Still working on that release-management and configuration tool at work. Writing tests for Net::ICal in my spare time, trying to get things working so we can get a snapshot release out. Oh, and sleep and bike-commute too somewhere in there.
Wow, time seems to just disappear, and suddenly I've not written in a week. Much has been going on.
So, I had this brainstorm about a timezone server--- a web service, probably XML-RPC, that I can ask questions like, "I think it's noon. What timezone am I most likely in?" or "When's my next Daylight Savings Time shift?". Reefknot could use that.
We've all been running scared from timezones--- they're Hard to deal with in a sane way. For one thing, every platform that Perl runs on handles timezones differently--- a web service would reduce some of that complexity. This idea looks interesting enough that I want to make it work though--- as well as figuring out in my own mind whether web services are really useful or just The Next Big Buzzword.
So, I poked around with Damon Chaplin's Olsen->VTIMEZONE conversions, which give useful data for accomplishing the functions I want. That means I'll end up writing Timezone.pm, which we've (also) all been throwing up our hands and running away from.
Read a post from Miguel about why not to run away from scary technical problems. That provided me with some energy. Now all I have to do is focus it.
In fact, focusing the energy into a release right now would be Really Good. All I have to do is find some time.
So, I seem to be cat-herding volunteers for the idea of a YAPC in Boston next year. I've been meeting interesting people and talking to them about finding space at a college in Boston where geeks can have a conference. If you want to help, mail me. (Or even if you just know someone in Boston who could help us find space.)
Rearranged our home office/server room/rackspace this weekend. That's a large task, especially the first time. Hence, not enough time for Reefknot. Such is life. Oh, and I've been biking, which is making me happy.
Spent yesterday at Damian Conway's Advanced OOP class, which was really interesting. Got a new set of ideas and some inspiration, which is always good. Now to get my home network to behave again, so I can actually hack on those ideas.
I've also begun to collect volunteers who are interested in organizing the YAPC conference in Boston next year. I don't really want to be the chief cat-herd of that event, but it may yet happen. (If you're in Boston and want to know more, particularly if you have ideas for colleges who'd be good for hosting it, or if you know of good corporate sponsors, mail me.)
Listening to Who is Jill Scott?. Good stuff. I recommend it highly. Bought the New Riders book on Vim, and I'm absorbing it slowly.
Worked on Reefknot some more--- trying to convince Date::ICal to do the right thing. I'm working on an automated deployment system at work, so that we can use a consistent process to make developer sandboxes, staging servers, and production servers. Other than that, nothing much is going on.
Hacked a little this weekend, but mostly spent it hanging out with friends. Recruited some new potential Reefknot people, tried to get Time.pm and its tests to work (yet again). Bought a bike computer, which is proving to be interesting for comparing my day-to-day rides. Other than that, nothing much.
Worked on Net::ICal some more last night, trying to make sure that my integration patches for Date::ICal are working properly. Chatted with Schwern for a while on the API for Date::More and on automated testing. Committed patches and tests that I think make Net::ICal::Time saner. One file down, 10 or so to go. If you like testing and QA, or if you're just looking for somewhere to help out, mail me, we could use you.
Packaging our internally-developed modules into CPAN-style packages, migrating from RCS to CVS, and thinking about automated testing of lots of homegrown Perl modules. Mmm, release management.
Worked on Date::ICal last night, adding a compare() function to it, and started trying to make Net::ICal inherit from Date::ICal in the right ways. Next up: finish that, work on testing.
Doing some cool stuff at work with our Perl modules there--- setting up tests and proper Makefile.PLs for the ad-hack modules we have lying around. Starting to think about how to do release management in a Perl development shop. It's fun.
More hacking last night--- lots of idea spewage, mostly, and defining the API for Rk::Storage in between chatting with Schwern about automated testing. Had an idea about distributed testing of our code, so that we know it does the Right Thing in different timezones. Essentially, I think we're going to have a developers' version of "make test" that sends mail to a "reefknot-testing" list. More as I get it figured out.
I've still gotta figure out how to balance hacking with having enough energy to ride my bike 50-80 miles per week. hrm.
Worked last night on Reefknot for a few hours-- the first good chunk of time I've gotten in on it since YAPC. Started trying to figure out how Storage should work, using some of LotR's code as a base. Wrote some testing infrastructure for the Reefknot::* modules.
Other than that, I did household stuff and tried to convince debian to play nicely with me on my (newly reinstalled) machine.
New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.
Keep up with the latest Advogato features by reading the Advogato status blog.
If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!