Nice writeup in Toronto’s Metro News about open source work that students are doing at Seneca College. Glad they’re getting the attention they deserve — it’s a great example for the rest of us.
All Canadians now have access to the Cochrane Library, a (more like “the”) collection of evidence-based medicine reviews and a critical health sciences resource. If you’d like to have a look, go to the Cochrane site; if you’d like access to continue, please sign the petition.
…so grows the developer (or something like that). Over at the “Practical Agility” blog, Dave Rooney has posted a piece about the way that pilots and developers both “revert to training” under stress, i.e., fall back on deeply-engrained habits when the going gets rough. Unfortunately, in the case of most developers those habits are ones that work in a student context, but not in the real world, like leaving things ’til the last minute. Fixing that is a bigger job than I can fit my head around… (Thanks to Adam Goucher for the pointer.)
Following up on last year’s post, take a look at Joey deVilla’s post about the employment.nil? Ruby job fair. Things like this are what “community” actually means—kudos to everyone who helped make it happen.
Interesting article by Navi Radjou about R&D investments in developing countries by multinationals requiring more anthropologists and fewer engineers than those companies may be used to. The same applies at a local level: companies that are used to working with large, closed organizations like banks often need to do a lot of mental re-gearing when they first tackle community-oriented mashups.
Interesting post from Simon St. Laurent about a recent TopCoder competition. As he says, “TopCoder runs programming contests designed to produce results for paying customers. The programming contests I joined as a kid were extremely abstract, put on by adults hoping to inspire us to learn. This approach turns competitive energy toward real problems, provided by companies who are even willing to pay for solutions.” Any philanthropists out there willing to put up the money to do this for public service software?
The 2009 Free Software and Open Source Symposium will be held at Seneca College October 29-30. I’ve enjoyed past editions; look forward to seeing lots of you there this year.
Nice summary from Diomidis Spinellis of a workshop titled “Software Architecture Challenges in the 21st Century“. Nice to hear Grady Booch say, “UML was never intended as a programming language. Models can be executed only in very narrow domains, mainly expressible by state charts. Those who think otherwise are fools.” However, I didn’t see any mention of what I believe is the biggest challenge: teaching this stuff. I taught “CSC407: Software Architecture” four times here at the University of Toronto, then lobbied to have the course removed from the calendar on the grounds that the material simply wasn’t connecting to the students. I still think something like my oft-proposed Architecture of Open Source Applications book (basically, software architecture by example) might work, but then, everything you haven’t actually tried might work.
Several times a year, it seems, I send around the link to Evan Robinson’s “Why Crunch Mode Doesn’t Work“. This time it was prompted by students who are trying to hold down full-time jobs while working on extracurricular projects that are almost as demanding. I recently came across this post on “Burnout” that I’ll include in future mailings; it talks about how good intentions lead to ever-lower performance due to fatigue, and what can be done about it.
I’m also going to start sending a link to “Unskilled and Unaware of It: How Difficulties in Recognizing One’s Own Incompetence Lead to Inflated Self-Assessments“. The authors showed that the less competent people are, the more likely they are to over-estimate their competence. Similarly, the more tired/burned out people are, the less able they are to recognize it—just think of how you felt the last time you were coding at 3:00 am, and how you felt about that code the next time you looked at it. The problem, of course, is that someone in the midst of an irrational frenzy is unlikely to listen when you tell them that…
Later: Leigh Honeywell sent links to a Fast Company artice about SAS and some quantiative data on the effect of sleep deprivation on group performance. Why isn’t there a required course on this stuff in first year?
Over on the Basie blog, Florian has posted an idea about using AJAX to get around one of the most annoying problems in DrProject: timeouts during lengthy batch creation of projects and/or users. Basically, his plan is to have the browser send one create request at a time, instead of sending a batch to the server and asking it to do them all at once (which often led to timeouts, since each creation can take about one second, and classes often have a hundred students or more). It’s theoretically less efficient (more network traffic, and N rewrites of the .htaccess file instead of one), but “slower and always works” is better than “faster and sometimes fails”. He’d welcome your comments…
Recent Comments