Category: software

Live chat for and about Wayne County citizens

I spent some time tonight getting a live chat feature working on the ProgressiveWayneCounty.org website. I believe it might be the only live chat room up and running that exists for Wayne County citizens in general...correct me if I'm wrong.

Hmmm, I wonder if we're ready for that. There's been some demand for it over time. On other community websites I've been involved in, it was quite a useful way for folks to connect, but that's often over a geographical distance. I don't want to make it any easier than it is to isolate ourselves from each other when we don't need to be, but for those who might not have time or inclination to meet in person about the issues at hand, an online chat can sure be handy. We'll see.

Tired of social networking sites

At lunch today we were talking about all the social networking sites that have popped up on the Intertubes over the recent years. Mark and I sounded a little curmudgeonly about it, noting that we've long since been ignoring invitations to join the latest fad in making virtual connections to the rest of the world. First I was on BBSes back in the day. They I joined some mailing lists. Then there was Friendster, which kept losing my profile and whose software sucked. And then there was Orkut, which I signed up for because it was Googly but I wasn't popular enough to do anything useful with. And that's when I sort of gave up. MySpace, Facebook, LiveJournal, Linked In and all the rest can call me or send a representative to do lunch if they really want me. Except Myspace - puke.

Part of it is probably the sense that I'd rather be spending time strengthening ties in my real-world community than in an online one. Another part is just not caring. But most importantly, it seems the trend is such that soon we'll have one social networking site per each person with an Internet connection, and we'll be back where we started. I've got enough passwords to remember as it is, okay?

But it is funny to me when the networking sites scrape information off of my website in an attempt to make me look like a member. Like Spoke, which just did this without asking. Unfortunately, they get some significant things wrong, e.g. listing Earlham College President Doug Bennett as the president of my company, Summersault. To be clear, Doug has not left his position as the president of an internationally known liberal arts college to serve as President of a website development firm.

If he is looking for that kind of position with us, we're open to, um, networking with him...without the help of a website.

Teaching software engineering

Today is the first day of the course I'm co-teaching at Earlham this semester, CS345: Software Engineering. I'm excited to be back in a classroom again and thinking in new ways and on different levels about a topic that's very much a part of what I do every day for Summersault (and why Summersault exists at all). Like the last course I taught at Earlham, this is sure to be a challenge and a joy. Wish me luck! I'll hopefully have some time to post general thoughts about teaching software engineering here, but the course website will be my main tool for collecting and sharing information and resources related to the class.

Homegrown RSS feeds for local news sites

I recently wrote in the Summersault weblog about how website content syndication is changing the way we use the web. Of course, if the sites you care about aren't syndicating their content, the phenomenon is a little less exciting. For me, the only ones I cared about that were in this category were local news sources. For a few years I'd been running hacked up Perl scripts that tried to retrieve headlines from the sites of local papers, but it tended to be hit and miss and I never really took the time to take a more structured approach until now. Using the Template::Extract Perl module and a handy article from Simon Cozens about spidering hacks, I was able to formalize and simplify the creation of custom RSS feeds scraped from sites without them built-in. So, here are the first few, to which you are welcome to point your own newsreader:

Please avoid refreshing these more than once an hour. If there are other feeds you'd like to see, let me know. I'll add new ones on my feeds page as they come. Enjoy!

Read your congressperson's blog

The eminent and celebrated E. Thomas Kemp points us to a wonderful and clever use of news aggregation and weblog technologies, Plogress. Using Perl and WordPress, the apparently anonymous administrator has created a site that sucks data out of the Library of Congress and displays a blog of the doings of individual Senators and Representatives. Now I can keep an eye on Mike, Richard, and Evan through my RSS newsreader! I'm sure they all read my blog, right?

Perl script to check Sprint PCS minute usage

After a few incidents of going over my Sprint PCS minute usage without any warning, I hacked together a Perl script that will log in to the sprintpcs.com website and check my minute usage for me. It e-mails me a warning if I go over a certain usage percentage. I thought I'd post it in case anyone else might find it useful: sprint-minutes.pl (link removed March 4, 2017, no longer works). I believe Sprint charges something like $3/month for a similar service. Note that you need a few extra modules installed, and you need the components for WWW::Mechanize that allow SSL connections. Consult your local Perl expert for help with this if you need it; I can't provide support here. It may well require tweaking depending on how standard your calling plan is, or other variations that I don't know about. And of course it's subject to breakage when Sprint updates the HTML code on their site.

Any comments or refinements are welcome. Enjoy!

Comair flight cancellations caused by a 16-bit counter

Over the holiday weekend, Comair had to cancel over 1,000 flights because of software problems. It turns out that, as I read in the F-Secure Weblog, the flight planning software they were using was using a 16-bit counter to keep track of flight staff changes...so after 32768 changes it would simply stop working. Details are available from an article in Cincinnati Post. This is the kind of madness I was expecting for Y2K, not four years later when we're supposed to have learned that lesson by now.