Nov 2003

Sat, 15 Nov 2003

Brain-Gain vs. Brain-Drain Cities

No surprises here. Indeed, halfway into it when I wondered when they'd start quoting Richard Florida, they quoted Richard Florida. The major comparison is between Cleveland and Seattle, an interesting contrast between turn of the century leading cities. They put Pittsburgh on the brain-drain list. No surprise there either.

Going the entrepreneurial route myself, I can attest to the difficulties encountered in a place like this, where just finding an audience that can comprehend what you are talking about is difficult, let alone the relative advantage or disadvantage from a hiring standpoint.

Around here, when you tell somebody that you are doing your own thing, a startup or something like that, you get the distinct feeling from the looks on their faces that they think you are an insane man. When I lived in Seattle or Boston, that was more a cause for high fives and congratulations.

That intangible cultural aspect is the biggest challenge that all of these rust belt incubators face. It's not really that there aren't fundable ideas, or people; it's that getting a critical mass of them together in one place, and having them all be part of the same risk-taking soup is nearly impossible. They were pickled in the wrong juice, and changing the pickling formula takes decades.

And it goes to the core of the culture. When I take my dog outside at 10 PM, there's not a light on in my entire neighborhood. If I go out to find something to eat at 10 PM, more than half the restaurants are either closed or stopped serving food. They're smart business people. Very few people here eat at that hour.

In Seattle, if you go out at 10 PM looking for food, you'll find a host of places ready to serve you, and they're bustling with folks that just came out of work. But why would anybody work that late unless they were heavily invested in the business?

Posted at: 11:20 | permalink

Fri, 07 Nov 2003

Linuxant DriverLoader Bugs

Well, taking the good with the bad, it turns out that the linuxant driverloader that I mentioned a couple days ago has bugs. No suprise there. The fact that it works at all is a small miracle. The bugs go like this... When I installed it, I had an eth0 and eth1 interfaces on the box. It created it's own eth0, destroying the 100 Mbit device that was there in the process. Linuxant tech support did finally respond and say yeah, that's a problem. Mark Boucher from Linuxant responds:

The problem with eth0 vs eth1 isn't driverloader-specific but a consequence of how network devices are registered and their names allocated under linux. Basically the instance number isn't persistent but set according to the order in which the devices are initialized. So a newly installed device which ends up being started before an existing one might be assigned the latter's slot by the kernel, which is what happened on your system. We would very much appreciate if you could correct the text on your website since it unfairly blames DriverLoader for this problem. When searching Google for "linuxant driverloader" your site comes up quite early and new users or potential customers could misinterpret your comments. Btw in version 1.25 we have added the --netdevname option to dldrconfig so that the "eth" prefix can be changed to something else in order to avoid any collisions with existing ethernet devices. We cannot use another name by default since some distributions and tools only recognize "eth" devices.

Also, I wrote to them last night to ask them exactly what you're supposed to do when you upgrade kernels, which isn't exactly clear, since the device has to be compiled against the current kernel and the driverloader configuration was not migrated when I updated kernels last night.

Mark Boucher of Linuxant responds:
After booting a new kernel just run "dldrconfig --kernel" (optionally with --auto to make the process automatic). This question is answered in our FAQ at http://www.linuxant.com/driverloader/wlan/faq.php#17

Posted at: 11:24 | permalink

Thu, 06 Nov 2003

Fedora Via Bittorrent

I decided to give bittorrent a try with the new Fedora Core release this morning and indeed, I'm getting sustained download speeds of nearly 150KB/sec! That's only 4 hours and change for the entire 1.8 GB fileset. Bittorrent only gets better with more users and it seems that this fedora core download has attracted just enough to make it compelling. My previous attempts with bittorrent were hideously slow, under 10 KB/sec. Join the torrent, it's easy!

Posted at: 12:04 | permalink

Tue, 04 Nov 2003

Russert Misses the Fatal Rumsfeld Question

I watched Tim Russert Interview Don Rumsfeld on Meet the Press on Sunday. You could say that Russert took it far too easy on Rumsfeld. People must be feeling sorry for him. The interview really caught my attention when it reached this juncture:
Russert: Time magazine reports this today, that this question was asked in the closed briefing with senators, "'What troop levels do we expect to have in Iraq a year from now?,' asked Senator Bill Frist, the Republican leader. And with that, the Pentagon chief began to tap dance." Do you believe that you have an obligation to tell our leaders in Congress what your best estimate is for troop levels in Iraq a year from now? Rumsfeld: You know, since -- any war, when it starts, the questions are obvious. The questions are: How long is it going to last? How many casualties will there be? And, How many troops will it take? Now, those questions can't be answered. Every time someone has answered those questions, they've been wrong. They have been embarrassingly wrong. I'll use another word: They have "misinformed." By believing they knew the answers to those questions, they've misinformed and misled the American people. I made a conscious decision at the outset of these conflicts to not pretend I knew something I didn't know. And what I have said is just that. I have said it is not knowable.
What's amazing is that Rumsfeld continues talking and a moment later says:
Now, the answer as to how many U.S. forces will be there a year from now depends entirely on what happens in the security situation on the ground, first and foremost. Second, it depends on how fast we are able to build up the Iraqi forces. What's happening is the total number of security forces in that country have been going up steadily. We've come down from 150,- to 130,000 troops. The coalition troops of about 30,000 have stayed about level. And what's changed is the Iraqi troops have come up from zero to 100,000, heading towards over 200,000 next year.
And Russert failed to deliver the follow-up question that I'd like to see Rumsfeld answer: If you can predict the number of Iraqi troops involved next year, why can't you predict the number of American troops? Are you misinforming and misleading the American people by stating the number of Iraqi Troops?

Posted at: 12:47 | permalink