Coding

Aaron (one of my co-workers), recently posted a link about OpenID. I’ve given OpenID only cursory glances over the last year, but the Coding Horror link in Aaron’s post had a comment to this Google Video where Simon Willison gives a Google Tech Talk on The Implications of OpenID. The video is nearly a year old, but to date, it’s done more to convince me to get on the OpenID bandwagon than anything else.

I recently had an odd situation come up, though it’s not incredibly hard to imagine. I needed to create a staging web server, identical to production, but it should live on a different hostname (eg, staging.sherman.bz, instead of www.sherman.bz). The content should not accessible to the public, nor to search engines. This isn’t really so much a privacy or security issue as it is a convenience and customer service issue. If search engines somehow picked up on our staging site (and they would, given it’s full of SEO triggering info, google analytics scripts and the like) then our search information could become poisoned with the staging URL. Also bad would be if a customer found the staging site and posted information to the wrong place. I’ve actually seen both of these happen in the past when measures were not in place to prevent it.

So there I was, poking around in some java / j2ee code, trying to learn how it all works. I did some testing on a Linux server and realized, something is broken. It seemed something was corrupting ALL the Jar files in WEB-INF/lib/. A co-worker guessed that the token filtering Ant was doing might be the culprit. He was right. It seems Ant has issues with detecting whether files are binary or not, given that it uses a Reader class which runs the files through a character decoder.