Wednesday, November 30, 2005
(BTW, is anybody else digging the new SourceForge L&F? Seems like they must have beefed up the server infrastructure a bit, too. It feels faster, but that may be me getting distracted by the generally better-looking interface.)
Tuesday, November 29, 2005
The author, who shall rename nameless because he didn't make it easy for me to find his name, of The Blog That Goes Ping has some thoughts on Practical Common Lisp (via Zach Beane). He sums up with a nice indictment of Common Lisp's freeze-dried-since-1994 state:
Fun Fact about Common Lisp: standard Common Lisp has no networking libraries at all, but it does have a built in function to print integers as Roman numerals — using either the new style (14=XIV) or the old-style (14=XIIII) Roman numerals. Huh.
The biggest problem is that he's right. Doh!
Saturday, November 26, 2005
Here's a twist on computing's oldest peripheral... I happened to be reviewing something on the Finding Lisp blog page this morning and one of the AdWords adverts happened to catch my eye. I clicked and found "Das Keyboard - The Blank Keyboard." In this case, the keyboard doesn't have any labeling whatsoever on the keys. They're all blank. The Das Keyboard folks bill this as a way to increase typing speed because it will quickly break you of the temptation of looking down at your fingers to determine which key is which. The keyboard seems like a quality product (as much as you can tell from a web site, without actually having your fingers on one). It retails for $79.95 with free shipping to the USA.
Personally, I found this most interesting because I have wanted to teach myself to type on a Dvorak keyboard layout for some time. While you can buy Dvorak keyboards or even dual-marked keyboards, the most common way to learn Dvorak is by just using keyboard remapping software present in every OS out there. But, then I'm tempted to look at my fingers. That's particularly true when I'm typing "computer things" like code or OS shell commands. If I'm just typing long stretches of prose, I don't find much need to glance down. When I have to type
find / -name foo -print 2> /dev/null, or something like that, there is much more tendency to look down, particularly since Dvorak moved lots of the symbols to (IMO) non-intuitive places. This keyboard may be just the thing to help with that.
Or you could just buy it because you're an übergeek and it's cool in a minimalist sort of way.
Update: If you're cheap, you can just do this.
Update: An anonymous commenter suggested the Optimus keyboard. This thing, if it ever gets produced, would be very cool. I don't know how well the keys would hold up over time, but you'd get major props from everybody at the office. Essentially, it solves the problem of remapping keys once and for all.
Monday, November 21, 2005
I love it when somebody thinks the way that I do. For quite some time, I have thought that it would be cool to create a distributed compute architecture based on Lisp. The ability to send a sexpr from one machine to another for evaluation is a very simple way to make a distributed architecture. Well, Brendan Kosowski has been thinking similar thoughts and sent me a link to his Simple Grid Protocol library. I haven't had a chance to play with it much myself, but it looks really neat.
The Simple Grid Protocol is designed to allow users on a TCP/IP network or the Internet to run programs on their computer which utilize the unused CPU resources of other computers on a network or the Internet.
Brendan has a nice set of documentation on his site, with instructions for getting the whole thing up and running, as well as some examples. The only thing that's missing is a performance report showing the sorts of speedups that you might be able to achieve on various problems.
If you're interested in building your own Lisp grid, check it out.
Friday, November 18, 2005
It's been interesting to see the play that Ruby has been getting recently. I first heard about Ruby years ago but didn't think much of it. It seemed to be a bit of Smalltalk rolled up with some other concepts. Interesting, but not compelling. Rails seems to have launched Ruby to megastar status in the Next Big Thing sort of way.
What I can't quite tell is whether this is just people fawning over more Lisp features that have again crept into the collective conciousness, like GC when Java first became popular--listening to some reports you would have thought Java invented it--or whether there is something more to it. I have checked out the Ruby on Rails web site and it all looked like stuff that Lisp was probably capable of doing better, but it was hard to tell without spending lots of time on it.
This article describes why some key Java folks are starting to look more at Ruby. Now the question is, why not Lisp?