Re: thing, device, application

On 7/6/11 10:19 AM, Cyrus Daboo wrote:
> Hi Peter,
> 
> --On July 6, 2011 9:30:38 AM -0600 Peter Saint-Andre
> <stpeter <at> stpeter.im> wrote:
> 
>>> I'd be very cautious about defining something broader than what we need.
>>
>> Simon, I find the arguments you and Cyrus have provided persuasive. I'm
>> now more comfortable with KIND:application for the XMPP use case. (Yes,
>> we might want to define subtypes even for applications.) If people have
>> use cases for KIND:device or KIND:vehicle or whatever, then I think they
>> can define those separately.
>>
>> One possible concern is a proliferation of KIND values. But I don't
>> actively worry about that, and I'd rather label things accurately.
> 
> Hrmm, so I wonder if we start re-using LDAP objectClasses for KIND,
> might it not be better to tackle the last item on our current charter:
> "LDAP mapping to the new vCard format without loss of data.".
> 
> I am thinking that objectClass applicationProcess probably has a whole
> bunch of attributes specific to it in the LDAP world and we should map
> those over.
> 
> That also begs the question of whether KIND should have a direct
> relationship with LDAP objectClasses.
> 
> So maybe we need to think more closely about the whole LDAP<->vcard
> thing and decide how far down that road we want to go...

Right, but that's a large task and I'm not sure that there's energy for
it. The KIND:application definition would be inspired by LDAP but would
not require a complete mapping. Or at least that's how I see it with my
individual hat on...

Peter

--

-- 
Peter Saint-Andre
https://stpeter.im/