- From: Dave Pawson <dave.pawson@gmail.com>
- Date: Tue, 4 Jun 2013 07:13:36 +0100
- To: public-vocabs@w3.org
Hi Chaals. On 3 June 2013 21:48, Charles McCathie Nevile <chaals@yandex-team.ru> wrote: >> We should have the description of http://schema.org/LocalBusiness be more >> filtered and fine tuned to what it really is representing. The current >> description does not capture the reality as well as it should. > > > Yes... But... (see below) > >> If we say that a LocalBusiness must have a few minimums to be typed as a >> LocalBusiness, such as mailing address at a minimum, than put that into >> the description. Or a telephone number additionally, if that is also a >> minimum requirement. > > > No, a phone number is not a requirement. In principle a mailing address > might be, but that starts to pin down the nature of the business. > > Although one of the businesses I rely on is a truck that appears in my > village on mondays and sells fruit, and a small car that appears six days a > week and sells bread. > >> I would rather see better descriptions in Schema.org that help frame up >> the realities of the Schema Types that we are trying to preserve based on >> those minimum Schema properties that signify that particular Type. > We're at something of a crossroads. We can either let things grow > organically, or we can try to shake up the vocabulary to get a bit more > structure. > > The first approach leads to such unexpected results as saying "the > courthouse is a local business", which people might think is unintended > humour - or worse, might actively use as intentional humour. > > The second requires us to do a lot of hard thinking about things we > describe, and learn more about what charateristics they really have. If the > goal is to create a lot of work in the short term, this is definitely the > answer - but it is not as clear which approach (or blend of the two) is > likely to be more successful. In part that hinges on our strategy for > versioning. > > Questions like "why can't I describe the different kinds of phone number"?, > or "should telephone numbers actually conform to the tel: URL scheme so they > can be used?" motivate thinking about this kind of meta-problem. Chaals points out some of the difficulties in adding semantics. Interpretation of nouns will always be local. If specific meanings are attached to classes the number will rise rapidly as users reject that definition and look for others. Contra. Without any semantic, we all guess / make our own assumptions and have a nasty surprise when we find out what others think it means. I'm supportive of very general semantics, would prefer users develop and agree over time anything further. Max was looking for a refinement (phone number to 'home phone number') Is there any convention for doing this in schema.org? regards -- Dave Pawson XSLT XSL-FO FAQ. Docbook FAQ. http://www.dpawson.co.uk
Received on Tuesday, 4 June 2013 06:14:07 UTC