W3C home > Mailing lists > Public > public-cwm-talk@w3.org > October to December 2009

Re: working on "A Model of Authority in the Web"

From: Henry Story <Henry.Story@Sun.COM>
Date: Thu, 17 Dec 2009 21:11:53 +0000
To: Dan Connolly <connolly@w3.org>
Cc: public-cwm-talk@w3.org, Jonathan Rees <jar@creativecommons.org>
Message-id: <769D7A86-DE90-40D1-9617-741026EC5BEE@Sun.Com>
Just last week I was arguing in a similar way [1] that one could use some such modal logic to prove that there is a missing HTTP server request, which is causing the latest HTTPS bug [2].

This modal logic would also need to make statements about degrees of belief, or degrees of trust in statements made by others whilst taking into account the quality of the connection.  

Particularly a rule that shows that merging information from different levels of trust brings the merged graph to the the level of the least trusted graph.

ie. something like

G1 trustLevel good .
G2 trustLevel bad .
G3 trustLevel veryGood .

[ union G1, G2 ] trustLevel bad .
[ union G1, G3 ] trustLevel good .

Anyway, there is an important issue there in the https stack that requires Web Arch involvement. It seems that very simply the server should be able to ask the client to resend its last request within a more secure channel.


[1] http://lists.foaf-project.org/pipermail/foaf-protocols/2009-December/001040.html
[2] we were made aware of this bug by Melvin's post:

On 14 Dec 2009, at 23:44, Dan Connolly wrote:

> It's been a while since I used cwm for real work, but
> it still seems to be a good framework for thinking.
> In last week's TAG meeting*, Jonathan brought out ABLP
> logic as a way to attack some HTTP semantics questions.
> His very brief presentation engaged much of the TAG,
> and I took the ball to write it up in prose.
> Of course, it's against my nature to write anything formal
> without checking it by machine. I just got the 1st full
> example working; feedback is welcome:
>  http://www.w3.org/2001/tag/dj9/story.html
> For details on how to run the example through cwm, see
>  http://www.w3.org/2001/tag/dj9/Makefile
> Next step: get cwm to spit out a proof and format
> it nicely. Wish me luck.
> Once I've got that much, I'll probably take it to www-tag...
> * minutes in progress, but you can pore
> over http://www.w3.org/2009/12/09-tagmem-irc
> if you're determined.
> -- 
> Dan Connolly, W3C http://www.w3.org/People/Connolly/
> gpg D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E
Received on Thursday, 17 December 2009 21:12:57 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:01:06 UTC