Re: slow down and organize

On 26 Feb 2011, at 13:06, Cosimo Streppone wrote:
> On Sat, 26 Feb 2011 22:26:22 +1100, Henry Story <> wrote
>> There are a number of other implementations out there, listed at
>> in perl, python, c, php, ...
>> I hope Clerezza gives them some competition soon, so they improve their user friendliness
>> What we need is to make sure we all interoperate, which is why I am really keen on getting a test suite out.
> I think this is an area where I could definitely contribute
> if Perl has decent support for everything involved
> (thinking RDF/SPARQL...).

I just opened ISSUE-49: "Keep a list of implementations of WebId up to date"
So that we can at least track new ideas in the mailing list.

I think what you want is Toby Inkster's

he's our perl guru here.

If you want to compare with a Java implementations you can look at

- keygen is the library that deals with the server receiving a request from a  
   <keygen> html5 tag 
   But really that has been moved to the apache repository.

- foafssl-java/foafssl-verifier-sesame
   is the package that deals with verification
  This is implemented in Scala, and will be more up-to-date in the apache code base
where it has been built to go right into the SSL layer

I think there are some C libraries that act as plugins into Apache that work at that layer too.

> [snip]
> Is there existing source code for the prototypes
> shown in the video?

On the certificate creation side it is just
the keygen element and some Velocity templates for XWiki that have not been kept up to date

The equivalent and probably a lot more readable for Clerezza is

especially the

which I am just working on to allow for friend additions.

The complexity in the presentation layer comes mostly from having to deal with Internet Explorer's keygen equivalent: an ActiveX component shipped with all the browsers. There is also a trick one needs if one wants to generate pure XHTML pages that work with Netscape. (you need to look at the javascript pages to find those)

>> ne.
>> That is what the following proposes.
>> Does that make more sense when put like that?
> Yes, it does. Thanks!

Now we don't have any implementations for that idea. The protocols is written out so that we
could work with that too, since we are working at the semantic level (what the document says)
rather than at the syntax layer (ASN.1 vs RDF/XML vs XML vs html vs JSON vs N3) 

The advantage of RDF and especially xhtml is that the profile page can
be in a human readable format so that when people click in their browser on their webid (a simple UI feature that could be added) they can be brought to the profile page Seeing a X509 file would probably put many users off. :-) 


> -- 
> Cosimo

Social Web Architect

Received on Saturday, 26 February 2011 12:53:42 UTC