Wednesday was the travel day, and fortunately I was able to check into the hotel early so I could get some sleep before the Oslo.pm meeting.
I left the hotel to go the the Oslo.pm meeting too early. My little tourist map didn't have a legend on it so I didn't know the distance I was going to walk. It turns out I got there way too early. Oh well. I took some notes on my projects for the weekend. Time away from the keyboard is a good way to avoid making bugs.
At the meeting, Salve started pushing people to think about what they would work on, and most of it has made its way into the QA wiki in a nicely organized form. I'm allergic to wikis so I've avoided updating my information there. Even if you aren't in Oslo, you might be able to help by working remotely by watching the qa IRC channel or whatever other communication methods people set up. I'm also allergic to IRC, but that's okay because I'm actually here.
I spent a lot of the meeting talking to Cosimo, who works in Oslo for Opera Software (I snapped a photo as I walked past their building). He tells me that they use a lot of Perl, and, apparently, if you want an job, you can work all over the world for them. Since they are a norwegian company, I'm trying to use Opera as much as possible while I am in Oslo. It's a free download! So far I like that clicking on the part of the tab with the go-away button doesn't make it go away; you first have to get focus. Any Firefox developers reading this? Steal that feature please!
I woke up way too early on Thursday, and after doing a bit of work on my Module::Release project I went down tothe hotel lobby for a bit of breakfast. I was surprised to see Gabor there because that would mean his QA class hadn't started yet and it was still really early. I sat with him during breakfast and we talked about the training business. How do you get a training coordinator to see the value in good training when the outcome doesn't affect the training coordinator? Many companies don't know the difference between attending just any training and students learning what they are supposed to learn.
For the rest of the day, I noodled with various things and wasted a lot of time figuring out that I had to reset my Sourceforge SVN password. Apparently they had some sort of security concern so they reset the passwords. This didn't affect my login to the web site though, which is supposed to be the same password, so it took me a while to figure out what was going on when I checked the site status page, which I had to google to find. A nice notice when I logged in telling me to reset my password would help. As always, Sourceforge's site documentation is either deficient or cleverly hidden. Because I use Module::Release, I can't release my code until I check it into a repository. Having fixed that, you can now play with Module::Extract::Namespaces and Module::Extract::VERSION. I'd like to see if anyone can come up with some perverse, but legal, Perl that can break either of those.
Later in the day I started working toward integrating multiple perl binary tests in Module::Release. I want to be able to test a module with several perls before I upload it. The trick there is a repeated cycle of unpacking the dist to get a fresh version, using the next perl to run the build script, and then running the tests to see the output. The next part is to figure out what happened. With the latest Test::Harness stuff, it's really easier, but the older Perls don't have that and I want to test with something as close to the standard library as possible. I'll have to think about that a bit more.