If you're not one of label's top artists with a renegotiated contract after your second album has made the label ungodly amounts of cash, you're treated like shit anyway.
Bravo for Harvey Danger. Download their music, if you like it, send 'em something for it.
FWIW, I sent them ten bucks because I hope more artists will take the opportunity to ditch the lables.
Juniper contributed a fair bit of code directly back to FreeBSD. In point of fact, the policy of the Engineering department was to contribute back absolutely everything that could be ported back into FreeBSD without causing FreeBSD grief or getting super-proprietary.
Why? It's good business sense. The company doesn't have to re-merge their changes later when importing new flavors of FreeBSD, and they have access to testers and developers who build on top of those contributions.
How has it helped the average FreeBSD user? Improved timer and time of day code in the kernel, modularization for a smaller footprint OS, tons of bug fixes, scripting support for system installation,...
Frankly, as a former lead of a software organization that makes commercial products on top of FreeBSD, it's difficult to imagine anyone being so short-sighted as to not wish to contribute code back. The amount of time and effort my company spent doing it was greatly rewarded.
p.s. The problem with the FreeBSD networking stack wasn't packet moving performance, it was routing support. The *BSD stacks are really meant to support end systems with tens of thousands of active sockets, not routers carrying over a million forwarding table entries.
The endpoint is way past geosynchronous orbit, but the counterweight is less than the cable. It's a win because if you put the endpoint out far, you get greater centripidal force for an extra-orbital launch.
What if end users could have a choice between a regulated and unregulated service, like, ostensibly, they have now with VoIP?
If I want the government to guarantee that I can get service, that 911 calls will make it to a dispatch center (where they will be promptly ignored), and that some form of lowest common denominator of service is available in rural Nebraska, then I'll subscribe to regulated service.
If I'm willing to play fast-and-loose and my phone's won't work if there's an earthquake, 911 may or may not work (market forces would eventually make it work, cf. Vonage), my voice quality may sound as bad as a cell-phone, and my phone will be about as reliable as my connection to Yahoo! or Google, then I'll eschew the regulated services.
Why does it have to be all or nothing? Yes, the unregulated services will eat the regulated services for lunch, for a while, but eventually a new ballance will be struck.
Hard coded IP address? That's pretty damned useless. What stops Verisign from returning different IP addresses every other day.
A better fix (which needs to be implemented in the nameserver, is to issue a query for a NS or SOA record for the second level domain first. If the NS or SOA record does not exist, any A records returned from Verisign will be bogus.
Of course, an even better fix is to configure your nameservers to not query the Verisign top level domain servers. They only control 3 out of 13 (a, c, and d if memory serves. The non-verisign domain servers don't do this BS).
Well, people are suggesting looking for some bogus address range of Verisign servers, which they could change any time they want.
However, it appears that if you do a NS or SOA query on a bogus domain, you'll not get back any A records. If you do an A query on said domain, you get back an A record for Verisign servers.
Patch is obvious. It sucks, because it doubles the traffic on the root nameservers, but what more can we do? Oh, I guess we could just configure the root nameservers to take away.com and.net from Verisign...?
And it will be a shining success like IPv6, and the communist state will wither away, and we will all eat ice cream.
Any proposal that defines an "NG" where there isn't a gun being held to everyone's head to jump to the "NG" will be doomed to failure. About the last successful "NG" transition in the internet was BGPv3 to BGPv4, which/had/ to happen so we could implement CIDR, otherwise we would have run out of routing table space and addresses. That succeeded because less than 100 major entities needed to make the change, and they were fiscally motivated to do so (otherwise they would have had dead Cisco AGS+ routers in 18 months).
Right now, there isn't a big enough hammer to solve this problem, and backwards compatibility is considered desirable. Look at open relays...why do spammers still use them? If everyone looked at ORBS/SPEWS/RBL lists they would cease to be valuable. However, if an administrator can't even fix their open SMTP relay, how are they going to upgrade to a version of BIND that supports RMX records?
For e-mail authentication and accountability to work, everyone needs to do it. Right now, SPAM isn't enough of a hammer to force folks to do it. Hell, I never see SPAM anymore, I've got spamassassin and I update it every time a new release comes out in Debian/unstable. That's less trouble for me than waiting for the rest of the world.
Right, but mineral or transformer oil would have a lower viscosity.
If you can get *some* convection going, it should radiate heat from all surfaces, thus turning the whole box into a big heat sink. I don't think they need the pump/heat dump.
I suspect a better packaging scheme would involve a flat thin pizza box design to maximize the radiating surface and minimize the distance between the hot parts of the system and the walls. On top of this, I would put a second surface, at least a few inches above, to act as a heat shield to keep it out of the light.
I don't like the whole design, I agree with the others who suggest keeping the computer separate and just running ethernet upstairs to the 802.11 module & antenna, but there may be circumstances where that just isn't practical, so...
The HDD is just a bad idea, they should have gone for a solid state disk (CF flash or DiskOnChip).
"The Scarlet Letter"
Gee Jim, you keep checking out the p0rn sites, we're going to have to bitch-slap ya...
Of course, what idiot is going to use a monitored computer.
The plane went through intensive spin testing during certification.
The plane has modern anti-spin enhancements that conventional aircraft don't have (wing cuff, wing tips). It is damn near impossible to unintentionally spin this plane. For example, you can use the alierons to actually recover a dropped wing (something that would cause you to have a really bad day in a Mooney).
Thise anti-spin enhancements do make spin recovery more difficult should you actually manage to jump off the cliff and enter a full spin despite the safety rails provided by the wing. That is why the chute is the officially sanctioned recovery for the craft.
The FAA inspectors actually requested that Cirrus be certified as spin resistant, rather than do spin testing because they felt it was more important to promote spin resistance than spin recovery in this, and all future certifications.
Sometimes I think Cirrus regrets this decision because they have taken a lot of crap from the marketing departments of historic airplanes.
From talking to other SR20 and SR22 owners, the SR20 is noticably slower when it picks up ice (15-20kts quickly), the SR22 seems to be able to carry quite a load with no problems.
The 22 has a larger engine and larger wingspan.
I've spoken to Cirrus about the plane that deployed the chute. They purchased it from the insurance company and are planning to fly it again. There was almost no damage to the plane.
The rule of thumb is, if you pull the chute, the insurance company has bought the airplane. However, there is "totaled" for insurance purposes and totaled in real life.
Fuel related accidents are due to fuel mismanagement. Just like pouring more money into a fundamentally mismanaged startup, or giving a guy more rope to hang himself, someone who doesn't manage their fuel is going to run into trouble no matter how big the tanks are.
Having more fuel will not save you from a mid-air collision (which happened to the CEO of Cirrus when he was learning to fly), nor will it enable your passengers to land your plane should you become incapacitated, nor, as in the recent example, will it save you should you experience a control failure.
You're right, it's not a panacea, but it is one more safety component to have at your disposal should there be a problem. I suspect Lloyd (the pilot) doesn't mind the fact that he gave up 88lbs of useful load for the parachute. Neither do I.
Obligatory correction:
The chute and all assorted equipment (lines, anchors, squibs) weighs 88lbs. The aircraft itself has an 81 gal fuel tank, and the usuable load of the aircraft *WITH*FULL*FUEL* is 647 lbs. The plane flies 4:50 with full tanks, which is far beyond most passenger's bladder range.
So if I get my DNA sequenced and copywritten, due to the great commonality between my DNA and the rest of humanity (over 99.99% the same), can I nail everyone else for copyright infringement? Can I get a temporary restraining order against further copyright infringment and order everyone else to be sterilized?
I guess it's good for us that Hitler didn't have the DMCA in his back pocket.
Slashdot Mirror
If you're not one of label's top artists with a renegotiated contract after your second album has made the label ungodly amounts of cash, you're treated like shit anyway. Bravo for Harvey Danger. Download their music, if you like it, send 'em something for it. FWIW, I sent them ten bucks because I hope more artists will take the opportunity to ditch the lables.
Juniper contributed a fair bit of code directly back to FreeBSD. In point of fact, the policy of the Engineering department was to contribute back absolutely everything that could be ported back into FreeBSD without causing FreeBSD grief or getting super-proprietary.
...
Why? It's good business sense. The company doesn't have to re-merge their changes later when importing new flavors of FreeBSD, and they have access to testers and developers who build on top of those contributions.
How has it helped the average FreeBSD user? Improved timer and time of day code in the kernel, modularization for a smaller footprint OS, tons of bug fixes, scripting support for system installation,
Frankly, as a former lead of a software organization that makes commercial products on top of FreeBSD, it's difficult to imagine anyone being so short-sighted as to not wish to contribute code back. The amount of time and effort my company spent doing it was greatly rewarded.
p.s. The problem with the FreeBSD networking stack wasn't packet moving performance, it was routing support. The *BSD stacks are really meant to support end systems with tens of thousands of active sockets, not routers carrying over a million forwarding table entries.
The endpoint is way past geosynchronous orbit, but the counterweight is less than the cable. It's a win because if you put the endpoint out far, you get greater centripidal force for an extra-orbital launch.
While reading this, a weird idea struck me...
What if end users could have a choice between a regulated and unregulated service, like, ostensibly, they have now with VoIP?
If I want the government to guarantee that I can get service, that 911 calls will make it to a dispatch center (where they will be promptly ignored), and that some form of lowest common denominator of service is available in rural Nebraska, then I'll subscribe to regulated service.
If I'm willing to play fast-and-loose and my phone's won't work if there's an earthquake, 911 may or may not work (market forces would eventually make it work, cf. Vonage), my voice quality may sound as bad as a cell-phone, and my phone will be about as reliable as my connection to Yahoo! or Google, then I'll eschew the regulated services.
Why does it have to be all or nothing? Yes, the unregulated services will eat the regulated services for lunch, for a while, but eventually a new ballance will be struck.
Hard coded IP address? That's pretty damned useless. What stops Verisign from returning different IP addresses every other day. A better fix (which needs to be implemented in the nameserver, is to issue a query for a NS or SOA record for the second level domain first. If the NS or SOA record does not exist, any A records returned from Verisign will be bogus. Of course, an even better fix is to configure your nameservers to not query the Verisign top level domain servers. They only control 3 out of 13 (a, c, and d if memory serves. The non-verisign domain servers don't do this BS).
Well, people are suggesting looking for some bogus address range of Verisign servers, which they could change any time they want. However, it appears that if you do a NS or SOA query on a bogus domain, you'll not get back any A records. If you do an A query on said domain, you get back an A record for Verisign servers. Patch is obvious. It sucks, because it doubles the traffic on the root nameservers, but what more can we do? Oh, I guess we could just configure the root nameservers to take away .com and .net from Verisign...?
And it will be a shining success like IPv6, and the communist state will wither away, and we will all eat ice cream.
/had/ to happen so we could implement CIDR, otherwise we would have run out of routing table space and addresses. That succeeded because less than 100 major entities needed to make the change, and they were fiscally motivated to do so (otherwise they would have had dead Cisco AGS+ routers in 18 months).
Any proposal that defines an "NG" where there isn't a gun being held to everyone's head to jump to the "NG" will be doomed to failure. About the last successful "NG" transition in the internet was BGPv3 to BGPv4, which
Right now, there isn't a big enough hammer to solve this problem, and backwards compatibility is considered desirable. Look at open relays...why do spammers still use them? If everyone looked at ORBS/SPEWS/RBL lists they would cease to be valuable. However, if an administrator can't even fix their open SMTP relay, how are they going to upgrade to a version of BIND that supports RMX records?
For e-mail authentication and accountability to work, everyone needs to do it. Right now, SPAM isn't enough of a hammer to force folks to do it. Hell, I never see SPAM anymore, I've got spamassassin and I update it every time a new release comes out in Debian/unstable. That's less trouble for me than waiting for the rest of the world.
Yet another money making opportunity for Google. :-)
Of course, I really don't want to meet the woman who's doing the same kind of searches I do... ewww.
What's amazing is not that a Senator can be bought, but rather how inexpensive it is.
$13,800 contribution from Lockheed Martin?
At that rate, the Slashdot crowd could own all of the Senate and Congress and still have money left over to buy a burger.
Why are we screwing around with the DMCA and RIAA all the time? Just buy your own congresscritter. Take two, they're cheap.
Math is hard! Let's go Shopping! Barbie
I don't like the whole design, I agree with the others who suggest keeping the computer separate and just running ethernet upstairs to the 802.11 module & antenna, but there may be circumstances where that just isn't practical, so...
The HDD is just a bad idea, they should have gone for a solid state disk (CF flash or DiskOnChip).
Yes, but doesn't this spawn yet another "industry" of professional copyright maintainers? Sort of like domain name squatters?
"The Scarlet Letter" Gee Jim, you keep checking out the p0rn sites, we're going to have to bitch-slap ya... Of course, what idiot is going to use a monitored computer.
s/whuffie/slashdot karma/ and we're already there.
Disclosure: I own one.
From talking to other SR20 and SR22 owners, the SR20 is noticably slower when it picks up ice (15-20kts quickly), the SR22 seems to be able to carry quite a load with no problems. The 22 has a larger engine and larger wingspan.
Jim Fallows wrote a book called Free Flight last year about the company that built this plane. It's an excellent read.
I've spoken to Cirrus about the plane that deployed the chute. They purchased it from the insurance company and are planning to fly it again. There was almost no damage to the plane. The rule of thumb is, if you pull the chute, the insurance company has bought the airplane. However, there is "totaled" for insurance purposes and totaled in real life.
Fuel related accidents are due to fuel mismanagement. Just like pouring more money into a fundamentally mismanaged startup, or giving a guy more rope to hang himself, someone who doesn't manage their fuel is going to run into trouble no matter how big the tanks are.
Having more fuel will not save you from a mid-air collision (which happened to the CEO of Cirrus when he was learning to fly), nor will it enable your passengers to land your plane should you become incapacitated, nor, as in the recent example, will it save you should you experience a control failure.
You're right, it's not a panacea, but it is one more safety component to have at your disposal should there be a problem. I suspect Lloyd (the pilot) doesn't mind the fact that he gave up 88lbs of useful load for the parachute. Neither do I.
Obligatory correction: The chute and all assorted equipment (lines, anchors, squibs) weighs 88lbs. The aircraft itself has an 81 gal fuel tank, and the usuable load of the aircraft *WITH*FULL*FUEL* is 647 lbs. The plane flies 4:50 with full tanks, which is far beyond most passenger's bladder range.
So if I get my DNA sequenced and copywritten, due to the great commonality between my DNA and the rest of humanity (over 99.99% the same), can I nail everyone else for copyright infringement? Can I get a temporary restraining order against further copyright infringment and order everyone else to be sterilized? I guess it's good for us that Hitler didn't have the DMCA in his back pocket.