R Allan Baruz: Computationally Minded

Sun, 13 Jul 2003 01:10:19 GMT

Scott Robert Ladd seems to have written a new book on parallel programming for Springer Verlag that I can’t find anywhere, so I’m assuming he just turned it in. I’m surprised he turned it in to Springer Verlag; he’s quite readable, and from my recollections of them, Springer Verlag books tend to be a bit dry and academic, with thin and reedy fonts that are a drag on the eyes. Of course, I could be wrong, but I don’t have any Springer Verlags in my library.

Apple WWDC Keynote

Sat, 12 Jul 2003 14:47:56 GMT

Finally watched this. While I am shaking with desire, let me note some of the grace notes that seem to have been overlooked in the bigger feature announcements that had the audience going ooh and aah. Labels have returned (about 14:00). Much as I liked Unsanity's implementation, I don't think it ever allowed searching for labeled items. If Apple brings this back from its OS 9 graveyard, there may be life for other OS 9 features. Postscript to PDF on the fly (about 36:00). Finally all those academic papers I downloaded from the ACM and IEEE libraries are readable without having to resort to Ghostscript. Waiting for the auto-negotiate of the iChat with Parisian addresse Jean-Marie Hullot, creator of InterfaceBuilder (45:30): “It takes a little while longer to negotiate with France.” Microsoft jab (51:00). Incremental compile, bringing XCode (the new IDE) closer to the Smalltalk ideal. Java has this in Websphere, which was built on Smalltalk technology, and perhaps Eclipse, which was dervied from the Websphere model. Xcode looks like it is starting to become a much more useful platform for Extreme Programming.

Thu, 01 May 2003 03:55:40 GMT

CODeDOC. Computers. And aht, dahling. How can it not go on these pages?

Sat, 26 Apr 2003 03:45:28 GMT

Both Mr Wiseman and Mr Robertson mention Alan Kay’s talk at the O Reilly Emerging Tech. J J had the link to the text, though.

Wed, 23 Apr 2003 04:00:30 GMT

Tuning in Smalltalk is the same as in any language. you either:

do expensive things less often, or
do less expensive things.

—Aner and Beck.
Lazy Optimization
“Patterns of Efficient Smalltalk Programming”
Pattern Languages of Program Design 2

Tue, 22 Apr 2003 18:19:46 GMT

Scapy, a Python-based packet... Swiss army knife, I suppose. By way of sweetcode.

Tue, 22 Apr 2003 03:54:38 GMT

Here’s something sad. Well, not sad, but certainly disheartening, at least to me. This is an example of a book that used to be prevalent in the early to late eighties. Books like this, and A. K. Dewdney’s column in Scientific American, taught children, enthusiasts, and amateurs not to fear computers, but to actively engage with them, as sort of partners in thinking or imagining. Nowadays, how many of these kinds of books do you find? I cannot think of anything right now that would fit this niche. Most books on computers are geared either to the professional or the consumer, that is, gamers and end-users. The Idiot’s and Dummies imprints are for professionals thrown into a field in which they feel out of place. There are no mass market paperback books for computer enthusiasts. There are no perfect-bound trade paperback books under twenty dollars for the computer enthusiast either. If a child were to look up and say, “Daddy, Mommy, computers are interesting. How do I find out more about them?”—what would you do? As the power of our personal computers has gone up up up, have we seen a corresponding rise in amateur literature? It is painfully obvious that it has not. While there have been individual projects that have focused on the experience of learning, there have been fewer and fewer books published for the enthusiast.

Sat, 19 Apr 2003 01:30:28 GMT

At GameDev:

Monday, April 14, 2003 Galaxy Dynamics Computer Simulation Posted by: Anonymouse at 18:50 The paper over at OSNews considers a mathematical model of the behavior of an assembly of N stars. The 'Kepler' Windows demo application based of this model enables to perform real-time simulation of star clusters dynamics for N~=2500. The paper also estimates the efficiency of the IPP application and provides an example of C-code with the IPP functions calls. Computer-simulated images of the spiral galaxy forming process, as well as the real galaxies photos, are presented.

Emphases mine. Heh. Considering getting a cluster of 2500 stars and the distances between them and the motions relative to each other in keeping it in a 1152x768 pixel screen, I could probably do a real-time simulation of 2500 stars by hand. What I want to do is get a faster-than-life simulation, say, a million years a second or so? :.)

Mon, 14 Apr 2003 04:16:22 GMT

Here’s an interesting tidbit: Peter van der Linden, who wrote one of my favorite C books (Expert C Programming: Deep C Secrets) has apparently left the Sun compiler group for Apple. He recently posted on Usenet (comp.sys.mac.programmer.misc) looking for an intern with some very high qualifications. I was skeptical of the e-mail (a Yahoo account) but the Apple employees seem to be backing him up. It looks to me from the job description as if they want someone to do automated testing frameworks using Java, perhaps JUnit—no, “internal Apple automated GUI test tools,” and Applescript and shell scripting, also C/C++. Java would make sense if it were PvdL, who did some work on Java. Hm. Automated test suites, and they’re expecting interns to handle the test plan? Pfft. Message-ID: <8f515400.0303270555.3662a9e9@posting.google.com> Quality automated testing is so hard to find. Oh, wait, no it isn’t—here we are! It’s just rather expensive.

Supershapes!

Thu, 10 Apr 2003 03:15:36 GMT

By way of Slashdot. Seriously. Here are more details. Here are some formulæ. Here’s a Java app implementation allowing you to tweak parameters to see what it can do.

Fri, 28 Mar 2003 02:47:17 GMT

It was written in C, which was a great help, except in those instances in which it wasn’t.

Sat, 15 Mar 2003 04:01:59 GMT

James Robertson of Cincom, the Smalltalk people, points out the results of a benchmarking study (yoinck!) between messaging services (sockets, CORBA, XML-RPC, and SOAP-RPC). BEEP not mentioned. [by way of Jay!]

Sat, 15 Mar 2003 03:27:23 GMT

Design decisions and best practices in application protocol design and rationale for BEEP, a protocol. [by way of The Fishbowl]

Sat, 15 Mar 2003 02:54:39 GMT

Doug Baskins tries to describe the Judy structure in ten minutes. Judy arrays (or Judy trees) are associative arrays/maps that are designed and optimized for modern memory architectures. Haven’t tried them yet, but they look to be of interest. [by way of sweetcode.org]

Sat, 15 Mar 2003 02:28:03 GMT

Graphical plotting programs for Macintosh by way of Slashdot (Apple). I can’t believe I missed something so relevant, um, to stuff.

Mon, 10 Mar 2003 03:46:04 GMT

Things to avoid...
How to write unmaintainable code

Mon, 10 Mar 2003 03:19:23 GMT

As a rule of thumb: A design is what the designer has when time and money have run out.

—James Poole
Chief Architect for Disneyland
Topanga, California

Tue, 04 Mar 2003 04:22:57 GMT

When and when not to use Numerical Recipes in C.

Mon, 03 Mar 2003 03:08:43 GMT

Multiplayer Design Patterns

Vectorized Mersenne Twister

Sun, 02 Mar 2003 17:42:56 GMT

Earlier today, I was thinking to myself, I could teach myself the Altivec instruction set by attempting to optimize the Mersenne Twister pseudoRNG (with its reference implementation in C) with it. That would be a cool project. Then lazin—er, sanity—prevailed, and I used the Google search engine (lawyers, please note unlitigable language!) to find a little message thread on an Altivec forum, wherein one of the mailing list messages responded with an Altivec-optimized MT PRNG coded by Doug Clarke as “a good exercise to keep [his] mind going.” Yahumina? Whuzifula? Do not worry, Darrin Cardini, I, too, feel like a deadbeat. While I was searching for that, I came across this Wiki for Mac game developers. I did not realize that idevgames.com had a Wiki FAQ.

Mon, 17 Feb 2003 01:12:07 GMT

Dave Hyatt has posted the table of contents to the beginnings of what will become the Safari WebCore documentation.

Sun, 16 Feb 2003 19:19:10 GMT

Gotchas when porting UNIX software to OS X by way of Stepwise. Ah ha, so that’s why X went wrong when I did Y. If only I had known when trying Z... Sigh.

Thu, 13 Feb 2003 02:27:55 GMT

What I should have done instead of saying, “I’m sorry, I’m afraid I’m not too familiar with your blog” all night long;:

curl 'http://babb2003.tripod.com/' |\
sed -e 's/<TR>.*<a href="\([^"]*\)">\([^<]*\).*/curl -L "1" > "\2.html"/g' |\
egrep "^curl" |\
sh

open -a /Applications/Navigator.app/ *.html

Browse at leisure, if you have any. It works best if you set the Chimerical Navigator to open up new pages in new tabs, in the background. It doesn’t work with framed pages, but who wants to associate with people who use those? Gotta love Mac OS X!

Sun, 09 Feb 2003 07:17:52 GMT

PragDave going onto the aggregator. [By way of the Accordian Guy, possibly the happiest Geek in Canada, if not on Earth.]

Fri, 07 Feb 2003 02:59:37 GMT

What was I saying? In the early stages of writing this script, I had declared an array in module space, which meant that any thread in the process could alter it. I had thought I had moved all my variables back into local storage on the stack, but after commenting every other call, I noticed that the string was changing in different places, which nixed my idea that one of the routines was overwriting it, and only the other threads could have been to blame. That seemed to be the error, as after I moved the variable back into a routine, I could no longer reproduce the error. In my defense, I usually work with process-based scripts. Ah well.