Re: Human-Opaque URIs

On Mon, 7 Jun 2004, Phil Dawes wrote:

>Do you use any opaque URIs for properties and classes?

Rarely, but I feel I should becaus it woud encourage a little basic
discipline to always provide labels and descriptions (which can be clearer
and more useful for people, as well as being easy to provide in multiple
language)

>The reason I ask is because I'm currently re-writing veudas (web rdf
>editor) to generate unique uris using a hashing scheme - this is
>useful because it allieviates the user from having to think about
>RDF/URIs when creating new resource data (which is a good thing
>because I want it to be used by non-rdf-savvy users).

I think this would be useful. It took me a while to get used to my HTML
editing tools making up ID's, but now I am perfectly happy about it.

>Currently I'm steering away from this mechanism for creating
>properties, since it makes rdf queries completely unreadable and
>hand-writing rdql virtually impossible.

Does it? Presumably the things are still of reasonable length, so you can
copy/pate them if necessary. It might be useful to use something that looks a
bit like a word - say the first 4 letters of the label plus a 2-digit number?

>But the advantages are so compelling to me that I'm wondering if
>opaque uris are the way forward for classes/properties too. Maybe
>developers could use a gui tool to generate their queries (including
>comments) before cut-n-pasting into code?

Yes, I don't see any reason why not. RFD Author already allows you to do just
that...

>Anyway - I'd be interested to hear of any experiences/thoughts you
>might have in this area.

The most relevant experience I have (I think) is using a WYSIWYG editor for
HTML - Amaya. Now that I trust it to do things right, I have got used to
letting it decide what to use as IDs, and I am happy about it. They are not
very meaningful to me, but they seem to work fine for the Web :-)

cheers

Chaals

Received on Tuesday, 8 June 2004 03:36:56 UTC