions.

>If you want to encode something based on month/day/hour/minute/seconds, it
>would be MM.DD.HH.mm.SS.  Years are 4 digits wide (including zero 
>extension where necessary for years < 1000 CE or > 1000 BCE.  If you need 
>to encode BCE dates, you should use BCE.YYYY.MM.DD for BCE and 
>CE.YYYY.MM.DD for CE dates.
>  
>
Don't know that we need BCE dates. For recording time/calendar of say 
medieval prayer book, they probably don't use AM/PM time, we have modern 
equivalents but that is not what you want to encode as part of the 
original text.

>I don't think there's any good reason to constrain people to a particular 
>calendar or even to request them to state the calendar they use.
>  
>
Could simply say this is how to do date/time for texts that are bound to 
a particular time. Best Practice. If you do something else, well you 
better say what it is in order for other people to use it. (Another best 
practice.)

>An important aimed-for feature is that, like ISO's format, the osisIDs,
>when sorted alphabetically, will also be sorted chronologically.  (So it
>is important to specify CE for CE values if you are also using BCE values,
>rather than assuming a default.)  BCE sorts before CE.  AM sorts before
>PM.  And the digits almost always sort in order.  (The exception,
>unfortunately, is that BCE years count down.  I don't see a solution to
>that.)
>  
>
See my note above on the need for BCE.

>A sample daily devotional in valid OSIS (Spurgeon's Mornings & Evenings) 
>is posted at the bottom the documents page at 
>http://www.bibletechnologieswg.org/osis/docs/ .
>  
>
Looks good to me! Would be interesting to take a slight more complex 
daily devotional, the Catholic prayer book for example, and see how this 
works with it. Oh, I forgot, it has poetry in it! ;-)

Hope you are at the start of a great day!

Patrick

-- 
Patrick Durusau
Director of Research and Development
Society of Biblical Literature
Patrick.Durusau@sbl-site.org
Co-Editor, ISO 13250, Topic Maps -- Reference Model