Ah, fixed it again. I gotta put that into my X startup scripts, so I can quit having to do it each time I login. For those who have no clue what I'm talking about, it was the following command:

xmodmap -e "keycode 0x31 = Escape asciitilde"

Mapping the escape key to the place where the ` (backquote) normally is. (And the keycode for a backquote on the escape key, so that they're switched (sort of): keycode 0x09 = grave asciitilde) Yeah, but I keep the shift + keycode 0x31 as an asciitidle, cause it gets used a lot for cd-ing to home, and other nice activities.

And people, I swear by this: It greatly improves efficiency. AND, unlike switching your whole keyboard layout to Dvorak (which is also well worth the effort, BTW), simply remapping Escape is quite a breeze to get used to. Just imagine, your left hand doesn't have to go way off into Timbaktu every time you want to switch back from input mode to command mode. With escape in easy range of the left little finger, it's as easy to hit as backspace is on the right hand. No need to even move the rest of your fingers off the home row for a millisecond!

And really, unless you do a lot of bash programming -- specifically bash programming which involves sucking the entire standard out of commands into variables -- how often do you really need to hit that backquote? And for those rare occassions when you do, you can move your whole hand up into La La Land and press what was once known as the Escape key. Much better than having to do it each and every time your cat sneezes, or your screen refreshes. (Or whatever)

And people wonder why I have such a hard time when put in front of one of those hideously un-optimized keyboards it somebody else's cube. lol.

So anyway, I guess the point is: Argo still needs a little bit of training when it comes to such matters. :)

As I have more or less stated before, most of the fest in my area was focused on getting Argo's hardware to work to its full potential. In fact, I pretty much put off taking care of other nicities, such as getting the window manager keybindings, colors, behavior, mouse bindings, menus, etc. Instead, the urge to just Make It Work drove me onward.

At around 0700, after having spent most of the night focusing on convincing Mesa to do hardware accelleration, I simply became too tired, and crashed in Jaeger's guestroom. cvs checkouts, config options, installation, kernel patches, and it still wasn't working. I needed a break.

When I awoke a couple hours later and returned to the alcove, I went back online to search for more documentation, scan more web pages, read more howtos... anything that might point me to how to solve my problem. (Of course, I knew the obvious: When dealing with pre-alpha code, this is often the way things are.)

I came upon a web page which had nightly build rpms of the glx code. Only problem: It hadn't been updated since March. Therefore, I was hesitant. But eventually, I decided it was at least worth a shot. I installed the package, restarted my X server, tried running one of the opengl xscreensaver demos, and viola! Hardware accellerated 3d! Smooth and crisp as a wafflecone Sundae.

Seconds later, we were called to breakfast. French toast. Mmmm.

After breakfast, and returning to Argo's console, I did some more playing with the various demos. Out of curiosity, I decided to try to see just how much of my main cpu was being used for this stuff, since the real grunt work was (hopefully) all being farmed off to the graphics chip. Went into an Eterm, and typed top. I never got to see the output. Argo froze cold, right in her tracks.

Somewhat let down, but not exactly disappointed. I had been warned many, many times in the docs that this sort of thing could -- and probably would -- happen. So be it. Reboot. e2fsck.

Took a look around again. Wiped out a couple of duplicate libraries in different directories that I thought might have been contributing to the problem. Again, I ran another of the demos, one that had been the most CPU intensive when using software rendering, and this time, successfully ran top. It showed my processor as about 33% occupied. Not bad. No noticable lag.

I ran a few more, and again the system froze, seemingly at random. Must be a memory leak somewhere in the AGP DMA transfers, or something. The truth is I have no idea what the hell was causing it. In fact, even if somebody who did know what was causing the problem tried to tell me, I probably wouldn't understand the explanation.

I decided that a flakey driver was better than a driver that didn't work at all. Proof of concept, at least. While I was there, I also grabbed a copy of Blender 2.0 off the net, which I will do some experimenting with. See if it has any crashes using these drivers. If not, great! If so, I see in my future more tinkering with cvs, subscriptions to glx developer mailing lists, and all sorts of other fun in my future. I figure even if my skill is not at the level where I can fix things myself, I might at least be able to help the real coders in testing, reporting, and debugging stuff.

I also pulled down a current snapshot of the emu10k1 drivers (the SB Live chipset) from Creative's site, which I have not yet done anything with yet. Reading the page though, I was given hope. As my experience reflected, Mandrake was reported to currently be the easiest distro to use when it comes to detecting and getting running with SB Live cards right now. The drivers aren't even included in kernel 2.2, so they had to hack that support up themselves. And, according to the page, they did a damn good job of it. But, the version of the driver that shipped with Mandrake 7.0 had bugs in... Tada! The mixer settings!

I'll have to compile this newer stuff, and see if it fixes my problem.

I was also tempted to try and snag a copy of the q3 demo. One look at the file size, and I knew it wasn't happening. I had already been sucking Jaeger's bandwidth to the point of verging on being a public nuisance, and that thing woulda taken hours over dialup. Argo was a very hungry horse on this fest! :)

As I had trying to hack up glorious multimedia capabilities on Argo, meanwhile, Jageer had been doing his share of insanity. He patched his kernel for USB support, which was a success. With this, he was able to interface Ziyal with his Visor. When I left off, he had been working on trying to decipher its mysterious database format, so he could seamlessly update his changelogs without having to retype them. Festing fun for whe whole family!

When I realized it was afternoon, I decided it was time to regretfully leave the dialup line behind, and began preparing Argo and Dagobah for the return journey.

I had Tobias all loaded up in short order, and was ready to say my goodbyes. Then, I got an offer that was just too sweet to resist: The opportunity to play a starring role in a Motion Picture Extravaganza of Millennial Proportions! And not only that, but it was in the scifi genre! Who could turn down such a thing? To paraphrase Jaeger, guerilla film making at its worst is also guerilla film making at its best.

And so it was that the project got underway. Things being decided on a shot to shot basis. Me and Willy being fed our lines, usually going through one trial run, and then doing it while the camcorder was rolling for real. Cutting to exterior shots, featuring the inflatable rocket against the folded up ping pong table space background. (oops. Did I just give away one of our special effects trade secrets?) The alcove being the set for the interior of the starship.

Then it was outside to the exteriour shot of the planet, where the base had been sucked out of the ground. lol.

And back to the ship, where that priceless final scene was scripted on the fly. rotfl.

After film making was complete, the entire cast and crew were invited to attend the premiere party, where we got to see what all the wizards in post production had done to enhance the film. My, they can work wonders, can't they?

We were then treated to three classic episodes of "The Why? Files", which were... well... something on par with the production that had just been premiered. Hound and Scaly on the case. lol.

It was then that I took my leave. Must've been just before 1600.

Upon returning to the Louisville Compound, I unloaded all my hardware into the Lair, wherever there was space on the floor, and... proceeded to crash on the bed. Immediately.

I must have been out for two or three hours. I awoke, took one look around my room at the utter chaos which surrounded me, and felt like going right back into a coma. Went to the kitchen, ate some leftovers. I straightened things up just enough to get Dagobah and Argo functional again. This time, they're together, side by side, in the corner which Dagobah used to inhabit exclusively.

They look so cute there! And they're getting along so well now! Like peas and carrots.

(/me recalls recently reading in another web content provider's page something about computers breeding like... was it rabbits? Looking at Dagobah and Argo... I'm not sure I want to know what's going to be happening around here after I leave for work tomorrow.)

"Now you two behave yourselves! And remember, if it's gotta happen, ALWAYS use a proxy."

Ok, this is getting just a little too insane.

And now, here I am, up late again, when I should be in bed. Must... resist... urge... to stay up until 3am configuring drivers. Must! Resist!

Alright, I'm quittin this. As a final thought, I will say:

Despite all the strangeness that happened on this fest, at least one good thing happened. We got to meet a lot of cute women. ;)