W3C home > Mailing lists > Public > public-swd-wg@w3.org > July 2007

Re: Why I don't like 'instanceof' (was Re: [RDFa] ISSUE-3: syntactic sugar for rdf:type)

From: Niklas Lindström <lindstream@gmail.com>
Date: Fri, 20 Jul 2007 13:37:19 +0200
Message-ID: <cf8107640707200437j695df2e2lf0d953314b00a59@mail.gmail.com>
To: mark.birbeck@x-port.net
Cc: "Ivan Herman" <ivan@w3.org>, "Hausenblas, Michael" <michael.hausenblas@joanneum.at>, "Steven Pemberton" <steven.pemberton@cwi.nl>, "Ben Adida" <ben@adida.net>, RDFa <public-rdf-in-xhtml-tf@w3.org>, "SWD WG" <public-swd-wg@w3.org>

I agree with Mark's defence of @instanceof. It may not be ideal
though, so I'll do some more brainstorming.

Of the one-word suggestions, I found @kind most likable. But I fear it
doesn't mean exactly the same thing. Perhaps nit-picking, it seems to
have a use as "the type of a type" in some settings, especially
<http://en.wikipedia.org/wiki/Type_system#Types_of_types>.

@category is probably a proper synonym
(<http://en.wikipedia.org/wiki/Type_%28metaphysics%29>,
<http://dictionary.reference.com/search?q=type>). Still, for me it
speaks more of a loose relation ("X belongs in category Y") than of a
stricter classification/type membership relation. Perhaps it's the
layman use that differs, and not the academic definition. I doubt
it'll "read right"..

I also agree with Mark that the "two words" problem may need more
argumentation. I am however somewhat sympathetic to it. But I cannot
think of any synonym to type (apart from class) which doesn't bring
along a possible difference in interpretation.

Following, I would also prefer something with the word "type" in it,
as Michael reasoned (with @typeOf, and Keith with @rdftype). I can
think of one, using the syntactic sugar of Notation 3 where you can
add "has" before any predicate. This gives us @hastype. Not sure if it
is less technical and more HTMLish, but possibly (and it reasonably
can't be expected for other use in future HTML). I can also think of a
one-word with type: @typed (to be read as "subject is typed as").

Finally, a pet "non"-suggestion of mine I really don't expect
following of, but it so craves to get out: @a. Yup, exactly like in
Notation 3, with all the virtues and flaws. :)

This makes my set of suggestions so far (including @a for
completeness), illustrated:

    <div about="#me" instanceof="foaf:Person">..</div>

    <div about="#me" hastype="foaf:Person">..</div>

    <div about="#me" typed="foaf:Person">..</div>

    <div about="#me" a="foaf:Person">..</div>

Best regards,
Niklas



On 7/20/07, Mark Birbeck <mark.birbeck@x-port.net> wrote:
>
> Hello all,
>
> Is 'instanceof' really so bad? It's been said that it is second-best
> to the more preferable (but taken) @role, @type and @class attributes,
> but I disagree. As a well-known British advert goes: "it does what it
> says on the tin", and I think that mark-up that indicates that we have
> an 'instance of' a foaf:Person (for example), is actually much closer
> to what we want than 'type', 'role' or 'class'.
>
> I also don't believe it is 'RDF-speak', any more than 'class', or
> 'type', or 'resource'. Of course, for someone versed in RDF these
> names will ring bellsand be very clear, but even for someone with no
> RDF background, I don't think they are particularly confusing. The
> notion of 'instances' of something will be pretty clear to most
> authors. And whether we like it or not, I think we do have to accept
> that we can't hide some of the basic concepts.
>
> The other objection I've seen is from Steven, that the name comprises
> 'two words', but with respect that does seem to be a personal
> preference, rather than based on some deep problem. It would be great
> if it was backed by an argument, otherwise we can't actually debate
> it.
>
> Regards,
>
> Mark
>
> On 20/07/07, Ivan Herman <ivan@w3.org> wrote:
> >
> >
> > Hausenblas, Michael wrote:
> > >> 'type' is an existing attribute in HTML
> > >
> > > ... that's why my mother always says: Think before you speak/write ;)
> > >
> > > But, still, IMHO the *new* attribute should be something
> > > containing 'type'. Now risking Steven will hate me, I propose:
> > >
> > > 'typeOf'
> > >
> > > Hope I did not open another Pandora's box ;)
> >
> > You did:-)
> >
> > Ivan
> >
> > >
> > > Cheers,
> > >       Michael
> > >
> > > ----------------------------------------------------------
> > >  Michael Hausenblas, MSc.
> > >  Institute of Information Systems & Information Management
> > >  JOANNEUM RESEARCH Forschungsgesellschaft mbH
> > >
> > >  http://www.joanneum.at/iis/
> > > ----------------------------------------------------------
> > >
> > >
> > >> -----Original Message-----
> > >> From: Ivan Herman [mailto:ivan@w3.org]
> > >> Sent: Friday, July 20, 2007 11:17 AM
> > >> To: Hausenblas, Michael
> > >> Cc: Steven Pemberton; Ben Adida; RDFa; SWD WG
> > >> Subject: Re: Why I don't like 'instanceof' (was Re: [RDFa]
> > >> ISSUE-3: syntactic sugar for rdf:type)
> > >>
> > >> 'type' is an existing attribute in HTML
> > >>
> > >> Ivan
> > >>
> > >> Hausenblas, Michael wrote:
> > >>> Again: Why do we refuse naming it 'type'?
> > >>>
> > >>> Because it is to RDFish? (BTW, we're doing *R*D*F*a)
> > >>> Or are there any (X)HTML (2) issues, I might have missed?
> > >>>
> > >>> Cheers,
> > >>>     Michael
> > >>>
> > >>> ----------------------------------------------------------
> > >>>  Michael Hausenblas, MSc.
> > >>>  Institute of Information Systems & Information Management
> > >>>  JOANNEUM RESEARCH Forschungsgesellschaft mbH
> > >>>
> > >>>  http://www.joanneum.at/iis/
> > >>> ----------------------------------------------------------
> > >>>
> > >>>
> > >>>> -----Original Message-----
> > >>>> From: public-swd-wg-request@w3.org
> > >>>> [mailto:public-swd-wg-request@w3.org] On Behalf Of Ivan Herman
> > >>>> Sent: Friday, July 20, 2007 10:57 AM
> > >>>> To: Steven Pemberton
> > >>>> Cc: Ben Adida; RDFa; SWD WG
> > >>>> Subject: Re: Why I don't like 'instanceof' (was Re: [RDFa]
> > >>>> ISSUE-3: syntactic sugar for rdf:type)
> > >>>>
> > >>>> If so... 'category' maybe the closest to what we mean...
> > >>>>
> > >>>> Ivan
> > >>>>
> > >>>> Steven Pemberton wrote:
> > >>>>> I think there are only 3 reasons why I think 'instanceof' is
> > >>>> a bad choice:
> > >>>>> 1. Multiword, which I already spoke of.
> > >>>>> 2. instance has another meaning in some existing and future XHTML
> > >>>>> documents.
> > >>>>> 3. It comes over as rdf-speak. Up to now we have done our
> > >>>> best to avoid
> > >>>>> exposing RDF terminology to the XHTML author; no subject,
> > >> predicate,
> > >>>>> object and so on, just existing HTML concepts where possible.
> > >>>>> Unfortuantely, most of the synonyms have already been taken (class,
> > >>>>> type, role), but I still think we should try and find
> > >> something that
> > >>>>> reads better than 'instanceof' or 'isa'.
> > >>>>>
> > >>>>> /me runs a thesaurus
> > >>>>>
> > >>>>> sort
> > >>>>> kind
> > >>>>> category
> > >>>>> realm
> > >>>>>
> > >>>>> depict
> > >>>>> portray
> > >>>>> represent
> > >>>>> embody
> > >>>>>
> > >>>>> like
> > >>>>>
> > >>>>> Steven
> > >>>>>
> > >>>>> On Thu, 12 Jul 2007 20:25:48 +0200, Ben Adida
> > >> <ben@adida.net> wrote:
> > >>>>>> Hi all,
> > >>>>>>
> > >>>>>> In today's telecon, we proposed and resolved to use a *new*
> > >>>> attribute,
> > >>>>>> rather than @class or @role, for the rdf:type syntactic
> > >> sugar. Thus,
> > >>>>>> @class and @role do not currently result in any triples
> > >>>> being generated,
> > >>>>>> although one may consider that they will in a future version.
> > >>>>>>
> > >>>>>> The question, then, is which attribute to use. Steven expressed
> > >>>>>> reservations about two-word attributes like "isa" or
> > >>>> "instanceof", and
> > >>>>>> instead proposed: denotes, depicts, represents, category,
> > >> ilk, kind.
> > >>>>>> Other thoughts?
> > >>>>>>
> > >>>>>> I'm partial to "instanceof" and "kind", and I have no additional
> > >>>>>> suggestions.
> > >>>>>>
> > >>>>>> -Ben
> > >>>>>>
> > >>>>>
> > >>>> --
> > >>>>
> > >>>> Ivan Herman, W3C Semantic Web Activity Lead
> > >>>> Home: http://www.w3.org/People/Ivan/
> > >>>> PGP Key: http://www.ivan-herman.net/pgpkey.html
> > >>>> FOAF: http://www.ivan-herman.net/foaf.rdf
> > >>>>
> > >> --
> > >>
> > >> Ivan Herman, W3C Semantic Web Activity Lead
> > >> Home: http://www.w3.org/People/Ivan/
> > >> PGP Key: http://www.ivan-herman.net/pgpkey.html
> > >> FOAF: http://www.ivan-herman.net/foaf.rdf
> > >>
> >
> > --
> >
> > Ivan Herman, W3C Semantic Web Activity Lead
> > Home: http://www.w3.org/People/Ivan/
> > PGP Key: http://www.ivan-herman.net/pgpkey.html
> > FOAF: http://www.ivan-herman.net/foaf.rdf
> >
> >
>
>
> --
>   Mark Birbeck, formsPlayer
>
>   mark.birbeck@x-port.net | +44 (0) 20 7689 9232
>   http://www.formsPlayer.com | http://internet-apps.blogspot.com
>
>   standards. innovation.
>
>
Received on Friday, 20 July 2007 11:37:41 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:17:29 GMT