Notes interleaved.
Thanks!
bill
--
On 9/18/14 1:59 PM, Gemmill, Craig wrote:
Ok
– that is what my original example had – I took out the dual
specification based on my (apparently incorrect)
interpretation of the call last week. I will replace the
example with both specified, to reinforce the notion that
one is required, both are allowed.
Also-
for the time of retrieval attribute name:
retrievalDate
retrievalTime
retrievedOn
retrievedAt
retrieved
And
should we stick with requiring abstime? The exact time
seems unlikely to be important, but abstime seems a more
common data primitive than date, since date is a 1.1
construct in OBIX.
abstime is an ISO 8501 "zulu" time, as I recall. That's fine - and
unambiguous. If you just want date, fine, but what about the date
the spec changed?
So I recommend dateTime or equivalent, which seems to be abstime.
From:
Considine, Toby [mailto:Toby.Considine@unc.edu]
Sent: Thursday, September 18, 2014 1:50 PM
To: Gemmill, Craig; William Cox; Obix TC List
Subject: RE: [obix] My pre-meeting comments, with
some post-meeting notes
My
understanding, as a conformance issue, is that it is
mandatory to have at least one (formal rev or retrieval
time) but optional to have both.
tc
“The
single biggest problem in communication is the illusion
that it has taken place.”
–
George Bernard Shaw.
Bill-
On
your comment about line 769, it looks as though you are
suggesting that the retrieval time be included EVEN IF the
version information is also available. This is different
than my understanding of our intent, which was retrievalDate
exists ONLY if version info is NOT available.
Just
throwing this out for others to weigh in.
Craig
Attached to this message as .rtf.
Thanks!
bill
|