Never one to follow the path of least resistance, I opted for a microATX system, which in principle sounds really neat: its compact, its light, its quiet, and it won't take up much space. However, in practise, what it means is the case is almost too small to fit all the stuff I want to put in it, and it takes forever to cram all those parts in there the right way without breaking them. One finds oneself staring at this tiny RS480M2 motherboard, wondering which of the components will take the force of a small pry bar needed to move the board into position. Once secured, I then set about attaching things that needed to be attached.
The array of manuals in front of me was impressive (yes I read the manuals, there's just too many tiny little obscurely labelled wires, for me not to). Each step required cross correlating the motherboard manual with the case manual or the cpu manual or the powersupply manual or something. Turns out that while there's four slots on the motherboard, you can't use three of them (like I had planned). You can use one, or two, or four; just not three. So one of my GB DIMMS was going to be warming the bench.
The first of many heart failures came when I went to plug the power in. Seems there are a couple of different kinds of connectors on power supplies and motherboards. My power supply had the 20pin kind, and my motherboard had the 24pin kind. Luckily someone was looking out for me, and the manual for the motherboard pointed out that the 20pin connector would fit in the right 20 pins of the 24pin connector, and that it would in fact work OK and not smoke $1,500 worth of parts.
After a quick test to calm myself and demonstrate that, yes, everything was indeed working and nothing was bursting in flames yet, I continued on my merry way; attaching USB cables, audio cables, Firewire cables, reset button cables, hard drive cables, etc. Compared with the initial assembly nightmares, installing the ATI FireGL V5000 card was kind of anti-climactic. Another test, and still everything was working. Amazing.
The next speedbump was the small note that Win2k and before could not boot from a SATA drive. Of course I had a SATA drive. Of course I did not have any install media for winXP. So I got the hard drive and DVD-R into the case and squished everything together, but I wasn't going to be installing an OS immediately. So I hooked up the monitor to it, and let it bake overnight.
The next day I started calling around to everyone I knew, looking for an install disk. Work would give me a registration code, I just needed the media. Wasn't having much luck though. Peter had one at CCC, but he was at home for the next couple of days trying to catch up on his homework. At the USGS and Gore, everyone was using Dells or someother brand name computer that came with its own restore disk. Joel was down in Phoenix for the week wrestling with the forest service, and Dan has disappeared off the face of the planet and his voice mail was already full. I was dreading a trip to Staples, but I was running out of options. Finally I convinced my friend to upload his media across his cable modem at a paltry 300kb. Ten hours later it was online, and I pulled it down in 50 minutes.
Now as old readers will remember, CDRs and I don't always get along. Once again, my new DVD-RW was out of commission (as it was currently tucked in the new case with the new machine), so either I had to pull everything apart again or I had to figure something else out. My laptop still had that CD-RW drive in it, and so I decided to dig up the new SW (that came with the DVD burner) and see if it would work in the laptop. Amazingly enough, it did; though asking it to record at 24X actually only got me about 10.8X. Still, 500MB goes quickly when you can pass the time watching movies on your Treo.
Called up work and got the key, entered it ... didn't work. Turns out the bulk licenses use a different media from the retail distribution. Arghh. Back to my friend Tim, who has ten keys as part of his developer pack from microsoft. He burnt me a new key and I was off and running. Droned through the install procedure for what seemed like forever, then it booted up and I was in. Time to load all kinds of drivers. Drivers for the motherboard, drivers for the ethernet, drivers for the firewire, drivers for the flash reader, drivers for the audio, drivers for the graphics.
Now, out of all the variables one has to choose from when building a PC from scratch, there were only three that I had specific requirements for: it had to be small (microATX), it had to be quiet (more or less), and it had to be able to drive the Apple Cinema 30. See the 2560x1600 dictates that the graphics card has to have a dual-link DVI interface, and that pretty much dictates that you have to have a PCI-Express card (unless you want the ATI X3-256 relic), so pretty much my choices were an ATI FireGL V5000 or a V7100. So I had a v5000 sitting in that pci-x slot (due to budget constraints), and I was in a hurry to get its drivers going so I could bask in the full glory of my hires display. The installer was kind of curious, because it offered the choice of installing the drivers off the CD, or installing from the internet. For some reason I chose the CD (thinking it was faster), and boy was that mistake. I got to experience a few minutes of joyous super resolution, and then after the next reboot I was in disaster land.
It started out ok: I got the BIOS screen, then the booting display, then my display shut off and the power light blinked three times. What does that mean? I did a quick google search, and the first entry there gave me another heart attack:
My experience was very similar. The upper half of my screen was very dim and the power light was flashing 3 blinks. I checked on Apple's web site, and it says that 3 blinks is the code for "backlight problems/failure". Indeed, my backlight had failed and sending it in for warranty service was the only solution. Even worse was the fact that my display was only 2 and a half months old.Arghh! My screen is broken already?
Just to be sure, I went out to the garage to get a spare monitor to hook up. Now there was already another monitor about fifteen feet away from this computer, but the problem is, its been there for a while; and thus is non-mobile due to the large quantity of stuff that has collected around it, in front of it, and on top of it. Going outside, and trudging through the mud, was a far easier task, then extricating the other monitor from where it had been buried. After pluging the other monitor in, it also wouldn't come up, so perhaps it wasn't the display. I rebooted and the computer decided to display on the new monitor at 1024x768, I started fiddling around with settings, occasionally getting the Apple display to work at resolutions of 1280x800 or lower, but I wasn't having any luck getting back to hires. I finally found a note on the apple care sight all about power light flashes. Turns out there's different ones. There's short-short-short, short-long-short, and short-short-long. Short-short-long is very bad; but short-short-short is just the monitor complaining about an invalid signal or unsupported frequency. So my monitor was ok, turning the suspicion onto the display card.
After fiddling with connectors & settings for a while, I decided to update the drivers from the ATI site. Turned out the CD was rather out of date. Replaced version 6.something with a much better 8.something. Rebooted again, swapped the DVI positions, and magically it worked. I wish ATI would do a better job QC'ing their drivers at hires. This isn't the first time their drivers have trashed my display. When upgrading my drivers on an older Radeon card, I had switched to 1920x1080 only to have the graphics card completely freak out and start displaying random blocks of colors everywhere as soon as I rebooted. It was only because I was able to walk through switching the video settings completely in the dark with just the keyboard (by tracing my steps on another computer) that I was able to switch the resolution back and then regress the drivers back to a previous version and have a usable setup once again.
Thanks to multiple panic attacks, I had managed to pull one of my floating ribs out of joint (which occasionally happens when I get over anxious), but hoping that the worst was over, I decided to soldier on. Too bad the worst wasn't over.
While downloading random programs necessary for operation: acrobat, netscape, gvim, mysql, cygwin, terraterm, cute-ftp, quicktime, bittornado, bzflag, etc. I also hooked up my 300GB storage drive in its USB case to move some things around. Just to freak me out again: the drive didn't show up. It felt like the harddrive was spinning, but the activity light on the case was stuck on, and the system wasn't recognizing anything. Frighteningly enough, this 300GB drive contains a fair amount of important stuff, and isn't really backed up. It was starting to look like it might never be. I took a moment to collect my thoughts, and after I started breathing again, I pulled the drive out of that case and stuck it in another USB case I had that wasn't quite in as good as shape (at least cosmetically). Thankfully the drive showed right up, and worked just fine despite the duct tape holding the case together; so I was able to install some more stuff from it and set it up as a secondary drive for this system.
Having cheated death three times in a row in one day, I decided to pack it in, and leave other battles (like my boot drive showing up as letter F:) for another day.
The Hubble continues to put out facinating pictures, while celebrating its 15th year in space, much to the consternation of the higher ups in the NASA administration that want to kill it. Fifteen years is pretty long for a satellite, though it could probably go another five with regular service. The funny thing is, that if the space shuttle hadn't blown up; the Hubble might have gotten one more fueling up, but it was destined for the scrap heap soon after that anyways. The James Webb Space Telescope is currently scheduled for deployment in 2009 with a lightweight beryllium mirror six times larger than the Hubble. Since part of its focus is more infrared observations, they're going to park this one way out at the second Lagrange point (about a million miles off). Guess the shuttle isn't going to be used to make service calls to this one.
Anyway, to top it all off, NASA also messed up the DART mission (Demonstration for Autonomous Rendezvous Technology), which is especially embarrassing, because here we're not doing any new investigation, we're just playing catch-up with the Russians who can automatically dock spacecraft together on a regular basis. After a 24 hour flight out to a derilict satellite (MUBLCOM), DART was supposed to maneuver within five meters of the other object on its own. While still 100 meters away, the flight was scrubbed when officials discovered the craft was already out of gas, but that wasn't the end of the bad news. Either because they decided to go home early (having nothing to do), or maybe because they couldn't steer any more, the DART craft apparently wandered over on its own and bumped the other satellite.
To you and me, this is just one more minor screwup in a long line of screwups (space can be like that), but to the people who have worked on this thing for the last several years, this is certainly a bigger disaster. We don't think much of the people who staked their professional reputations on the design and engineering goals of the mission, fighting for the funds year after year, gathering the teams together to build and assemble the satellite, and then pack it all together in the rocket, ship it off and see it launched. As someone who has put years into various projects that have occasionally crashed and burned, I feel for those guys.
It was never actually supposed to be a full time job, and its been a pretty good run actually; luckily I've found another project to jump into the middle of, which is also exciting. My wife is especially glad, considering we have another rug-rat on the way. I think she's gotten used to it in any case (I changed jobs right around the time each of the last two boys were born), or perhaps she understands that my identity is not tied to my job; but whatever the reason, it is a nice thing to have a wife that is calm as I weather this transition.
It does help somewhat that the economy is back, and Java hype has long since crested so that people can go back to using other tools like perl and C++ where they are appropriate. Its not that I can't write Java, it's just that the kinds of thing that Java is good for aren't that appearing to me. Give me: bare to the metal, runs forever, resource use is an issue, all the performance we can get, C++ anyday. Its a bonus that the new project has lots of 3D graphics, and we're required to do our testing in a MMOG. Also, the VPE is very cool (that's him dressed up in fatigues for a publicity shot for an article in SFWeekly), and had no problem having me start weeks before I'm going to disappear off the planet, tending to my wife in the hospital.
Still, the caliber of the people on games grid was phenomenal. I've never worked on a project before with a group of programmers that understood technology, software engineering in general, and C++ specifically; to the same degree. I haven't always agreed with the designs and decisions of the others, but I've always respected their choices.
The group also sports an impressive background, having known each other through times at Apple, Taligent, Palm, Be and Netscape. Red even reached a certain level of infamy as the manager of the billion dollar DTS (Apple Developer Technical Support), a group of guys who went on to create (or help create) several billion dollar business and products, such as Pierre starting EBay (with another alumnus, PT Helme), Rob Tsuk helping launch the iPod, Steve Maller a million dollar tech gun-for-hire, and Jor Bratko heading up developer relations for Palm, and Dean Yu heading up Yahoo Gaming. I just hope his luck continues to hold, as Gordon was one of his employees, and it would be nice if Games Grid became yet another billion dollar business (and made me rich in the process).
So a fond farewell to Max, Gordon, Chris, Red, Buzz, and the rest of the gang as I head out (back to my little cubby in Flagstaff), and on to bigger and better things. I wish GamesGrid the best of luck, and will be looking for that check in the mail.
There was even a handy locator, so I did a quick search for Flagstaff ... and turned up zip. No icecream for me today, I guess.
But all is not lost. I might not be able to get any frozen fat out of the deal, but that doesn't mean I can't get something. I know lots of people that do live in an area big enough to have a B&J near by, so this gives me a good excuse to call them and catch up. I'm horrible at networking anyways, so this is a limited time opportunity to quickly say hi and pass on the tip. Plus then, they all owe me a tip sometime.
Anyways, if you live in a big city--go get an icecream, and if you don't--call a friend that does.
Now if you're a founder, and you've been talking to the investors yourself, then this probably isn't really a big surprise. You were hopeful, and maybe a little to optimistic, but really in the back of your brain you were thinking that those guys with the big checkbooks weren't really going to move anytime soon. However, if you are one of the ones back at the office, hunkered down at a computer screen, working on the latest changes to add support for call data records from yet another carrier (because the CEO can't make up his mind which carrier to use), and you are hoping that funding will come in sometime before summer so you can get a real paycheck, and maybe get the air-conditioning fixed in your car; then I regret to inform you that its not going to happen.
The earliest you're going to see any action on the funding at this point is September, or maybe October. But that's like six months away, what the heck is going on? Well, unlike you and me, venture capitalists (and all their minions), have real lives, and do things like go on vacation. So, even though they still have meetings for the next month or so (just to put up a good show for their limited partners), they're going to be sitting in those meetings trying to figure out if they should vacation out in Martha's Vinyard this summer, or maybe try out that villa in Italy their friend told them about. And because any important decision requires all the partners to meet and discuss it; the whole summer gets wiped out by one vacation after another. You're just sunk at this point.
I don't mind so much when a startup I'm working for is not going to be able to make good on its promises right away, but its really annoying to be planning on one thing, when all along someone else knew it wasn't going to happen. Its not a problem that's unique to VCs or startups though, I run into it in business all the time, like when I try to buy a house. After having the loan manager go on vacation (usually for two weeks) right in the middle of my mortgage application on three different purchases, I finally added it as a question on my qualifying list for mortgage companies:
I was so annoyed at this latest example of stupidity, that I wrote a scathing article all about common household items that you could easily bring through airport security, then combine to create chemicals far more dangerous than butane. (Those rocks are calcium carbide, if you want to get an idea of the kinds of things I was thinking.) I sort of knew while I was writing it though, that nothing good could come of it. The department of homeland security would still continue to put out idiotic rules that don't help airline safety at all, and I'd probably alienate some readers who thought I was passing secrets to the bad guys.
There's also the bigger picture. Right now I'm working for a company that provides training technology to various groups, including the military. While geeks like sharing ideas as widely as possible (even threats), the military doesn't like anybody giving any ideas out at all. And, they have no sense of humor about these things either. So you don't really want to piss these people off (as a friend of a friend in Iraq found out after getting on the wrong side of the base commander by posting an insulting rant about checkpoint security on his blog). So not only would my original posting have been irresponsible, but it also could have caused problems for the people I'm working for, and they're pretty nice guys.
While reacting, and posting my rant would have been satisfying in the short term; understanding that what I write effects others, and that we have to act like we are members of a civilization (at least some of the time), is a far better lesson for me to take away.
Maybe its because the car is 15 years old, or maybe its just from the abuse I give it, but my car needs repairs on a pretty regular basis. This time I managed to accumulate a number of problem, which gave the mechanic quite a hard time.
My previous regular mechanic had retired a while back, and the last garage that worked on it hadn't done that good a job, so I was still hunting around for a good one. Based on a friend's recommendation, I ended up over at Larry Hobbs' place. Poor him, he was in for some trouble.
Working on a strange system is always a challenge, but in this case, that was the least of his troubles. Instead of chasing one fault, he was chasing three (and all three faults were intermittent faults as well-the worst kind). It got so bad that at one point he called me up and told me that my car was cursed. He didn't want to look at the thing any more, so I better come get it, or he was going to shoot it. My car can affect people like that.
I didn't drive it much for a while, the weather wasn't that good and a laser isn't really a snow car; so it mostly sat around for a few months. I took it to the airport for travelling (which works fine, since it sits around for several days before I try to start it again), and I got back home just fine, but of course I left the door ajar and then ran down the batteries. So, eventually I hooked up the charger, fixed things up, and took it for a spin.
At first it worked pretty well. Maybe the charge helped, or maybe I was imagining things, but it was going ok until I stopped over at my parent's house, and parked in the driveway. Then it wouldn't start, at all. Some how, after rolling backwards down the driveway, turning around, and stuttering down the hill under the power of the starter & gravity, I somehow annoyed the electronics enough that it started for me and I made it home. Not willing to let a sleeping dog lie, I again took it out the following week and at every stop, it wouldn't restart right away, and only got me to my next stop after much cajoling & abuse.
This turned out to be real progress, because as any software engineer knows, an intermittent problem is much harder to find than a consistent one. So I waited a little while longer, then called up the mechanic to let him know the good news. After some convincing, he finally agreed to take another look.
After some more poking and prodding, he had finally discovered three separate problems which he believed were responsible for the trouble. Unfortunately, the solutions are turning out to be even a bigger problem.
So I now have a new mass flow sensor, fuel pump, and ignition security relay; the sensor malfunction was leading to serious confusion on the part of the injector routines and was causing hesitation, while the low pressure from the old fuel pump was starving the engine at higher RPMs. Larry was hoping the relay replacement would fix the starting problem, but no such luck.
Unfortunately its pretty hard to get parts for old cars, which is a shame. First it was trouble getting parts for my 77 chevy, now its a 1990 thats entering the twilight zone between mainstream and vintage. Once I find the electrical manuals, we're going to see if we can hotwire this thing, or else it may be time to start working on that electric conversion.
While I always hated videotape (or any tape for that matter), I've collected movies in just about every other format that came out. It all started with the RCA CED format which featured a vinyl grooved platter which was read with a phonograph like needle. I eventually moved on to laserdisks (spending a small fortune to amass several hundred), and I have now settled in to a more laid back pace, collecting DVDs.
Besides buying new movies, I sometimes try to get a DVD of something I have on laserdisk or CED, as I imagine someday not being able to get spare parts or players for those formats and ending up with unplayable disks. Also I feel a bit better about my kids handling a DVD disk than one of the larger twelve inch platters that the other formats use. DVD's certainly are more portable as well. One can only imagine the stairs one would get if one tried to watch a laserdisc movie on an airline flight (never mind the size of the laptop required to have an on-board laserdisk player).
Still, all the old movies I have on Laser or CED aren't necessarily worth getting again. Some of those movies were bad 15 years ago (and haven't improved with age), and others are just kind of odd.
The last unicorn is one such movie. I thought I liked this movie when I first got it, but after watching a copy I picked up last week from the library, I'm really scratching my head about it.
The story is pretty good: an ancient unicorn overhears hunters talking about how she's the last unicorn left, and goes off in search of the others (since Unicorns are immortal and they have to be somewhere). Along the way she deals with country-folk who can no longer recognize her for what she is, and a witch who does recognize her and captures her briefly for a travelling freak show. As she learns clues about the other's fate, she gets near to the land of the red bull who has supposedly herded up all the other unicorns and to protect her, a friendly magician turns her into a human.
If you make it this far through the movie, things get even more difficult. Through to the end (no I'm not going to tell you how it ends) the movie feels like half the dialog was edited out, or that maybe this movie was poorly translated from another language. The animation was actually done in Japan, though the movie never came out there. It was in theaters in the US back in 1981, and since then has wandered out on video in a number of formats. This is probably one of those stories where the original book is much better that the film, or maybe I'm just jaded after 20 years of hollywood endings where at the last minute, wesley saves the ship, and everyone lives happily ever after. Perhaps that's not how this story goes, but the way things conclude in the movie is jarring and a little depressing.
The entire script with pictures from the movie.
Due to a random fluke, I was over on the other side of town with some minutes to kill; so I popped into the nearby Salvation Army to see if there was anything interesting in the bins. Nothing grabbed my attention, so I wandered into the back, where behind a row of pressboard shelves and steel office furnature from the 1950s, I discovered a row of bikes, from adult size all the way down to just-starting-out size. The price seemed right too: $5. So I grabbed one from the end, and checked out, very pleased with myself.
Max was pleased as well, though there was still some work to be done. None of the bikes had training wheels, so something would have to be arranged. Also, on closer inspection, the back tire had been worn clean through and the innertube was flat. So there was some fixing to be done.
After some misdirection, looking for a wheel size that does not exist (how I got 10" in my head instead of 12" I'll never know), I had a new tire and a patch kit. Back in the old days, patching a tire took some real skill. I never had to use the 'melt-it-on' style, but we still had to cut out our patches from some square stock, prep the area, apply the glue, and press the whole thing together until it set. Now the entire kit is a couple plastic bars for prying off the tire and a sheet of stickers. Kind of takes the mystery out when a three year old can do it.
So with the tire fixed, it was time to think about some outriggers. Once again the internet turns out to be pretty worthless, as shipping for a set of wheels is more than the wheels themselves. Guess its all that steel and bulkiness. The local bikeshops weren't much help either, as new bikes generally come with a set. Special ordering something for $20 that usually comes for free with a $40 bike didn't sound very good. Time to come up with Plan C.
A few years ago, my only option would be to scrounge around, asking other parents for leftovers; but now thanks to hundreds of hours of professional training at the local community college, I had a little knowledge of metal fabrication. And as we all know, a little knowledge is a dangerous thing.
After picking up some wheels at the local hardware store, I set off to the metalshop to concoct my solution.
It started off pretty straightforward. The ironworker was able to turn a four inch wide strip of quarter inch plate into two two inch wide strips with one quick chomp, then put some holes in the ends with another chomp (of the punch end), and then after moving the press brake into place, I had put two 90° bends into each piece. So far, so good.
The next part to do was the axle for the wheel. The plan was to find some half inch bar stock, weld one end to the bent strip, and drill a hole for a cotter pin in the other end. It sounded like a good plan except for the fact that there was no half inch bar stock in the metal pile (school stockrooms are often annoying that way).
Well, there was some 5/8 stock and a full machine shop. A lathe is especially good at making a large round thing into a smaller round thing, so onto the lathe. One has to be especially careful with school machines when following the readings on the dials. One would think that turning the crank in 0.005" would reduce the diameter 0.01", but its just as like to reduce it 0.015" on the machine thanks to the hundreds of impatient highscool students who have come before. Luckily I've learned never to take anything for granted, so I hit my target just right. With a final polish pass using some 120 grit sandpaper, I had a nice & shiny axle and I was just about ready.
The last hard step was putting a hole in the end for the cotter pin. Here the taper in the bar stock (I only milled down the end for the wheel) helped, as I was able to fit the rod through the plate, add a washer on the back, the wheel, and then a washer in the front, and measure where the pin hole would need to be. Unfortunately, drilling a small hole in the side of a hard round object is rather difficult, and the drill kept wanting to swerve off to the side, but after some adjustments and a couple of attempts, we managed to get a hole drilled (though not in the center exactly) and the fine work was all done. Taking all the pieces home, I finished assembly with my own MIG 200 welder which did a nice job joining the two thick pieces
So after putting it altogether, we took it for a spin. Everything seems to work great at this point, though the outriggers are so strong and large (I was really aiming at building a trike for starters), that occasionally Max rides over a rut and the back tire starts free-wheeling, but I'm sure he'll figure it out. The bars should be strong enough that his little brother could stand on them and ride behind once Max gets his peddling strength up. And I'm sure someday down the road when he no longer needs training wheels, those wheels will see use in some other project.
Was it worth it? If you just look at the costs of the parts and time vs the cost of a new bike, probably not. But there's non-tangibles that balance it out. For one, Max got to learn that getting something doesn't always mean going to the store and buying it. Sometimes to get what you want, you have to make it, and that takes time and effort. I also got a good review on the iron-worker, the lathe and the mill; as well as having something else I was able to weld with my welder, which was all good. To paraphrase "throw mama from the train": a metal fabricator fabricates–always.