Seth Ladd's Blog

Over at Craig's stack trace , Crag has a nice post on a terse configuration for the Acegi Security Framework for Spring . Now that version 0.7 of Acegi is out, I had to make two small changes to Craig's acegi-common.xml file. This file holds all the default and stock beans needed for a generic install of Acegi. By moving all the boilerplate config out of the main configuration files, Acegi suddenly becomes usable. To make this config work in the Acegi 0.7 + Spring 1.1.4 world, you must make (at least) two minor modifications. First, for all the beans in acegi-common.xml that will be redefined in your application specific security configuration, you must add the attribute abstract="true" . This is in addition to lazy-init="true" . The abstract attribute forces Spring not to load up the bean, even if another bean references it . This is the case in acegi-common.xml where the httpSessionIntegrationFilter bean references other beans. Which brings me ...

We now have two extensions to Thunderbird that can read the X-FOAF header. Phil Wilson took the MessageFaces source code and modified it for X-FOAF . Woohoo, Phil! Meanwhile, Chris Schmidt has released and announced Thunderbird FOAF . It, too, is based on existing code . Woohoo, competition! A big thanks to Jens Bannmann, who created the MessageFaces extension. He outlines the difficulties of working with Thunderbird headers in extensions: Just to warn you: There are a few obstacles here. The first is that unless the user sets "View all headers", only the standard headers are sent to the UI. Unfortunately, the set of headers is hard-coded into the core (see https://bugzilla.mozilla.org/show_bug.cgi?id=236954 for future work to change this). To work around this, MessageFaces has a mechanism to switch the All Headers preference on just before the message is loaded, and to switch back before it actually gets displayed. This is a really, really fragile hack, but we hav...

I'm looking around for a dedicated hosting provider. I need a place to run my FlickrRdf code, plus other semwebby goodness. I'd like a dedicated machine so that I'll have all my options open. Any suggestions for a good dedicated hosting provider? I just need the basics, and I'm looking to spend around US$100/month. Oh, and it has to run Linux. Thanks!

Leigh Dodds created Slug: A Simple Semantic Web Crawler . Nice job Leigh! Because it's in Java, I'll be able to integrate into my FOAFverse work. Back in March I was tinkering with writing a Scutter. I'd never written a web crawler before, so was itching to give it a go as a side project. I decided to call it Slug because I was pretty sure it'd end up being a slow and probably icky; crafting a decent web crawler is an art in itself.

After taking a brief survey, it seems that X-FOAF is used more. I changed my Thunderbird installation to send out X-FOAF headers instead of X-FOAF-URI headers. Just trying to remain consistent.

Ian Davis saw my post on FOAF and Thunderbird , and found out how to add a FOAF URI to Thunderbird as a mail header . Thanks Ian! Ian said: Seth Ladd had an interesting idea: adding an x-foaf header to outgoing emails. I wondered if this was possible in Thunderbird and after a bit of digging I found that you can do it by adding the following to prefs.js in your Thunderbird profile directory I changed the code for the prefs file just a bit. I used: user_pref("mail.identity.id1.headers", "foaf"); user_pref("mail.identity.id1.header.foaf", "X-FOAF-URI: http://www.picklematrix.net/foaf.rdf"); I changed X-FOAF-Uri to X-FOAF-URI , and the values of the idXXX.headers pref must match the idXXX.header.YYYYY . Now I'm sending FOAF around the world! Part two: Setting up gmail to send arbitrary headers.

Thinking about FOAF and Thunderbird integration , for it all to work, FOAF documents should be signed ( PGP / GPG for instance). A nice little trust web would emerge. Signing a FOAF document (or, more accurately, a set of triples) would mean "I verify that these triples are correct to the best of my knowledge" Same principles would apply from signing others' PGP keys (from a trust perspective). So, in an effort to see if we can sign RDF triples, I stumbled upon the Tech Report: HPL-2003-142: Signing RDF Graphs . The Jena guys do it again! From the article synopsis: However, it is possible to define a large class of canonicalizable RDF graphs, such that digital signatures for graphs in this class can be created and verified in O (n log (n)). Without changing its meaning, an arbitrary RDF graph can be nondeterministically pre-canonicalized into a graph of this class, before signing. The techniques in this paper are key enablers for the use of digital signature technol...