hash uris (was Re: excellent compilation of quake3 stunts

from "Mark Baker" <[email protected]>, Tue, 12 Sep 2006 10:36:22 -0400

Replies:

Parents:

None.

On 9/9/06, Gerald Oskoboiny <[email protected]> wrote:
> Hmm... urn:sha1: doesn't seem to be on IANA's list of URNs [2];
> what's the recommended way to refer to files by their checksums
> if not urn:sha1?

http://en.wikipedia.org/wiki/Magnet:_URI_scheme

There's support for them in some p2p clients.

Mark.

Re: hash uris (was Re: excellent compilation of quake3 stunts

from Gerald Oskoboiny <[email protected]>, Fri, 15 Sep 2006 15:19:56 -0700

Replies:

distobj

Parents:

distobj

* Mark Baker <[email protected]> [2006-09-12 10:36-0400]
>On 9/9/06, Gerald Oskoboiny <[email protected]> wrote:
>>Hmm... urn:sha1: doesn't seem to be on IANA's list of URNs [2];
>>what's the recommended way to refer to files by their checksums
>>if not urn:sha1?
>
>http://en.wikipedia.org/wiki/Magnet:_URI_scheme
>
>There's support for them in some p2p clients.

I have never understood the point of the magnet: URI scheme: it
seems like a pointless wrapper around urn:sha1:, which says
exactly what I want: the file whose contents have the given SHA-1
checksum.

I can't believe there's still no standardized way to refer to
files by their checksums; proposals for md5: etc must have been
around 10+ years ago?

--
Gerald Oskoboiny <[email protected]>
http://impressive.net/people/gerald/

Re: hash uris (was Re: excellent compilation of quake3 stunts

from "Mark Baker" <[email protected]>, Sat, 16 Sep 2006 12:48:38 -0400

Replies:

None.

Parents:

On 9/15/06, Gerald Oskoboiny <[email protected]> wrote:
> * Mark Baker <[email protected]> [2006-09-12 10:36-0400]
> >On 9/9/06, Gerald Oskoboiny <[email protected]> wrote:
> >>Hmm... urn:sha1: doesn't seem to be on IANA's list of URNs [2];
> >>what's the recommended way to refer to files by their checksums
> >>if not urn:sha1?
> >
> >http://en.wikipedia.org/wiki/Magnet:_URI_scheme
> >
> >There's support for them in some p2p clients.
>
> I have never understood the point of the magnet: URI scheme: it
> seems like a pointless wrapper around urn:sha1:, which says
> exactly what I want: the file whose contents have the given SHA-1
> checksum.

I don't disagree, but whatever the reason, it can identify by hash and
is dereferenceable by some (reasonably) widely deployed software.
Seems like a win to me!

Now excuse me while I run off to the first W3C WG meeting on African soil ... !

Mark.