Just a notice that this blog is coming to the end of it’s life.
When it comes time for me to renew my hosting arrangements in the spring of 2018, I probably will not do it with this provider.
But, don’t despair! All of the content has been transferred to a blog on GitHub Pages. Just amble over to https://clartaq.github.io/yo-dave/. It’s all over there now.
River5 is the latest in a series of RSS News Aggretators written by Dave Winer. You can read his announcement here. These notes describe how to set up a River of News on a Virtual Server (VS) purchased or rented from CloudAtCost and freshly imaged with Ubuntu 14.04.
These notes assume that you have done some work in configuring your VS with a user, other than root, that has super user privileges. If you need help doing the initial setup of your VS, look here.
As some of you know, we’ve spent the last several months moving around looking for a nice place to settle down. That’s finished. We’re now in Salem, Virginia.
Before starting that search, we bought a laptop to take over the duties of the desktop. On the night before we left Tucson, I dumped a cup of water on it and fried the keyboard and other parts. It turns out the model was too new to get replacement parts. And since we had just spent $1,400 on this thing, my wife wasn’t about to let me buy another.
As a result, we’ve been using iPads for our computing needs with occasional jaunts to the library.
Oh my gosh! I thought I was going to die. Typing anything of any length on the iPad is horrible. Being back on the desktop with a keyboard is just so… good! I never want to go through that again.
(Still don’t have the laptop repaired.)
Well, Happy New Year everyone.
It seems like I’ve been hearing a lot of buzz about getting unconnected. Everything from hotels offering special deals on stays where you are forced to turn off (or turn in) all your gizmos, to advice to disconnect to improve your productivity. Could it be that the pendulum has started to swing in the other direction?
Sit back and enjoy the quiet.
My wife and I just returned from a long cruise to Hawaii and back. During that time, we were just about completely disconnected from the mobile world — no access to email, cell phone, fax. There was sporadic television available.
On balance I didn’t miss it too much. I continued to work on problems in my head, looking forward to the chance to get back to work on them when we returned. That eagerness to return was almost exactly balanced by the dread of dealing with all the emails that would be waiting when we got back.
Still, a nice chance to step back and look at the world with fresh eyes.
Today I came across a blog that describes a sentiment that comes up periodically. It describes a development group that does not believe it needs to follow the agile development process. The team is composed of experienced developers with a track record of delivering the goods.
I’m not much of a fan of agile myself, but I believe in process as part of software development. I think the real sentiment being expressed is that experience trumps process. Maybe that is an appropriate cost/benefit tradeoff in this case. But it is certainly not appropriate in all cases. It’s really a statement that failure of your system is essentially inconsequential.
Consider an airline pilot. They have gone through the process of flying a plane from one point to another hundreds of more times than you have started new development projects. Yet they go through the process of filing flight plans, the pre-flight checklist, and all of the other steps of assuring a successful, safe flight. Would you want them to act differently? No, because the consequences of a failure could be catastrophic.
My personal heroes in the area of software development are the Shuttle Flight Software development team. They write software that never fails. And they do it over and over. There are no super rock-star ninja code cowboys on this team. They achieve that performance by strictly following a well-defined, detailed process.
Nobody really cares if your social network clone crashes every so often. But if you write software that does something difficult and important, like controlling the braking system in a car, controlling a medical device with the potential to blind you, interprets medical information where an error could injure or kill a patient or cause an unwarranted abortion, or software that launches a space shuttle, you want every possible advantage you can get. A good development process contributes to those advantages.