[13:46] <GabeW> dnm: I think its hilarious your job options are a job in starbucks or a $125k a year security position ;-)

Yeah, you're not the only one.

Do people using those little orange XML buttons for linking to their RSS info on their sites realize that it's completely misleading and pointless? I might as well have a button reading "Text" or "Bits". The key is that it's RSS, not that it's XML.

The strictly hiearchical chain-of-command isn't working.

Found on this Blogdex, thought I might as well feed the meme beast: Weary, Bush mocks reporter. I'm simultaneously amused and frightened. Pay no attention to the multinational protests behind the curtain. Oh, and how dare you be a decent journalist and actually use the language of the country you're in to address its head of state? Foolish mortal. I find your lack of faith disturbing.

Oh yeah, I get Blogdex now.

Two diatribes for now:

XML:

Well, let's just say that beyond bandying about buzzwords for effect (be it half-genuine effect in front of suits, or completely dis-genuine effect in front of friends who know better), I wouldn't want to discuss XML in polite company. The best presentation I've read about XML lately is Aaron Crane's "Does XML Suck? or: Why XML is technologically terrible, but you have to use it anyway". The PDF is gold.

Programming languages:

There's a thread going on about Python v. Lisp at ll1-discuss as part of the fallout from Paul Graham's latest article. Yawn. I've gotten past the point where I care all that much about language X v. language Y, really. At the end of the day, I think it comes down to personal matters of taste, comfort, experience, and effort. If we assume that all Turing complete languages allow you to construct systems with greater levels of expressiveness (via abstractions) on top of what they already possess, then all languages are equivalent modulo effort. The argument about using "the right tool for the job" isn't wrong as much as it is uninteresting. I don't have any problem with that sort of pragmatic thinking, but if technology didn't evolve, the right tool for the job for making fire would still universally be flint, stones, and twigs. These days, Bics are more common, though you may still fall back on the former in a survival condition. I'm much more interested in making better tools than and advancing the state of the art than I am doing what the majority of programmers need to do in the majority of their work the majority of the time. Forgive me, but majorities almost universally bore me to tears. That's a rut I'd like not to get stuck in.

What does this all mean? For one, I think it's pointless to wage wars about Python v. Lisp on terms of which language is better. In all but a few cases (let's say in the majority [gasp!] of cases), people are using a particular programming language because it suits their needs the best. This is also a self reinforcing pattern, because when you use something and it turns out not to work as much as you might like, you are more apt to try something else. People solving all or most of their problems in Python are unlikely to switch to Lisp, just as people solving all or most of their problems in Lisp are unlikely to switch to Python. The issues isn't that Lisp is better necessarily so much as it is different, and the same goes for Python. When I encounter programming language arguments these days, I tend to mentally see them as though they were arguments involving spoken or natural languages -- "French is better than English! German over Russian! Esperanto uber alles!". I want to be able to draw ideas and inspiration from as many wellsprings as possible, and I would like to be able to communicate with as many different people and cultures as I could. Therefore, learning new languages is fun!

Such it is with programming, with only a few modifications. Lisp is great; I like Lisp. I'm not interested in defending or championing Lisp. I'm interested in using Lisp. All things have their time. C is great; I like C now for what it's good for, but only given a lack of better options; I'd love to get past C. I'd love to get past Lisp too, eventually.

I'll write a better, more coherent screed on this topic soon, if people bug me about it.

Note that this doesn't mean I don't like reading these sorts of threads -- on the contrary, I quite enjoy them.

So what's the coolest new language out now? REBOL. REBOL rules for all things Internet, and as we all know, that's pretty much all anyone cares about these days. The only real downside I see to REBOL at the moment is the single vendor-provided closed implementation. It's still pretty darn impressive though. And REBOL/View is even neater. Languages like GOO and SmallScript and Mondrian are on the horizon. And I have to plug Who, Me? just because. Even given the weird name.

Other stuff:

Linking in weblogs is more difficult than I'd prefer it to be.

I almost forgot: MzScheme and MrEd went to version 200alpha19 in the PLT Scheme CVS tree yesterday. New features include a portable 'subprocess-kill' function and a plethora of squished bugs.

Coolest new Lisp toy for today: AP5. Thanks go out to Brian Rice for sharing AP5's existence with me on IRC. My brain is now horribly twisted into all sorts of delightful little thought pretzels.

From Adam Langley's weblog: Paul Graham has a new article out on Lisp. In the reading of it, I had though several times that I had read this paper before, but no, it is new. However, what Graham has to say and how he says it does seem an awful lot like his previous articles. That's fine, I have no problem with that, it was just eerie.

Ah. That's why. A lot of the material in this was originally posted to ll1-discuss.

I recently mailed the following to a fellow Lisp hacker and thought it might make for mildly informative reading:

You know of any free tools that support GUIs for CLisp?

There _might_ be ports of Garnet and CLIM, or derivatives thereof, for CLISP, but they're likely to be uh, less than impressive.

Or one where I don't have to make alot of FFI calls?

Depends on what you want to do.

I myself recently cut down on some dead weight and now pretty much stick to Corman Lisp and CLISP on Windows (after having run limited trial versions of both Franz Allegro CL and Xanalys LispWorks forever) and OpenMCL and CLISP on Mac OS X.

CLISP isn't great, but strangely, I'm much less averse to using it than say, CMUCL, which I can't stand. I like SBCL as a free Lisp on Unix platforms, but it needs more work before it's something I'd consider using in anything close to a "production evironment" (which it seems to be getting, thankfully).

Allegro CL and LispWorks strike me as being fine commercial products, I just have no money with which to buy them at the moment. Franz's pricing scheme is legendary for being obtuse. Xanalys seems much more approachable. The one commercial Lisp I'd buy in a heatbeat is MCL, and they reduced the price of their 4.3.1 version (which is built for Mac OS 9, but will run in OS X inside the Classic environment) to $95. Why haven't I bought it yet then? I don't really have a use for it. If I were writing Lisp professionally, or if I had a steady job with regular income, I probably wouldn't think twice, even though part of me would like to hold out until 4.4, the OS X native version Digitool is working on, is released. Plus, OpenMCL, although without the polish and GUI development features of MCL, is one hell of a Lisp (fitting enough, due to being the MCL core), and it's free.

Oh yes, I almost forgot.

I downloaded and installed the .NET Foundation SDK and SP 1 update yesterday. Thanks go out to Joey for help in figuring out which pieces I needed.

I have commenced playing with Mondrian (beta 5) and I await the July 2002 beta release of SmallScript for .NET. I'll be taking a look at Chris Double's work on GOO for .NET, as well as Hotdog, soon.

I got sucked into reverse engineering some software these past few days, which was devilish fun. The problem with having interests in so many specializations of a given field is that you tend to flip-flop a lot. One day I might be hacking on BEEP in Ruby, and the next I may be knee-deep in disassembly and hex.

I am still job hunting. One new development is that I'm considering attacking the technical writing market, given my technical background and this book project. I also like writing, which helps.

The archives are still not working correctly. I'll be fixing this shortly. Watch this space.

New additions to the list of blogs above include Karl Schroeder, Nicholas Riley, and Zimran Ahmed and Faisal Jawdat. I know of Faisal through Joshua, of Memepool fame, but I haven't yet had the pleasure of meeting him. Karl and I were co-workers at OpenCola. If you like scifi, I highly recommend you pick up Karl's "Ventus" and his upcoming new novel "Permanence".

I think the object of the game is to spread good ideas as far and wide as possible. Convince everyone - not just the people we like - to think for themselves, share their thoughts, publish their ideas in whatever format they can. A book like Coercion, for example, which attempts to untangle the marketing dream in which so many people are living, is much more important for people at the mall to read than people at the cybercafe. It's easy - maybe too easy - to exchange ideas with the other leftie intellectuals at the coffee house. The hard part is getting mom and dad, Mr. and Mrs. Gap, or the people marketing Starbucks to understand the real dangers of communicating in a coercive fashion, and to think for themselves instead of letting some corporation's bottom line think for them.

-- Douglas Rushkoff <http://www.rushkoff.com/2002_05_01_archive.php#85098594>

I forgot to chime in about Frost, which Chris Double was talking about today. I saw Frost on sweetcode and I just wanted to throw in a pitch (since Frost reminded me of it), such as it is, for LPP, from the fine folks at Interhack.

John Wiseman just rocked my world. I've followed Chris Double's work on the Corman Lisp mailing list and through his website, but I didn't know he was hacking on GOO (makes sense, since I know he's into Lisp and Dylan). I knew of Hotdog, but didn't know about it's creator. I kept reading excepts of Patrick Logan's posts on Paul Snively's blog, but hadn't thought to look at it myself. I'm happy I've got new and interesting links to add to my morning/afternoon/evening clickery. Thanks John!

On the programming language design and implementation (PLDI) front as regards interesting .NET things, I found out about Mondrian ages ago, but I don't have the .NET runtime, so I haven't been able to play with it. Anyone have a review?

Besides the TOPS-20 fun mentioned earlier, I compiled rc under Cygwin (although, unrelatedly, I still can't update Cygwin) yesterday. In the paraphrased words of Steve Jenson: "Ah, rc, how I've missed you". Also, some fun has been had with new C/C++/Java coding aids, including an amazing lint.

A friend clipped an article (in French) out of what I think was Photo on the new Leica M7. "It will be mine... oh yes! It will be mine."

@LOGOUT

 KANKAN, PANDA TOPS-20 Monitor 7.1(21733)-4

@

I'm so giddy right now.

I added all the blogs I have in my Opera bookmarks to the top of the page. Enjoy the tasty linky goodness!

What is with Maplesoft? They must have picked up some agressive marketing and sales people that are driving development these days or something. I checked their site today out of idle curiosity, and they've released Maple 8. I remember Ethan buying Maple 6 when it was brand new during winter (beginning of the year) 2001. I'm using Maple 7 on my Windows box at the moment. Why would I upgrade to 8 already?

It just seems like they're releasing too often. I hope they're giving deep discounts on upgrades, because otherwise I don't know how average folks keep up. Unless, like me, they don't bother upgrading with every major version.

Netinstallers have been giving me trouble lately. Netinstallers you ask? I mean the installer program that's used by Cygwin and XEmacs (on Windows anyway) which connects to a site, downloads some setup files (like mirror lists), and walks you through the process of downloading which pieces you want and installs them locally into a working system.

Neither the Cygwin nor the XEmacs netinstaller I have are working. Working as in, when I tell them to connect to the Internet (either via a direct connection, using IE 5's settings, or via my HTTP-only localhost proxy), they send packets and do the typical sort of network things. They aren't. I checked.

They're also not getting far enough along to write any error output or logs, apparently, which is even more frustrating.

From Neil W. Van Dyke's weblog, I found this New York Times story of a power company that's buying a whole town in Ohio to the tune of $20 MM USD. Why? Well, for one, they pump lots of toxins in to the air, and the residents aren't pleased. $20 million may seem like a lot, but all the citizens of the town are getting roughly three times the market value for their homes in exchange for never ever suing the company property damage or health issues. I cannot begin to grasp how wrong this is.

Dare we believe it's actually out?

On Tuesday, May 14, Stephen Wolfram's book A NEW KIND OF SCIENCE is published and should start being available at all major bookstores. Representing the culmination of nearly 20 years of work by scientist, Mathematica creator, and Wolfram Research CEO Stephen Wolfram, the book is being heralded as a major scientific achievement, as well as being highly readable. Wolfram credits Mathematica as the tool that made his science possible. The notes at the back of the book contain nearly eight hundred Mathematica programs that illustrate the power of Mathematica programming and can be downloaded on the website...

http://www.wolframscience.com

Happy Mother's Day Mom!

Highly strange traffic issues occurred recently.

I uninstalled an old executable build of Standard ML off my Windows box in an effort to reclaim some space. Also killed off in the purge were a few MP3s, VisualWorks 5.4i, Clean 2.0, Aonix ObjectAda, Joy and a few other things. My OS X box has taken up the task of keeping current with SML and Joy, which is easier to accomplish there anyway.

I resurrected and partially updated old pages which were missing from the site for a long time, notably the CLAS and R6RS project pages, which I saw some people were still visiting from Advogato.

Built and installed the new scsh 0.6.2 release on my PowerBook. It's very nice, with many additional SRFIs implemented in this revision, and a clean up to align with the Scheme 48 module system.

Two new tasty treats arrived my way via Apple's Apple Developer Connection (ADC) email newsletter #300 recently: First, updated documentation on Cocoa, with no more "description forthcoming" placeholders (yay!) and second, new docs on kernel hacking, which means more kernel extension (I love microkernels) tom-foolery in the near future (yay!).

I leave for Houston tomorrow just before five, and spend all day there in a slew of interviews. I'm back Wednesday, later in the day.

In the spirit of keeping up with the Joneses, or in this particular case, the Jenson, I did a cursory search of my books packed up downstairs recently, in an effort to reestablish my chess library. I found a Pandolfini book on endgames and my copy of Bobby Fisher's classic text, but no other books. I can't imagine for the life of me where I put or who I let borrow my copy of "Modern Chess Openings" or "My System", but I do remember noticing MCO was missing before these books were packed in May of 2001. I also had the four part series by Yasser Seirawan, but I believe I let my good friend Mike Carey borrow those when I was still in high school (Mike, on the off chance you're reading this, get in touch! I'd love to hear from you!) It might have also been high school AP Physics chum Neb, who I played occasionally (Neb, on the off chance you're reading this, get in touch! I'd love to hear from you!).

It was Ross Hayes, a dear friend, who really got me into capital-c Chess. I had played off and on before with friends and relatives of course, and had even been in small chess clubs, but it wasn't until I played Ross and learned from him that I started realizing the enormity of the work I still had before me. It was from Ross I got MCO and the Seirawan books, along with his old copies of Chess Life and a small, portable chess set from Germany. I still have the board, and I'm glad that I do. I think of those memories every time I use it. I hope whoever I left the books with is making good use of them and has learned more about chess. If they're still lurking in with my packings, they'll be unearthed soon enough.

These days, I futz around with computer chess engines, mostly Crafty, which plays a strong game. I also dabble with Nemjet and Sjeng for interesting games when I want something outside the norm. One of the cooler modified chess games I've played in real life was Steve Jackson Games Knightmare Chess, which is loads of fun. I don't have a copy of Pocket Fritz or Fritz yet, but I've read interesting things about both, and I await Jenson's first-hand review. I keep myself somewhat busy with PocketChess on my Visor, and Erik "Possum" Stewart hooked me up with the shareware ChessGenius awhile back as well.

Oh yeah. Steve: bring it on!

I'm almost solidly convinced this Toshiba laptop is stuck in a power-conserving SpeedStep mode. It has a SpeedStep capable Celeron, but I installed Windows 2000 Professional on it back in January without the benefit of Toshiba's original driver disk. Its been nigh-impossible to get anything resembling reliable drivers from Toshiba's website(s), since they don't even seem to admit the existence of a Satellite Pro 4270, which is the model I have (actually, that's the "system unit", the official model number is "PS427C-3Q152", serial number "20012620J"). The closest I've found are "4260DVD", "4280XDVD", and "4280ZDVD" models on http://www.csd.toshiba.com/.

One Windows 2000 compliant Intel SpeedStep applet I downloaded from Toshiba's site installed fine, but gives an error on start up. Likely a conflict with the imperfect model matching.

This is annoying because all the diagnostic tools I've been able to run show me running at around 252 MHz to 286 MHz, but identify the chip as being a ~500 MHz Celeron.

I'm also running out of disk space, but that's my fault, and an entirely different problem.

More writing on the chapter done today, as Gabe heads off for a holiday in Hawaii. Enjoy Gabe!

Some mucking around with the new Plan 9 release as well. Ideas percolating, as always...

Realized once again that I really can't handle the load of mail squeak-dev pumps out and unsubscribed today after only being subscribed about a full day or so.

Sent resumes off to prospective employers in both the southwest of the US and Ireland, as well as a local-ish place.

Feeling the need to hack something different. Bored with current project list and lack of means. Fiddling with mobile phones.

After a brief email exchange, Mark Simonson has updated (for Windows) his good-looking, great-tasting, less-filling font, Anonymous. It now shows up in those "monospaced-fonts-only" limited font selection dialogs you get with some apps on Windows. Go download it and his other wonderful fonts today! Thanks again Mark!

Wow. Zooko actually reads my blog. Who knew? Hi Zooko!

I just lost half a month to execute on various plans.