W3C home > Mailing lists > Public > semantic-web@w3.org > May 2007

Re: vCard/iCalendar RDF process document 2007-04-06

From: Garret Wilson <garret@globalmentor.com>
Date: Wed, 02 May 2007 13:06:26 -0300
Message-ID: <4638B702.9020306@globalmentor.com>
To: Bruce D'Arcus <bdarcus@gmail.com>
CC: Kjetil Kjernsmo <kjetil@kjernsmo.net>, semantic-web@w3.org

Bruce D'Arcus wrote:
>
>>
>> Again, sorting is not my issue here.
>
> But it IS an issue. Why do we bother with breaking apart names like 
> this except to be able to know how to order them in a list or table, 
> recompose them for different kinds of display, and so forth?

Your logic here seems to be, "Sorting is an issue. Breaking apart names 
may not help with sorting. Therefore we should not break apart names." 
My logic is: "Sorting is an issue. Breaking apart names may not help 
with sorting. We should still break apart names for other reasons." 
That's why I said that sorting is not the issue I'm raising here.

The separation of names itself is semantic content; I don't want to 
throw that away, whether or not it helps in sorting.

Imagine you have the following properties:

<vcard:name>van Buren bin Laden</vcard:name>
<vcard:nameSeparationIndex>9</vcard:nameSeparationIndex>

By saying that you don't want to break apart names (or at least allow 
it), it's like saying, "You must always discard the 
vcard:nameSeparationIndex property" (from the example above). For me, 
it's all about not losing semantics that may be useful---whether or not 
it helps with sorting.

>
>
>>> Yet vCard has no support for articular names. So isn't the question 
>>> in this case which representation is more wrong, rather than which 
>>> is right?
>>
>> vCard supports articular names just fine:
>>
>> N:bin Muhammad,bin 'Awad,bin Laden;Osama;;;
>
> I mean in the sense there is no such structure in the spec; "bin" is a 
> meaningless piece of text.

I don't follow you. "bin" is a meaningless piece of text, sure. so is 
"Laden" and "Osama", as far as the processor is concerned. My point is 
that the division is significant: there is a difference between "van" 
and "Buren"; and "van Buren".


>
> If the consensus is that we also ought to chop up those properties 
> according to the comma delimiter into lists, then so be it, but I 
> thought it might bear some in-depth discussion before going down that 
> road. Your note that the approach was "controversial" seemed to 
> recognize the same.

That was my nice way of saying you disagreed with me. ;)

Seriously, though, my whole problem is that I don't want to throw away 
data found in existing vCards (and potentially found in vCards ones they 
start being used more internationally). Ideologically, I would hate to 
think that the "recommended new semantic way of doing things" (RDF) is 
less expressive than the "old brute force proprietary way of doing 
things" (vCard).


>
> Relational databases also don't support order by default. And I've 
> never actually seen a GUI that could handle these cases well (if Osama 
> bin Laden filled out a contact record somewhere, I am betting he would 
> have a single family name field!), but ...

If nothing else, I think we had better wrap up the discussion quickly 
before the Department of Homeland Security shuts down this list. ;)

Garret
Received on Wednesday, 2 May 2007 16:07:05 UTC

This archive was generated by hypermail 2.4.0 : Tuesday, 5 July 2022 08:45:00 UTC