10 ms·
This article confuses "open source" and "the Linux desktop". I consider myself an open source person. I spent a good chunk of my day today using open source t
by btilly 14y ago
This article confuses "open source" and "the Linux desktop".
I consider myself an open source person. I spent a good chunk of my day today using open source tools, writing an article that I will give away for free.
And yet, I am typing this on a Mac. I don't care about gnome or KDE. Yes, I used several Linux machines today. But I have not seriously used a Linux desktop in the last 2 years, nor do I care what is going on there. (The last window manager that I liked was fvwm, but many years ago a Debian upgrade added integration with Gnome I did not want or need, and it stopped working in the way I wanted. I have not bothered caring about a window manager since.)
I am far from alone in my views. (Except for liking fvwm. Most people consider that plain weird.)
- jmspring 14y agoAgreed, though I went with XFCE (over fvwm) until it got a bit bloated. My desktop requirement (laptop, in this case) is - unix based and just works.
- nwzpaperman 14y agowhere do you write?
- btilly 14y agoVarious places. Occasionally to http://bentilly.blogspot.com/ http://bentilly.blogspot.com/. But in this case it is an article on A/B testing, which is a response to http://www.evanmiller.org/how-not-to-run-an-ab-test.html http://www.evanmiller.org/how-not-to-run-an-ab-test.html. It will be part of a series. The one I'm polishing now is a rigorous technique that lets you look at results as often as you like, but lets you run the test forever. If there is any bias at all, eventually you'll get an answer, and have high confidence that it is right. The next in the series addresses the question, "I have a limited sample size I can realistically obtain for an A/B test, how should I run my test, and what kinds of guarantees will I get?" (With, of course, multiple looks allowed.) Hopefully this will be coming to a website near you in the next day or so. :-)