" /> BP's Weblog: March 2001 Archives

« February 2001 | Main | April 2001 »

March 20, 2001

Looks like it's California for me!

Well, PARC offered me a job. The details are well within the lines of "satisfactory". And, the job still sounds great. So, to end the saga of "what will BP do when he graduates" - we find the answer: go to California.

I guess I'm going to follow the old "Go West Young Man, Go West" attitude. I'm excited about that. Now, it's just time to, you know, finish my academic requirements. I'll probably take up residence in some part of the Palo Alto/San Jose area about June. Expect geekdom.net to vanish for a few days/weeks while waiting for DSL or Cable modem to go up there.

More later. You'd think things would get easier right after Spring Break, not much harder. Ugh.

March 12, 2001

California is pretty this time of year....

So, I just got back from California, where I was interviewing for a position at Xerox PARC. And, boy, is it pretty out there! Man. If you haven't been, you should definately check out the SF/San Jose area. Maybe it was just a nice late-winter weekend, but, really, it was very nice.

With that said, I have to talk about my trip some. Interviewing at research institutions is great - they give you a little intro to what they're doing, run you through your paces on things they think you should be able to do, and then you get to talk about the interesting stuff they do. I've now interviewed at two different research labs (the other, MIT Lincoln Labs, in November), and both experiences were great. In this case, I was interviewing with the Advanced Systems Development group, a part of PARC that's responsible for finding interesting projects nearing completion (or, well, near enough to be moved on) within the full-scale research going on elsewhere in the labs, and build a hardened ready-for-market sort of project. Or, possibly, discover that a technology isn't good enough for deployment. In any case, they get to play with everyone else's toys, and see a conceptual semi-working research prototype through to the finished stage. Kind of like a bird's mother seeing the baby safely to the point of flying and self-sustainence. It looks like a lot of fun.

Anyway, here's hoping they find me a worthy member of their team. California's beautiful, their lab's beautiful, even the research is beautiful. Hrm. Sounds like a line from Cabaret - "In here, life is beautiful- the girls are beautiful- even the orchestra is beautiful!" Sorry. Theatre withdrawal. I hope there's an active community theatre somewhere near Palo Alto. Too much work makes Bryan a dull boy. Hrm. Anyway, back to PARC - the work looks challenging but fun... what more could I want for a first job out of the academic world?

March 03, 2001

Knowledge transfer in Computer Science, Grade F.

Computer Science as a field really needs to get organized. There is lots of research and information discovery that happens everyday, but that knowledge is not being transfered properly. Many of papers written are hard to understand and even harder to code. For example, there are several papers written on the 'Realistic Animation of Liquids', i.e. water, etc. The paper describes equations with incorrect mathmatical expressions, unmatched variable names, and overly simplified explainations. Other problems include neglecting to mention important equations and their permutations. Why should Computer Scientists accept this? As a contrast, the medical community will not even look at such a shody work, not even considerable for publishing. Incomplete details in their field means certain death, and no doctor wants to take that chance. Do you want your doctor to operate on you while he experiments on the proper technique? I think not. So why do we accept 'published papers' that only describes half of an algorithm? The researchers may have discovered the meaning of life, but by publishing an incomplete answer, they are getting people's hope up, and wasting other people's time.

March 02, 2001

Looking for some good writers!

Do you like the premise of this site? Do you think you have something interesting to contribute yourself? You, too, can be a contributor to geekdom.net! Just go to the left side of the screen, and select "Submit News". Looking for ideas? Check the forums! All submissions are welcome! Know someone who you think is really smart? Recommend the site, and suggest they contribute some of their dizzying array of knowledge back to the rest of us. That's what this site is all about. Don't delay!
Ed: This is an old question - the blog world has pretty much decided where to go since then.

Pain comes in waves....

So, the cold's not even entirely gone. Sheesh. A friend of mine suggested it might be a sinus infection following the cold which is hold me up. Frankly, I don't care, but I do want to stop being congested. Sheewiz.

This sure has been a long week. I've had demands all week long. Big problem set in Machine Learning. Exam in Astronomy. Demands from my residual project for Dr. Kimura. Grading the first set of labs for CS431, and holding offer hours, and a "response" to some theory papers that Professor Loui had us working on. And that was just Tuesday through Thursday. Ugh.

Anyway, heading into the weekend is nice. Next week really only has the uncertainty of the first exam in Machine Learning. No one really knows what to expect, so I'm not sure how nervous to be.

But, my travel to interview at PARC has been confirmed. This will be my second trip to the West Coast. I'm sort of glad I already interviewed at Microsoft, it sounds like Seattle had a little trouble with that earthquake yesterday. I haven't heard from any of my friends who work out there, which, unfortunately, is not unusual. I hope everyone's alright.

Put up the LinkExchange banners on the site. I did this years ago for B.P.'s Domain, the website, and it worked out really well. But, I was also in the first, oh, 50,000 sites to sign up for LinkExchange. The number for the account I got for geekdom is obscenely huge. Worked up a banner, in typical simplistic style - lots of animation. But, it's under 10k, and its done moving in 7 seconds. Hopefully it'll draw in some new visitors. If you're reading this after following a banner ad, chime in below, eh?

Random musings: technology that might allow "printing" of computers keeps popping up on Slashdot. I can't tell you how excited this makes me. If it works out, the result would be the ability to make custom computational hardware on the cheap. Really cheap. Like on the order of 10s of dollars for entire machines. Which more or less means to me we won't be calling them "machines" so much anymore. That full-fledged computers will become literally as common as electrical outlets. Why use a light switch when you can print out a custom control panel?

And, even more interesting, is what this kind of technology could do for the field I'm just about to enter. It seems as if the field has just begun to enter the era of the conventional computer everywhere. Several of the embedded proejects I've been involved with or seen working are targetting small form-factor, but essentially totally complete PCs for their systems. They're just cheap enough, that why would you hassle with a reduced system and require the programmer to mess with it? But, if we can print out new circuits to test them, CS might venture back to the EE world, studying VLSI and circuit design much more closely. Compilers might once again see a resurgence, as people work out how to build efficient multi-layer boards from high level language code. If prices drop so greatly, researchers will be able to get access to so much more CPU time. Put this together with the growing wireless and broadband and distributed networking trends - what do you see? You see some Sci-Fi futures coming true. Computational power everywhere. Enough CPU power in every room of your house to run a nice voice recognition/voice synthesis. Why not give up on switches, and switch over to Star-Trek style voice commands to control our appliances? Suddenly Big Brother is cheap and could be commonplace. We must focus on privacy and security as the barriers to truly ubiquetous computing continue to drop.

Hrm. I'd probably really better rescuplt the above into an editorial. Look for that soon.