Synthesis: Operating system of the future?

The design principles of Synthesis, while powerful and generic, still have some major drawbacks. The algorithms are difficult to understand and implement for regular human beings (or kernel programmers, for that matter). As Linux has demonstrated, making kernel development simple enough that a wide variety of people can contribute has some significant payoffs. Another drawback is that two-word compare-and-swap is, shall we say, not a common feature of modern processors. Lock-free synchronization can be achieved without this instruction, but it is far less efficient.

Posted May 24, 2008 3:23 UTC (Sat) by olecom
As a matter of fact eight-year developer of PowerPC Linux port Cort Dougan has citations of Massalin's work in his research and application publications.
[...]

No, i don't want to prove anything, just share information and pumping up SNR in the damned www.

(btw, is it hard to do url wrapping with <a href=...> in <blockquote>?)

I started to read your text, but I tend to fall asleep (Awake since 16 hours), so I read on tomorrow. But as you say most are programmers or hackers and not software developers. If you do one thing on the easy way it will certainly fuck the next guy who is trying to use or maintain that crap.
I've done three month ago my bachelor of science in business computing, but they only showed us some java hacks and fast, quick and dirty development (more quick solution developing).
Today @work I can see the same bad stuff happen every day... and nobody is complaining.

Good night, I will do more reading tomorrow!

Laziness is cited because programs are often written to automate repetitive tasks. A lazy programmer will automate the stuff, which leads to better programming skills through real world applications, since if something isn't automated yet, it's probably because no one would or could figure out how to do it. Likely this is because nobody was lazy enough to not want to do the same dull IQ sucking repetitive tasks day after day.

If you do, you learn something and make everyone elses life easier, which leaves more time for you to focus on more intellectual pursuits, like writing a better tree search algorithm or implementing better code security, as opposed to always being reactive because someone forgot to do something and data is spilling all over the floor.

If you are not lazy and do the manual stuff by hand instead of automating it, that will quickly become all you ever do, boring repetitive tasks, which will cause development skill rot.

Laziness is probably the most important oft cited programmer characteristic. It's far more important than hubris or impatience for freeing up time to do more constructive stuff than vacuuming a database or checking your patches in one by one (for example).

Therefore, a lazy programmer is usually a very productive one. Someone who is not lazy and doesn't automate things, will find themselves doing very little coding before long. They'll eventually be overwhelmed with menial repetitive tasks and be very unhappy and stressed out. I'm convinced that the IT workers you see in the news, who shoot up the office, are doing so, in large part, because they weren't lazy enough, and lost "it" due to stress.

-AC

There are certifications for programming on the market. The author of the article assumes that developers of Open Source projects don't do testing. That's a blanket statement, just because he's seen it in the projects he's been involved with, doesn't mean that that can be applied to all Open Source software projects. OpenBSD, for instance, is known for going through a solid code auditing process.