Re: Need explication about rfc2426

On 2002-08-09 12:01:28 -0700 Paul Hoffman / IMC <[email protected]> wrote:

> The first format you should want to support is vCard 2.1, since that is the one that is most interoperable with applications out in the world.
Yes... that's my next step :-)

> vCard 3.0 is a better format, but is not as widely used, unfortunately.
Yes, i think it's more explicit and more complete.

> A vCard can also represent a group of people, such as a workgroup,
A workgroup... difficult concept to represent easily with a gui :-/

> or a person with more than one name.
Why use many N fields to do that ? The BNF grammar of the rfc2426 describes the N field like this:
n-value = 0*4(text-value *("," text-value) ";") text-value *("," text-value)
So, each part of n-value is a text-value-list ; if a person have more thant one name, the vcard writer should write:
family1,family2;given1;middle1,middle2;etc
I think this description should use the text-value-list instead of "text-value *("," text-value)".

>> More important, is there a standard way (or a very used way) to
>> generate this UID ? Should i use the Message-ID way (rfc2822) ?
> 
> I really don't know on this one. The idea of the UID was to allow a globally-unique ID.
So... why reinvent the wheel ? :-) I go to use this experienced algorithm :-)

> It is almost never used, which probably is why no one noticed the bug in the 3.0 spec.

By the way, according to the paragraph 3.2.1 (ADR Type Definition), page 12:
When a component value is missing, the associated component separator MUST still be specified.
So, i think the corresponding grammar part should be (according to the ABNF rfc):
adr-value = 6(text-value ";") text-value
and not:
adr-value = 0*6(text-value ";") text-value.

Oh... i've a new question :-)
Why do not create a 'vcard-set' like with the xml format ? That would be more easy to regroup many vcards.

PS: i've already report a bug in this rfc to the rfc-editor : in the grammar, it appears two times a 'text-list' instead of 'text-value-list'.

-- 
Vincent