On Monday 06 October 2008 12:15:56 Pedro Melo wrote:
> On Oct 6, 2008, at 7:38 PM, Justin Karneges wrote:
> > On Monday 06 October 2008 10:45:06 Remko Tronçon wrote:
> >> So, I agree with Pedro that resources should be opaque, and that what
> >> we currently abuse resource names for (mainly historical because not
> >> many clients did much disco'ing before caps).
> >
> > Do we abuse resource names that way?  When I direct stanzas at a
> > particular
> > resource, it's a human decision.  It's not like I wrote code that
> > scans for
> > resource strings that contain "laptop" in them.  Who has even done
> > that? :)
>
> The Psi GUI? :)

No.  Psi lists the resource names but my point is that it never parses them 
for meaning.

> If I right click a contact, all the "ACTION to" will open a sub-menu
> with the available resource.
>
> The concept is correct: you should be able to select the resource
> when you want to execute commands, we just should not identify the
> connections using the resource as is the current practice.

And instead we should identify them with the disco identity name, which is 
also a free-form, untranslatable, opaque string?  Fair enough, that's the 
same thing.

> cool, you should name the connections. I think we all agree on that.
> I'm just suggesting that you use the name attribute of the disco
> <identity> to do it.
>
> That's what he is there for.

I'm reminded of the DNS-SD specification.  I suggest reading section 4.4:
  http://files.dns-sd.org/draft-cheshire-dnsext-dns-sd.txt

The authors (likely Cheshire here) argues that there is not much point in 
using an ugly hex value for the real identifier, and a discoverable friendly 
name, when you could just use the friendly name as the real identifier in the 
first place.

-Justin

Reply via email to