Category: technology

It had to happen at some stage I suppose. (although I’m still not quite sure why???). Blogshares enables users to trade blogshares similiar to a fantasy stockbroker game.
Blogs are assigned monetary values based on the number of incoming and outgoing links to other blogs. It’s similiar to Google in that it measures ‘connectedness’ Currently this blog is very lowly ranked 🙁 Probably due to the fact that most of my friends don’t actually maintain blogs so the usual web of trackbacks is avoided. It could also be that I haven’t said anything interesting. I hope not. Also my blogroll is generated using javascript and it appears that the blogshare parser has failed to pick up these links…

Picked up the following link from Jon Udell about the CAPTCHA (Computer Automated Public Turing test to tell Computers & Humans apart) preventing blog spam. This is a really (should that be raelly) tasty idea from Rael Dornfest. It can be summed up as can’t add, can’t post. He uses the Blosxom Writeback function which provides weblog comments with write-backs. An arithmetic sum is embedded in the writeback and no commets are allowed unless the answer is posted correctly. An example of this is
5 + 2 =
Neatly sidestepping more general blog spambots. The numbers are generated randomly. A definite improvement would be image obfuscation (a la Captcha!) and a bigger range.. He currently only uses 0-9 meaning a 1 in 20 chance you’re gonna get the right number. I’m not sure I want to encourage blogspammers to brute force my site, especially when a post is so tantalisingly close
I’m working on my own interesting weapon in the battle against blogspam. It currently has the catchy title of blogassasin (Apologies to jmason & the rest of the spamassasin team). Also, it doesn’t kill blogs but early versions come close. Active blacklist generation is another tidy feature. So spammers should think before thy HTTPiss Off innocent bloggers. Personally I don’t believe that my blog (or anyone else’s for that matter) needs to become any less relevant or increasingly grbled. So let’s say NO to blogspam 😉

Just read an interesting post on Steven Vinoski’s middleware matters about the lack of multiple-port support in the End-Point Reference EPR currently under review by the WS-Addresing working group and augmenting WSDL 1.1 by allowing for more dynamic usage patterns. Currently the EPR doesn’t suport multiple ports. Ports, for those that remember the original WSDL spec, enable a webservice to be accesible through multiple protocol/transport/format alternatives. Steve Vinoski proposes a useful “business card” analogy to explain the practicality of multiple ports, covered by one EPR. Personal addressing on the internet has arguably evolved this way anyway. Here are some examples of ports associated with the person Shane Dempsey (of Geesan Tech, for SPAM avoidance purposes) with basic URI: sdempsey@geesan.com

• EMAIL: mailto:sdempsey@geesan.com (TCP, port 25)
• MULTIMEDIA SESSIONS: sip:sdempsey@geesan.com (UDP/TCP, port 5060)
• SIP INSTANT MESSAGING: im:sdempsey@geesan.com (UDP, port 5060)
• JABBER INSTANT MESSAGING: sdempsey@geesan.com (TCP, port 5222)

In some cases the URL scheme is provided, indicating a particular port (e.g. mailto: implies SMTP). The use of schemes is far from uniform however, meaning that there is not a direct port-scheme correlation. In the service domain, this is better. For example a SOAP service where information is transferred over an alternative application layer protocol such as SOAP, SIP or SMTP is possible. A hyperlink to such a SOAP service would take a form similiar to mailto:soap@mydomain.com

Save time writing a difficult and boring java style guide! Refer to William Blake’s handy page which covers everything from Java exceptions to method and variable naming. I used to have something similiar in my lecturing days but this version is definitely better.