I like this solution, but, unfortunately the value of the normal field in
this case would be invalid against the EAD schema. I'm not sure if that's
not a problem with the EAD schema or not, however. If "PT0S" should a
valid as part of an ISO8601 date then the pattern of the normal attribute
might need to be changed.
BTW, one might also, I suppose, give an end date with the maximum
allowable year value to express an "open-ended" date. It is "2999" in the
EAD schema.
<unitdate normal="1970/2999">1970-</unitdate>
This might be easier for machines to process than nothing after a "/".
/Terry
On Thu, 6 Dec 2007,
Clay Redding wrote:
> Hi Jenn,
>
> We've been researching this issue at LC for MODS, PREMIS, etc. I have
> an inelegant solution to this that we didn't end up adopting for open
> ended ranges. It still complies with ISO8601, however, and it seems
> that a number of systems (Microsoft uses it in a number of apps) have
> used this format. If you use the string "PT0S" as the end date, that
> indicates an open or infinite duration. "P" is the 8601 designator to
> show an interval or duration, "T" means time, and "0S" means zero
> seconds. Some variations of this are PT0Y (year), PT0D (day), etc.
> They all mean the same thing when qualified as zero.
>
> TEMPER didn't help with this either. I think it also recommends just leaving the unknown end date blank. No spec that I've seen has a way of indicating "a future date that /will/ end at some point, but we don't know when".
>
> In your example, if you choose to use this, it would look like:
>
> <unitdate normal="1970/PT0S">1970-</unitdate>
>
> Clay
>
> ~~~~~~~~~~~~~~~~~~
> Clay Redding
> Digital Project Coordinator
> Network Development & MARC Standards Office
> Library of Congress
> LA308, Mail Stop 4402
> 101 Independence Ave. SE
> Washington, DC 20540
> [log in to unmask]
> 202-707-7196
> ~~~~~~~~~~~~~~~~~~
>>>> "Riley, Jenn" <[log in to unmask]> 12/06/07 9:04 AM >>>
> Any other suggestions for how to encode an open-ended date range? Even if DACS says not to do this, not all EAD implementers will be using DACS. I know it's convenient to make @normal conform to the ISO8601 data type, but we're really excluding a significant use case here...
>
> Thanks all!
>
> Jenn
>
>> -----Original Message-----
>> From: Encoded Archival Description List [mailto:[log in to unmask]] On Behalf
>> Of Kellams, Dina M
>> Sent: Tuesday, December 04, 2007 7:33 AM
>> To: [log in to unmask]
>> Subject: Re: open ended-date range
>>
>> Because you'll probably hear this from others, according to DACS, we
>> are no longer supposed to leave anything with open date ranges, but to
>> instead give it a closing date and update it as necessary. I know, I
>> know, I still have some finding aids out there with an open date range
>> (BFC & Trustees) but I just haven't found it necessary to go back and
>> clean those up yet.
>>
>> (DACS rule 2.4.8: When further accruals are expected, record the
>> inclusive dates pertaining to the holdings currently in the custody of
>> the repository. Record information about expected accruals in the
>> Accruals Element (5.4). When the accruals are received, revise the date
>> as necessary.)
>>
>> -----Original Message-----
>> From: Encoded Archival Description List [mailto:[log in to unmask]] On Behalf
>> Of Riley, Jenn
>> Sent: Monday, December 03, 2007 7:36 PM
>> To: [log in to unmask]
>> Subject: open ended-date range
>>
>> Hello all,
>>
>> In working with finding aids encoded according to the EAD 2002 W3C
>> Schema instead of the DTD, we've run into an issue with the
>> unitdate/@normal attribute. The data type for the unitdate/@normal
>> attribute doesn't see<unitdate normal="1970/">1970-</unitdate>m to support open-ended date ranges (because it's
>> based on ISO8601, which doesn't support them either, is that right?).
>> So what's the best encoding to use for a date range that, say, starts
>> in 1970 and continues to the present day? These aren't valid:
>>
>> <unitdate normal="1970/">1970-</unitdate>
>> <unitdate normal="1970/9999">1970-</unitdate> (that's a hack, but I've
>> seen it in use before)
>>
>> I can think of these other options:
>>
>> <unitdate normal="1970">1970-</unitdate> (treat it as a single date in
>> the normal attribute)
>> <unitdate>1970-</unitdate> (leave the normal attribute off entirely)
>>
>> I don't like either of those options, though. What would you do?
>>
>> Jenn
>>
>> ========================
>> Jenn Riley
>> Metadata Librarian
>> Digital Library Program
>> Indiana University - Bloomington
>> Wells Library W501
>> (812) 856-5759
>> www.dlib.indiana.edu
>>
>> Inquiring Librarian blog: www.inquiringlibrarian.blogspot.com
>
Terry Catapano
Special Collections Analyst/Librarian
Columbia University Libraries Digital Program
212-854-9942
[log in to unmask]
|