- From: Frank Manola <fmanola@mitre.org>
- Date: Mon, 09 Dec 2002 18:59:20 -0500
- To: pat hayes <phayes@ai.uwf.edu>
- CC: w3c-rdfcore-wg@w3.org
pat hayes wrote: > >> pat hayes wrote: >> >>> >>> I forgot to mention. You used the term 'primary value' to refer to >>> the value of rdf:value (if you see what I mean), and I avoided it >>> deliberately because in this context particularly it might suggest to >>> someone with a DB background that rdf:value was being used to >>> indicate a primary field in a DBS, and nothing could be wronger. I >>> havnt checked whether or not you follow this usage elsewhere in the >>> primer. >> >> >> Pat-- >> >> (minor point) >> I have a database background and I wasn't confused (nothing like a large >> sample size to prove a point!). > > > You werent confused because you WROTE it. Everything I write seems > transparently obvious to me at the time I write it, but I can't > understand it the next day. At least I don't *think* I was confused *when* I wrote it (I'm not sure about now though). What you say is why I wait until the next day to read the stuff again (it usually changes then). > >> I think the database term you have in mind is "primary key" rather than >> "primary field" > > > Yes, whoops, sorry. (Im NOT a DB person....) > >> , and we've never suggested anything like "key-ness" of >> values in any of this discussion (A "key" is a unique identifier; a >> "primary key" is selected from among the possible "candidate keys" to >> identify rows in a relational table; a row in one table might identify a >> specific row in another table using a "foreign key". Note that all >> these involve using various adjectives on "key". I'm not familiar with >> "primary field"). However, if people think it'll be a problem we can >> change it to "main" (as you did in your rewrite). > > > I know we havnt said anything explicitly, but I bet that any DB guy > would feel right at home reading your explanation of how to handle > more-than-binary relations, thinking happily about 'complex values' as > DB rows, and to throw 'primary' at him at that point with no warning is > a low blow, seems to me. Particularly in that example, where one might, > just, at a stretch, think that the weight of a tent could be the primary > key. > Actually, I can't imagine it normally occurring to a DB guy to think of weight as a primary key. However, it gets worse. After composing my reply, I looked at the Concepts document, and section 3.5 has an example (the "John's address" example from the Primer) that actually uses the term "primary key" (it uses "staffid" as the primary key, which is reasonable). So whatever text we use, it'll need to avoid "primary". --Frank -- Frank Manola The MITRE Corporation 202 Burlington Road, MS A345 Bedford, MA 01730-1420 mailto:fmanola@mitre.org voice: 781-271-8147 FAX: 781-271-875
Received on Monday, 9 December 2002 18:41:51 UTC