 |
 |
At 11:04 AM 9/29/97 -0400, Ray Denenberg wrote: >i find the argument un-persuasive that there should not be an element that >cannot be used unqualified. There already is one: coverage. That has no >semantics without qualification. As I just said in a message to Misha, the semantic is something like "spatial locations and temporal intervals characteristic of the resource" (e.g. 19'th century France). Similarly, the semantic of an unqualified Date is essentially a publication date. The argument on unqualified elements is that you should be prepared to deal with an unqualified instance of any element. Pretty much this means that if you are getting DC records from an outside source, you have decided in advance what unqualified elements you will retain for fielded searches(Title and Creator), what ones will all be tossed into a full-text index (subject, description, perhaps others) and what ones you will silently delete as being meaningless (your choice). We can certainly issue recommended practice guidelines, but you should have decided in advance how to handle records that don't have the level of detail you want. <rant> Not to pick on Ray, but I'm bewildered by this sudden gush of messages on changes to the core and pleas for more restrictive semantics. All 15 elements have semantics. The semantics are, unarguably, very loose on some elements. That does not mean they are non-existent, it means they are very loose on some elements. My contention is that for the simple cases, the rules are adequate. The goal, IMHO, is to let the simple stuff remain simple. If people want to do something complex, great. However, complexity comes at a cost, and that cost should not be amortized across unwilling users. </rant> Ron Daniel Jr. voice:+1 505 665 0597 Advanced Computing Lab fax:+1 505 665 4939 MS B287 email:[log in to unmask] Los Alamos National Lab http://www.acl.lanl.gov/~rdaniel Los Alamos, NM, USA, 87545