Print

Print


Ray (and all),

First of all: the revised specs look great to me. Thanks for all the great work.

Next, I apologize for being late to the call for use cases and for the call for feature #318 (before/after) feedback. 

Just to make sure I understand the move of putting the before/after syntax on "last call": is the rationale for that move that anything expressed with the '.af' (and friends) syntax could be expressed otherwise anyways?

Below, under "===use cases START===", I have tried to come up with some EDTF encodings for natural-language date intervals I regularly encounter in the Founding Fathers documentary editions. Please let me know if I'm misunderstanding anything.

In the below, the string in quotation marks immediately following the case number is natural-language date (range) information as frequently encountered in the documentary editions of the Founding Fathers Papers. Hopefully, the lines following each case identifier (EDTF, B/A last call, "Mathy") are self-explanatory.

===use cases START===

[Case 1]: "before March 11, 2011"
EDTF: unknown/2011-03-10 (note date adjustment)
B/A last call: .bf.2011-03-11
Mathy: ]unknown,2011-03-10] (left-open interval for "unknown"?)

[Case 2]: "after March 11, 2011"
EDTF: 2011-03-12/unknown (note date adjustment)
B/A last call: .af.2011-03-11
Mathy: [2011-03-12,unknown[ (right-open interval for "unknown"?)

[Case 3]: "not before March 11, 2011"
EDTF: 2011-03-11/unknown
Mathy: [2011-03-11,unknown[ (right-open interval for "unknown"?)

[Case 4]: "not after March 11, 2011"
EDTF: unknown/2011-03-11
Mathy: ]unknown,2011-03-11] (left-open interval for "unknown"?)

[Case 5]: "on or before March 11, 2011"
EDTF: unknown/2011-03-11
Mathy: ]unknown,2011-03-11] (left-open interval for "unknown"?)

[Case 6]: "on or after March 11, 2011"
EDTF: 2011-03-11/unknown
Mathy: [2011-03-11,unknown[ (right-open interval for "unknown"?)

===use cases END===

The following questions occur to me (as far as I can see, (B) and (C) may be irrelevant at this point, but (A) may be worth re-considering):

(A) In any of the examples above, would it be better to assume a value of "UNSPECIFIED" rather than "UNKNOWN" for reasons of semantics?

(B) Would we be better off by any chance if we adopted a more mathy notion/notation of left-open/right-open/left-closed/right-closed intervals?

(C) As far as implementations are concerned, should "unknown" ("unspecified"?) always be treated as an open interval?

Thanks much,

Markus

On Wed, 9 Mar 2011 15:35:45 -0500
 "Denenberg, Ray" <[log in to unmask]> wrote:
>I am adding the "Before/after indicator" feature, #318, to the list of features on LAST CALL.
>
>Please speak up for this feature if you need it.   
>
>(If we can drop it then we do not have to resolve the open issue of its syntax.)
>
>--Ray

--
Markus Flatscher, Editorial and Technical Specialist (Web Developer II)
ROTUNDA, The University of Virginia Press
PO Box 801079, Charlottesville, VA 22904-4318 USA
Courier: 310 Old Ivy Way, Suite 302, Charlottesville VA 22903
Email: [log in to unmask]