Andrew:
I've taken the liberty of copying your comments, and my replies, to the
DC-Userguides list, since you said you'd been unable to do so. I also
copied meta2, since discussion on this element is now going on in both
places. My comments are interwoven below:
>On Mon, 1 Jun 1998, Diane I. Hillmann wrote:
>
>>
>> 6. Included coverage wording I suggested in my message of 5/23, which no
>> one seemed to have much objection to.
>>
>
>Well, er, I have ...
>
>
>http://128.253.70.110/DC5/UserGuide5.html
>includes as examples
> <META NAME="DC.Coverage" CONTENT="17th century">
> <META NAME="DC.Coverage" CONTENT="Upstate New York">
>
>which is exactly the kind of vagueness that I'd rather not encourage by
>writing it down anywhere official.
>I'd been waiting for some further response from Mary or other members of
>the coverage WG, but it's possible some of my mail fell into a hole; I
>know I had picked up some from home with IMAP and got rejected by the
>userguide listserv for not being subscribed, but thought I'd forwarded it
>through my work machine....
I agree that simple naked coverage is not optimal for many DC users, but
the fact is that it will be used by some, whether or not we "write it
down." We have no way to "enforce" qualification, no matter that we might
think it a good idea for this element in particular. We did say: "For more
complex applications, consideration should be given to additional
qualification." I don't think we can go further than that, in a simple DC
document, without saying "DON'T USE THIS ELEMENT" in neon lights.
>As far as I recollect from the last year or so, the unqualified Coverage
>element was never condoned or defined. The qualified elements may be
>crosswalked into geographic information systems; the unqualified
>ones are back to the "throw all the elements into the keyword pool
>and see what the search engines make of it" idea.
And don't you think that this is the way some people will be using most of
the elements? I think we should recognize this, get comfortable with it,
and move on.
>I have pruned my earlier suggestions somewhat at
>http://andrew.triumf.ca/coverage/UserGuide3d.html,
>though I'd like to see some wording encouraging people to at least
>look at the full specification, as I believe that we have good, workable
>definitions for most of the qualified elements and that promoting
>a "dumbed-down" usage will lose data and cripple applications.
I don't think we're promoting "dumbed-down usage." We're recognizing that
people will use it, because it's there. Whether they use it appropriately
or with care and forethought, is beyond our ability to control--but then,
this is true of the format as a whole. When we finally do have guidelines
for qualified DC that we can point to--great, let's do that. For now, I
suggest we not spend a lot of time gnashing our teeth over this.
>(I have recently been concerned in a project which, though it currently
>uses a custom record format not DC, nevertheless uses
>placenames and simple position data to generate maps in a manner
>impossible using only ill-defined names,
>viz. http://www.slac.stanford.edu/xorg/iepm/pinger/map.html
>http://sitka.triumf.ca/net/pingER-sites.html)
>
>
>regards, Andrew
>
>Deniable unless digitally signed
>Andrew Daviel, TRIUMF, Canada
>Tel. +1 (604) 222-7376
>http://andrew.triumf.ca/andrew
/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
Diane I. Hillmann
Head, Technical Services Support Unit
Cornell University Library E-mail: [log in to unmask]
107B Olin Library Voice: (607) 254-5290
Ithaca, New York 14853 Fax: (607) 255-6110
WebGoddess: http://www.library.cornell.edu/tsmanual
/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
|