[foaf-dev] embedded file-based metadata - for .torrent

Peter Williams pwilliams at rapattoni.com
Sat Dec 25 04:13:07 CET 2010

Perhaps I should ask it other way arounds.

It's easy to go to google and search for file types (such as those implied by .torrent). Similarly, I can find RSS feeds whose elements wrap .torrent content types. I can get Outlook to recognize the opening of a torrent ref in an RSS feed, and it fires up the torrent client. It will then locally store the released content as a local resource, and then open the resource referring to it by its local reference.

I can also read all about a microblog feed about a particular torrent @ http://techcrunch.com/2010/08/06/twitter-now-even-more-torrent-friendly/ noting its main claim:

" From BitTorrent VP Simon Morris:

	"The point of Torrent Tweet is to adapt the powerful referencing system built into BitTorrent to the incredible social interaction engine that Twitter has built such that people can have conversations about things they are downloading, and they can be sure that they are talking about the same thing." "

I suppose my original question was too technical (how does one describe .torrent urlrefs in semweb-land, and does it work like .sig urlrefs in wot-land?)

Perhaps a better question is:

What have folks done with foaf cards when working with: a twitter-like capability aiming to confirm an claimed identity relation in a graph (e.g. bittorrent's tweeint integration), web2.0-era RSS feed of refs to .tweet resources/things, or when simply creating .torrent streams that contain one or more foaf PPDs persisted as file streams?

I dont see why its inappropriate (nowadays) to let a webid ( in its client cert) be ref to a container (of a foaf card) - where the container is essentially playing the role of a transfer-syntax and its transfer-encoding. At this point we have a foaf protocol (and the topic should probably move mailing list...). But, in short, once the PPDs in the .tweet stream are accesses, the individuals PPDs may well have either a local http uri (to the unwrapped resource) or more likely a local file: ref.

-----Original Message-----
From: Melvin Carvalho [mailto:melvincarvalho at gmail.com] 
Sent: Friday, December 24, 2010 11:35 AM
To: Peter Williams
Cc: foaf-dev at lists.foaf-project.org
Subject: Re: [foaf-dev] embedded file-based metadata - for .torrent

On 24 December 2010 20:06, Peter Williams <pwilliams at rapattoni.com> wrote:
> Using the definitions in the wot ontology, assume that some instance 
> of a metadata document makes statements that are (to humans) 
> references to files (.sig etc) to be pulled using HTTP(s). Since those 
> refs are related to descriptive statements that type the urlrefs, a 
> reasoning engine might know that .sig is a PGP signature blob (without 
> having formally defining the value syntax or internal-coding of such a 
> blob). Presumably, a semweb processor is supposed to fire off a 
> "plugin" that uses non-semweb metadata to fully interpret the blob; 
> and ultimately present the information in a browser-centric manner.

Yes, though WOT and FOAF will hopefully come together a bit more next year.  We just need to nail the use cases, terms and definitions.
Some notes are in progress here:


> In the case of FOAF-related ontologies (such as the proposed ontology 
> for RSA  keys), is there any experience (a la wot) with pointing to a 
> .torrent file (vs a .sig file), qualifying its urlref as torrent 
> stream (rather than a PGP .sig  blob).. and firing off a 
> file-type-centric plugin in the browser that renders the torrent 
> stream (rather than firing off PGP sig verification of a transfer-syntax/encoding of a RDF stream)?

Only one I know of the music ontology torrent field


> Ive been playing with outlook, embedded IE browser, email-pushes of 
> openid-signed comments on wordpress blog sites, torrents, and 
> twitter-microblogs about individual torrents (and torrent metadata). 
> Seems an ideal mechanism for mass distribution of self-signed root 
> certs for individuals..where (when combined with DNSsec) the seeding 
> and peering process addresses spoofing threats attaching the authority 
> of the binding to a domain-name.

Sounds great.

> If nothing else its complements the cached distribution model of the 
> web, leveraging the survivability nature of packet switching (as 
> intended by the early internet designers).

The bottom up approach of the semantic web means that if something isnt already created, you can make it yourself.  What you will tend to find is that if someone is prepared to well maintain a vocabulary, then its tends to get used more and more, and the necessary code will follow.

> _______________________________________________
> foaf-dev mailing list
> foaf-dev at lists.foaf-project.org
> http://lists.foaf-project.org/mailman/listinfo/foaf-dev

More information about the foaf-dev mailing list