Wow:

I can still remember watching this on my family’s old TRS-80 Model I, absolutely entranced by the thing. I would’ve been what? Four? Five?

“There is the theory of the moebius, a twist in the fabric of space, where time becomes a loop.”

Just found one of our DNS boxes (running Solaris 9/Sparc) doing this:

bash-2.05$ date
Thu Nov  8 17:24:25 EST 2007
bash-2.05$ date
Thu Nov  8 17:24:26 EST 2007
bash-2.05$ date
Thu Nov  8 17:24:27 EST 2007
bash-2.05$ date
Thu Nov  8 17:24:27 EST 2007
bash-2.05$ date
Thu Nov  8 17:24:25 EST 2007
bash-2.05$ date
Thu Nov  8 17:24:25 EST 2007
bash-2.05$ date
Thu Nov  8 17:24:26 EST 2007
bash-2.05$ date
Thu Nov  8 17:24:27 EST 2007
bash-2.05$ date
Thu Nov  8 17:24:27 EST 2007
bash-2.05$

That’s just me hitting ‘date’ repeatedly. Its the results which are rather strange.

Guess its time for a reboot, and hope that the machine comes back.

Update:

Reboot was successful. Pulled a copy of syslog off the machine before doing so, just in case Really Bad Things ™ were about to happen. Found the following interesting bits right before the time loop began:

Dec 30 19:00:00 xyz named[3721]: [ID 873579 daemon.crit] timer.c:447: fatal error:
Dec 30 19:00:00 xyz named[3721]: [ID 873579 daemon.crit] RUNTIME_CHECK(isc_time_now((&now)) == 0) failed
Dec 30 19:00:00 xyz named[3721]: [ID 873579 daemon.crit] exiting (due to fatal error in library)

Note the odd timestamp. The epoch, maybe, corrected for GMT-5?

Don’t really know what is going on, and Google was, surprisingly, no help with the error message.

So I’ve been slowly working on transferring the feeds I read from Bloglines to Google Reader (hey, I’ve pretty much sold my soul to Google…what’s one more service?)

One of the more interesting features is the ability to share links of blog articles that you find, and make them available for others to peruse (and even subscribe to that as an RSS feed). You can find mine at http://www.google.com/reader/shared/10605254425022528773.

In setting up this new hosting environment, I decided to go with lighttpd to handle webserving duties, as opposed to Apache. All-in-all, I’m quite happy with lighty. It’s pretty damn fast, and the configuration file is both easy and amazingly powerful.

Of course, one of the things that lighty is missing is an .htaccess-type file. And some of Apache’s more powerful rewriting capabilities. In particular, the ability to see if a file or directory actually exists, and then rewrite the URL based on that. All of this, of course, would be particularly helpful in setting up Wordpress, with its “clean” URLs.

Most of the solutions that I have seen to this problem involve a lot of lines of rewriting individual, hard-coded filenames. For me, not terribly optimal, especially since I’m mainly looking to rewrite things like http://blog.jgumby.com/something-that-doesnt-really-exist/ into http://blog.jgumby.com/index.php/something-that-doesnt-really-exist. So, basically, my feeds, archives, &c. What I’ve come up with is the following:

And that’s it. I’m sure I’ll find that I’ve missed something really obvious, particularly with a feature of Wordpress that I just don’t use very much. But I’ll cross that bridge once I get there. The main thing to remember is that if you have other directories in your Wordpress install directory that do actually exist, you’ll have to add entries for them (or just modify the first line to include them, you regex guru).

Napster rues Microsoft, player glitches

NEW YORK (Reuters) – Technical glitches by Microsoft and the digital music device makers have hampered Napster Inc.’s ability to close the gap with Apple’s iTunes, the dominant online music service, Napster’s chief executive said on Tuesday.

“There is no question that their execution has been less than brilliant over the last 12 months,” Napster Chairman and Chief Executive Chris Gorog said at the Reuters Global Technology, Media and Telecoms Summit in New York.

“Our business does rely on Microsoft’s digital rights management software and our business model also relies on Microsoft’s ecosystem of device manufacturers,” he added.

Yeah, way to blame everyone but yourself. Perhaps there’s a chance that your music subscription service just isn’t the greatest thing in the world?

Google Reader is available. It certainly has a pretty, and new, interface. Not quite sure if it’s going to replace Bloglines as my newsreader of choice, though. Guess only time will tell.

No more turns

Latin, I love it. I love Latin. I love the sound. I love all the Xs. I love that JLo. She’s Latin. She’s got what all the kids are calling today…She’s got that much back. Or, mucho espalda, as they say in esapañol.

Today, I asked Google Maps for directions to Chantilly, VA. It told me:

1. Head north from N Vermont St – go 0.1 mi
2. Turn left at Wilson Blvd – go 2.2 mi
3. Turn right at Arlington Blvd – go 0.3 mi
4. Bear left onto the US-50 W ramp to Fairfax – go 3.0 mi
5. Take the I-495 ramp to Richmond/Rockville (I-66) – go 0.1 mi
6. Take the I-495 N ramp to Rockville (I-66) – go 1.0 mi
7. Take the I-66 W exit 49C to Vienna/Front Royal – go 6.7 mi
8. Take the US-50 W exit 57B to Fair Oaks/Winchester – go 0.9 mi
9. Continue – go 0.2 mi
10. Continue on Lee Jackson Memorial Hwy – go 7.0 mi
11. Continue on US-50 W – go 1.9 mi
12. Chantilly-specific directions

As far as I can tell, I didn’t ask for the longest, and most convoluted, route to get to US 50 past the Beltway possible.

Even better, if I ask for directions from the intersection of Wilson Blvd & Glebe Rd (less than two blocks from my house), it tells me to drive up Glebe, hang a left on Washington Blvd, and then get on I-66W when they intersect.

My house is less than half of a mile from the Fairfax Drive on-ramp to I-66W. The Wilson & Glebe intersection is even closer to said on-ramp. I ask you, what does Google Maps have against Fairfax Drive in Arlington?!

Just some quick stats:

Now, the sad thing is, this isn’t really a fair comparison.

For some background, this is happening on a Windows XP box with dual-P4 CPUs (3.4Ghz each) and 2GB of physical RAM.

The Webtrends profile is reading 2.2GB worth of log files, spread out over two servers (load-balancing). The start date is Aug 1, 2004.

The Urchin profile is for only one of the load-balancing servers, and it’s start date is Jan 1, 2005. It has 20.0GB of logs to read.

Both applications are running on the same machine, obviously at the same time. Just to reiterate, Webtrends is still working on it’s report. It’s much, much smaller report. At the time of this post, it had made it through January 23, 2005. So at least somewhere past the 50% point.

*sigh*

While walking back from Lazy Sundae (hey, they did good on their last few health inspections), I got to thinking: eating bubblegum ice cream is like eating watermelon…just with tastier seeds.

Just thought I’d share the kind of nonsense that fills my head while walking around and eating ice cream.

Also bought some new earphones for my iPod from the Apple Store up in Clarendon. Sadly, they no longer carry the old skool, white earbuds. Well, at least, not without a remote control for the dock-style iPods. Which I don’t have. Thanks for making me regret my early adoption of the iPod, Apple.

They’re trying to sell those damn in-ear buds. Which I bought a pair of a while back, and I just hated them. My ears must be freakishly designed (and no comments from the peanut gallery), because no matter what combination of attachments I used, I could never get a comfortable fit. So it goes.

Instead, I bought the Plantronics MX100S, which has the added advantage of being a handsfree headset for my cell phone. After some fiddling around to get the bastards in my ears, I was very happy with how they felt on the walk home, and the sound was great. $40 well spent, I think.