Home

Search Posts:

Archives

Login

January 2014

S M T W H F S
1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30 31

Recently I saw, again, the final round of a US Open golf tournament at Pinehurst. I'd seen the last one at that venue in 1999 as well, and my first trip to Pinehurst was for a PGA Tour tournament in the early 90s. I've also seen the GGO (now the GGCC) probably half a dozen times at Forest Oaks in Greensboro.

Golf - especially as a spectator sport - is one of those mysterious things that many people (including almost all females) just don't "get." "Isn't it boring to watch a bunch of guys hitting tiny balls with a stick?"

I don't know. Is it?

I like golf a lot. I don't suppose I like it because it's fast paced or action packed, but on a certain level the excitement is still there. I don't normally watch golf on TV (except for the majors, occasionally), but there's something compelling about seeing, in person, a group of people who are undisputedly the absolute best at what they do, even if it is something like hitting a tiny white ball around.

This year's Open was pretty exciting (as far as golf tournaments go) as Campbell, who had posted solid but unimpressive scores for the first 3 rounds, kept rising up the leaderboard by simply sticking near par. When I first saw him under par on the front 9, I had a feeling that he was going to win the whole thing - everybody else had been sliding, and sliding hard. Being the leader going into day 4 of the US Open doesn't necessarily mean anything, as Retief Goosen demonstrated - the leader and odds on favor to win as of his tee off on round 4, Goosen sank 11 strokes in the final round to drop off the leaderboard entirely by the end of the day.

Despite not winning, Tiger Woods was the guy who made this tournament interesting. While Campbell won by just playing solid golf all day long, Tiger's ups and downs had the crowd enthralled. Opening with 2 bogeys off the start (I saw both of them), he looked to be backsliding rapidly going in - and he was sitting about 7 strokes off the lead at one point. I was ready to write him off then, but he made several birdies to get back under par for the day and make it a great finish. Campbell walked up to the 18th hole with a 3 stroke lead over Woods, who was sitting in the clubhouse, and ended up bogeying the hole to win by 2 strokes. At that point you have to look back at Tiger's opening two bogeys, or his dreadful 3 put on 17 from 10 feet, and you start to realize just how close this thing really was.

All in all, it was a great experience, and I don't know that I could've asked for a better way to spend Father's Day with my dad and uncle. My Grandfather unfortunately cancelled on us the previous night, and I can't really tell you how disappointing that was - golf is one of those few hobbies that the men in my family share, and to see my Grandfather ducking out of something like this on Father's Day makes me really worry about his mental health and wellbeing. It really was a disturbing shadow hanging over an otherwise wonderful event, and I just wish there's something I could do to help him.

I've added another Dell PowerEdge system to my arsenal. You just can't beat these things.

After thinking (mistakenly) that I had fixed my old machine's lockup problem (or at least had worked around it by underclocking and using a generic kernel), it came back to haunt me with more lockups the following week. Instead of dicking around any longer trying to track down the hardware failure, I bought a Dell PowerEdge 420 for $330 shipped.

Now that Dell is fully operational, and is running everything the old system did.

I had toyed with the notion of migrating to FreeBSD, but when I woke up this morning to find my Athlon had locked up yet again - and that it wouldn't even *post* anymore - my hand was forced. I had received the Dell on Tuesday and had thrown a basic Debian install on it just to make sure everything was working, so this morning I pulled my old system, threw the old drive in the Dell, configured networking, and went to work.

Since it was a slow day at work, I was able to set up everything in short order. The Linux (and Unix) directory structure makes migration from one system to another truely trivial. The Debian package system makes installing all the software I needed equally trivial. All told, the migration cost me maybe 2 hours of actual work, and for most intents and purposes the system picked up exactly where the old one left off.

So cheap, so easy. Dell + Linux - you just can't beat it.

I was talking with Louie the other day and he observed that, due to huge rises in productivity, a good chunk of the population does virtually no "work" - they are effectively employed solely to be consumers.

When speaking of the United States, we have plenty to cover the "basic" needs of everybody. Food, shelter, clothing... we have no lack of resources or manpower to create these things. There are of course people slipping through the cracks, but it's not for a lack of resources (instead it's due to a societary misallocation of them).

As capitalism chugs on, the things we actually need to live become more and more trivial to produce, and there are fewer jobs involved in their production. Of course, the rest of us have to do something...

I find myself asking just what exactly it is that I do. At its most basic level, I lend support to help a business operate.

But let's follow through on that chain. What does my company do? They aid in software and hardware UI design and usability research for larger clients.

But what do the larger clients do? They take that research and create gizmos and gadgets and sell them to people.

What do the gizmos and gadgets do? They (in some cases) help people work more efficiently. In other cases they provide diversions, comfort, or entertainment.

There are essentially two potential outcomes, if you follow the chain:

- I lend support to a process that ultimately creates products that lend support to other processes
- I lend support to a process that ultimately creates products that results in trinkets and diversions for entertainment value

Of those two outcomes, I find the first especially interesting. It's so incredibly cyclical - the products we create may shave seconds off of a person's workload, and that person in turn may be working on products that shave seconds off of another person's workload.

It seems that a good chunk of the population works only to make other people's work easier. When you follow that line of thinking to the natural conclusion and "close the loop," it starts to seem pretty ridiculous - my job is to make Bob's job easier, and Bob's job is to make my job easier.

So how much of the economy is really smoke and mirrors, and why are the people actually producing tangible results (the guy down at Burger King making you a sandwich, for example) valued so much less than those of us who are shuffling work back and forth between each other?

I can't help but think at some point it's all going to collapse.

I did 3 things at once, some combination of which appears to have stabalized this system (knock on wood, I guess): downgraded to Testing, installed Debian's generic i386 kernel instead of the Athlon optimized one, and disabled the second NIC in the system.

Generally when troubleshooting you want to minimize the number of factors that change at any given step of the process so that you can isolate the failure point. However, given the potentially long timeframe between me taking an action and being able to verify whether it had the desired result, and given that I really need the system to be working, I just took the "shotgun" approach - eliminate as many failure points as possible in an effort to restore minimal functionality.

Something is really messed up with this system. Debian Sarge is frozen, and ever since the last update I've been locking up like crazy.

It could be Debian, it could be hardware (this stuff is cheapo, who knows?) I'm trying to DOWNgrade to Testing from Unstable (done by pinning Testing to a priority over 1k, if you're interested) so god only knows what'll happen next.

The blog is backed up, so is my email and such, so... well... we'll see. I may disappear for a while, though.

Futurama was the best TV show ever made.

Period.

It astounds me how some of the most creative, humorous, touching, and simply brilliant shows have failed so miserably. Futurama continued in that tradition, perhaps failing because it was simply too smart for the average viewer.

I just finished watching the episode "Luck of the Fryrish" again, and as it always does it brought a tear to my eye. I have never seen a better 22 minutes of television, and since Futurama was cancelled a couple of years ago I doubt I will ever see anything like it again.

The Simpsons died its creative death years ago, but without its accidental success there would have been no Futurama. While Futurama is buried, The Simpsons is still a Fox cash cow - I remain hopeful that we will see more from Groening in the future.

One can always hope.

What an amazing film.

I finally got a chance to watch this movie at Movie Night this week, and was simply blown away by how well done it was. The animation was flawless, and Pixar obviously understood the same basic principles that have been evident in World of Warcraft - stylized graphics make you forget that you're looking at something that's generated by a computer. The theory is that no matter how good a computer simulation of a human looks, it will always miss the real thing by enough to appear "wrong" - such "wrongness" can be easily seen in the Final Fantasy movie, which (on top of being a bad movie) tries to actually reproduce humans digitally instead of simply hinting at them.

The stylized look works marvels. The characters can do things that seem perfectly plausable in their world, which is a cartoon world, and we never have to question it.

But as with all the great Pixar movies, what make The Incredibles a winner are the simply astounding storylines, the clever dialogue, and the superb acting. It's amazing, really, and it's possibly my favorite film in recent memory - how they can manage to make something so good that it can transcend all demographics is just beyond me. I'm sure kids will love this movie, I'm sure adults will love this movie - it's difficult for me to imagine anybody who fails to love this movie. It's at turns remorseful, hopeful, exciting, hilarious... it manages to somehow superimpose the family life that we all recognize upon a family that has powers that are completely unfounded in our reality. But when the two main characters (Mr. Incredible and Elastigirl, husband and wife) start arguing about which route to take, with argumentative kids in the back, all the while barreling down a crowded highway at super-sonic speeds, it's so charming and somehow so believable that you have to imagine that these characters are more realistic than what we see in so many live action dramas.

Sadly, though the end seems to come far too quickly and leaves us praying for a sequel, none will ever be made by Pixar. The rights to this "property" will be retained by Disney, and any sequel will inevitably be befouled by their money-grubbing taint. But that's OK, really - The Incredibles is good enough to stand alone, and we'll just treat anything else Disney tries to spew at us with that name on it in the same light as Highlander 2 - to us, it never happened.

Robert Holt put together a great little piece on the game.

At the very end of the article, he actually gives out his server/character info - I don't suppose NPR correspondents have too terribly many online stalkers to worry about.

New version of Wordpress is up and running. New default theme as you have certainly noticed, this one I may actually end up changing (I never did stray from the old default).

I've got to set up the anti-spam hacks again, at some point. I've got some WoW to do at the moment though...

I recently removed Fedora Core from my life by installing Gentoo on my mythtv system. I remain skeptical as to Gentoo's overall value for most circumstances, but it's particularly well suited for this task because:

1) The compile USE flags can be used to customize MythTV and its miscellaneous requirements to your tastes
2) Many of the funky drivers and software packages exist in Portage

Gentoo is to my knowledge the only distribution that provides an "official" way to install MythTV and the crazy tv-related drivers, if that means anything to you. Fedora of course has atrpms and other third party repos, from which you can pull the packages you require with yum, but you end up with a bastard system with a mish-mash of vendor supplied packages and 3rd party versions. Everything may work - maybe - but your system is at best a confusing mess.

Gentoo has a lot of weaknesses, which I mentioned before. But if there's any task it's particularly well suited for, I believe MythTV is it.

Let's look at the big issues and why they're tolerable:

1) Installation. Installation still sucks. This time I used stage3, and it sucked somewhat less, but it definitely still sucked. I'm willing to put up with the horrendous and laborious install because, once the system is up and running, I should NEVER have to go through something like it again (compare to fedora core, which you need to effectively reinstall from disc with every version upgrade).
2) Compilation. Yes, it takes a while. Yes, it's a pain in the ass. But on my P4 2.4 GHz it's not unbearable, and given that I only have the bare minimum number of packages installed it didn't take *too* long all told.
3) Moving Target. Gentoo is a moving target, as are all the ports. There is no freeze, it's always changing. But then, so is MythTV - in this case, that's pretty much irrelevant.

Gentoo is, basically, a less mature clone of FreeBSD built around the Linux kernel with a core set of GNU tools. But given that all of the funky software and drivers you need for MythTV are included, it's actually very well suited for this task.

So, here's the gentoo-specific parts of the procedure (instructions on the configuration of the various services can be found in the mythtv docs and/or the man pages and/or the fedora mythtv howto, which is undoubtedly the best mythtv-related resource out there).

1) Install Gentoo (check out the handbook). I recommend stage 3 unless this is just a learning exercise. When you're configuring your kernel, make sure to use the drivers you need, and include support for i2c (for LIRC), alsa, and bttv (if you'll be using brooktree). Be sure to modify your /etc/modules.autoload.d/kernel-2.6 file to include stuff such as your NIC driver if you didn't compile it into your kernel. After you get done with this crap, you probably want to take a break.

2) Get sound working. Make sure to include "alsa" in your /etc/make.conf USE flags, and if you're feeling daring you may want to add "-oss" there as well. emerge alsa-utils. If you compiled your kernel with oss emulation support, you will want to emerge alsa-oss here as well. Start the alsa daemon (/etc/init.d/alsasound start) which should run alsaconf the first time you do it, if not run alsaconf manually. Make sure to de-mute your sound sources with alsamixer if alsaconf didn't do it for you. Check with mpg123 on an mp3 file to ensure that things work, then "rc-update add alsasound default" to make it active on boot. Note that for some reason having the emu10k1 (soundblaster) driver compiled into the kernel itself caused issues for me (probably due to alsasound daemon wanting to load a module), recompiling as a module fixed them. You don't need to load emu10k1 module in your modules.d, the alsasound daemon loads it when it starts.

3) Get your video drivers, X, and mplayer. I installed nvidia-glx which (I think) pulled in nvidia-kernel as a dependency. The 4k stacks option is no longer needed when compiling your kernel, just FYI if you used older nvidia drivers. emerge nvidia-glx, xorg-x11, some fonts (freefonts, corefonts, maybe some others as needed), mplayer, and your window manager (I used evilwm). Note that this will be a LONG compile, so it might be a good thing to do right before bedtime. When you feel like digging in again, run xorgconfig if you don't want to modify the default xorg.conf file manually. Make *sure* to specify the "nvidia" (not "nv") driver here (again, consult the documentation for the more essoteric options you may want for TV out). You can try using the x font server, but I gave up on that after a few minutes of tinkering unsuccesfully, instead just using the local font directories in my xorg.conf file. Fire up xdm and make sure you can log in ("/etc/init.d/xdm start"). You'll probably want to run xdm by default, unless you'll start x some other way... "rc-update add xdm default"

4) Get your Hauppauge drivers working. Modify /etc/portage/package.keywords to include something like "media-tv/ivtv ~x86" to allow for dev versions. emerge ivtv and pray! If you can insmod it afterwards then add "ivtv" to your /etc/modules.autoload.d/kernel-2.6 file. This DID work for me, but I've seen mixed reports and other instructions, so it's entirely possible that the ivtv ebuild will be broken for you - if so, you'll need to do some more work and install it the generic, non-gentoo way, which I won't go into here. You'll want to test this with mplayer - start up mplayer and try to play "/dev/video0" - if it works, move on.

5) Get and install mysql. Emerge mysql - easily done. Make sure permissions are right and note your root database password. You need to set up a database for mythtv at some point, but I can't remember what's involved exactly - check the mythtv docs for info (since I was migrating from an existing install, I already had my mysql database which I just reused). Start mysql on boot - "rc-update add mysql default"

6) Set up lirc. Unmask development packages again in /etc/portage/package.keywords "app-misc/lirc". Set the flag to compile for the hauppauge IR device (add "LIRC_OPTS="--with-driver=hauppauge"" to /etc/make.conf). Go ahead and add "lirc" to your USE variable in make.conf. emerge lirc, load the modules (lirc_i2c), then test it with irw to make sure it works. Add to /etc/conf.d/lircd "LIRCD_OPTS="-d /dev/lirc/0"" to specify the device file to use. Start the daemon ("/etc/init.d/lircd start"). Create (using irrecord) or borrow a lircd.conf file, a ~/.lircrc file, and a ~/.mythtv/lircrc file for your remote. Add "lirc_i2c" to your /etc/modules.autoload.d/kernel-2.6 file.

7) Get and install mythtv and friends. emerge -s myth first to locate all the various mythtv packages you'll need. Modify /etc/portage/package.keywords to include entries such as "media-tv/mythtv ~x86" for *all* of those mythtv packages to unmask development and get the .17 branch (this is a PITA, but myth* has a TON of deps and you probably don't want dev versions of them all - if you do "export ACCEPT_KEYWORDS="~x86"" and then emerge all the myth stuff you'll end up with a lot of development versions of the deps, which may not be what you want). After it's all compiled, run mythsetup within X and set things up as desired. Fire up the backend ("/etc/init.d/mythbackend start"), make it default ("rc-update add default mythbackend"), start the frontend. Pray everything works.

8) Tidy up, create your .xinitrc and .xsession files as needed.

All in all, there's not much gentoo-specific stuff to it. I've set up mythtv on a few different distros now and doing so on gentoo is actually relatively painless, though the compile times are a bit frustrating if you have a slower machine. The best thing is that we can (in theory at least) now use portage to update every single piece that we've installed.

So, while gentoo linux does still suck, I wholeheartedly recommend it in this role.