Maybe if they focused on "stories that didn't make the TV news", they'd make more sense. But almost all the stories mentioned got some major attention in the print media, and they're all on line.
The TV problem has more to do with what looks good on TV. Which is why we have actors in politics.
I'm not sure I believe those numbers. The Private Equity Pool, Stanford's participation in the venture capital industry, hasn't been marked to market yet.
This is a bit off topic, but the VC business is currently stalled. It's not the companies that have died that are the problem.
They've been liquidated and require no further attention. It's the thousands of
companies left over from the dot-com boom that aren't going anywhere. They're not successful enough to ever pay off the money invested in them, but they bring enough cash to meet their current expenses. (VA Systems, Slashdot's parent, is an excellent example.) There are thousands of VC-financed companies still in lingering-death mode. Much VC time goes into tracking these companies and trying to work out exit strategies.
It's hard to value these walking dead. Some of them might, someday, be worth something. They tend to be carried on the books for a value more appropriate to their glory days, since no big event has occured to force a repricing to market value. This overvalues private equity portfolios, like Stanford's.
During the dot-com boom, Stanford was getting about a 20% return on the endowment, and they got carried away. Then when the market tanked, they started hitting on us alumni for more money.
There's nothing there, yet we need help
on
Fabian Pascal Reacts
·
· Score: 2, Insightful
That article is just a long flame.
There are real issues, but the article doesn't address them.
Tree-based databases are thirty years old. See MUMPS. Explictly linked databases are also thirty years old. See the CODASYL DBMS. XML database enthusiasts need to read up on those old systems to avoid making the same mistakes.
Relational databases aren't enough, either.
When you find yourself putting columns of serial numbers in tables so you can link tables together, the relational model isn't fitting the problem.
These issues are not being addressed all that well.
The New York City subway value card system, MetroCard, has much greater security. NYC knew they had to come up with something that really was secure. They were losing $20,000,000 a year on turnstile fraud before they switched over.
MetroCard stores the data both on the card and in a database. They're crosschecked every time a card is used. The data on the card is encrypted, of course. Cards are checked by the station computer and by a central database. Cards are read both at entry and exit from the subway, and if entries and exits don't match, it's noticed.
There are protections against fraud by insiders.
This system was intended to be Mafia-proof.
There's a big generation gap on this. Younger people have grown up surrounded by surveillance cameras and cell phones. They assume they're being tracked.
And it doesn't bother them.
I've talked with teenagers about what it means when their cell phone has GPS. They're not bothered by having their location reported. They like the idea of knowing where all their friends are. Then they'd know who's nearby, and could hook up. It's a feature.
Lois Bujold's best work is the Lord Vorkosigan series. She won a Hugo for one of those in 1992. But she has to pay the rent, so she cranks out those fantasy novels. The Vorkosigan series is too complex and unsettling for many readers.
The most common example of this is the slow progressive adaption of Newtonian Physics, and then of Einstein's Relativity: doubters are in abundance, until they are won over to the new paradigm.
Wrong. Once experiment confirmed special relativity, it was widely accepted. Rather quickly, too. Einstein published the key paper in 1905, and by 1910, the physics community accepted it. Definitely not "slow and progressive". Even though Einstein was a nobody at the time. He was still working at the Patent Office.
In the hard sciences, progress is made by experimental confirmation. In areas where theoretical change is "slow and progressive", there's not much progress. Change, yes, progress, no. Consider psychiatry.
And then, there's the specter of "vote buying". If ballots are secret, it's much more difficult to buy votes, because you cannot be sure that they guy whose vote you bought really voted the way he promised;-)
This, incidentally, is one of the problems with paper ballots. In Chicago, the way it worked was that the ward heeler stood outside the polling place. He'd voted early, but didn't put his ballot in the box. Instead, he kept it and marked it. As he paid off voters, he'd give them a pre-marked ballot to deposit. When they came back with a blank one, they were paid.
This sort of thing is rare today. Those old-line regular party organizations are dead. Nobody has enough people in place any more to pull that off. Effort today goes into sucking up to donors and buying ads.
The US Army tried this in the 50s
on
Port-A-Nuke
·
· Score: 1
Assuming that Roland the Plogger didn't get this as confused as he usually does, it's worth noting that
little packaged nuclear plants have been tried before.
The American Locomotive Company built three packaged nuclear plants for the US Army between 1958 and 1962. One was installed at Fort Greely, AK, one at Camp Century in Greenland, and one in the US as a test and training unit.
They worked, but weren't useful enough to be worth the trouble.
Re:As a former nuclear navy reactor operator
on
Port-A-Nuke
·
· Score: 2, Informative
I can tell you that US Navy subs have had few catastophic disasters, and perhaps none at all for a long time.
USS Thresher and USS Scorpion were lost at sea. USS Guitarro sank alongside a pier during construction for reasons that can only be described as Really Dumb, but was refloated and repaired.
No US subs have been lost since the 1970s, though.
I recall hearing about "distributing video on hard drives" a while back.
Why bother? DVD-sized media with more capacity are coming, and compression good enough for HDTV on existing media is already available.
Besides, hard disk replication would be a pain. We'd be back to where things were in the early days of VHS replication. Huge farms of consumer grade VCRs slaved off one good digital tape player. A bunch of people running around changing media. (Later, better ways to duplicate tape were developed.) CD/DVD replication is a high-volume process that's incredibly cheap in volume.
The "broadband problem" is something created by CLECs looking for a better regulatory deal and politicians looking for an issue. It's not a real problem.
Second, because the US has free local calling, good line quality, and plenty of telco switch capacity, dialup works well in the US. In many countries, dialup involves per-minute costs, and you can't stay on all day. It the US, it's been flat-rate monthly for years. And dial-up is really cheap.
Third, more people in the US have Internet access than buy books or subscribe to newspapers. The literate fraction of the population is already on line. If you can't read, even AOL isn't useful.
Most "one time pad" systems found on the Internet aren't. They generate their "one time pad" with a pseudorandom number generator. That is not a one time pad. It's a short-key system based on the seed of the random number generator.
A real one-time pad key generator using current technology would have a hardware device like a noise diode that generated random bits at a high rate and created two identical DVDs. That would be useful. With 4GB of random bits, you could encrypt about 5000 minutes of phone-grade voice.
SCO just adopted a poison pill, intended to prevent a hostile takeover.
SCOX has been in the $3.50 to $4 range all week, with light trading. Six months ago, it was at $14.
Next court date: September 15, Judge Kimball. Motions in both the Novell vs. SCO and SCO vs. IBM cases will be heard. SCO has been stalling to put off this date, but it looks like the judge won't tolerate any more delay and some major issues will be decided in two weeks.
A programmer just isn't going to be competent in a modern game without a decent grasp of basic linear algebra, 4D as well as geometry in 2D and 3D. We often use 4D representations for basic operations (4D homogeneous coordinates for general linear transformations, and the quaternions to represent rotations5) so the ability to reason about higher dimensions is extremely useful. Basic calculus is necessary for all kinds of simulation and rendering tasks. For many rendering tasks, signal-processing mathematics is very important--both linear signal processing as well as the murkier study of spherical harmonics. For any kind of sophisticated simulation, you'll want experience with numerical analysis and differential forms. For networking, information theory and the statistics behind compression and cryptography are necessary to build a robust system.
A good engine programmer should have working familiarity with a great many algorithms--so many that attempting to list them here would be silly. The most necessary algorithms perform tasks like spatial partitioning, clustering, and intersection and clipping of geometric primitives. Most algorithms will be mainly focused on one task area, like rendering or physics, but these algorithms are often very deep and take a while to master. For years we have been mining academic research to find and modify appropriate algorithms. However, a game engine must meet soft realtime requirements, and most academic work in the relevant subject areas is geared toward batch computation. (Most of the past research in graphics has applied to offline cinematic rendering. Most physics algorithms are unstable and can fail outright, which is solved in a batch setting by tweaking the initial conditions and trying again. These algorithms do not adapt successfully to a soft realtime setting.) As games are now starting to be taken seriously by the academic community, this is beginning to change, but most academic research is still pointed in directions that don't do us much good. So, creating a technically ambitious game engine will often require a substantial amount of original research.
Hacking isn't enough any more. You need theory.
Lots of theory.
Game development has some really smart people.
I've met some of them. The best theoretical work today is being done in game development. Game AI is better than academic AI, and it's gaining.
Slashdot Mirror
The TV problem has more to do with what looks good on TV. Which is why we have actors in politics.
This is a bit off topic, but the VC business is currently stalled. It's not the companies that have died that are the problem. They've been liquidated and require no further attention. It's the thousands of companies left over from the dot-com boom that aren't going anywhere. They're not successful enough to ever pay off the money invested in them, but they bring enough cash to meet their current expenses. (VA Systems, Slashdot's parent, is an excellent example.) There are thousands of VC-financed companies still in lingering-death mode. Much VC time goes into tracking these companies and trying to work out exit strategies.
It's hard to value these walking dead. Some of them might, someday, be worth something. They tend to be carried on the books for a value more appropriate to their glory days, since no big event has occured to force a repricing to market value. This overvalues private equity portfolios, like Stanford's.
Enough accounting for one night.
During the dot-com boom, Stanford was getting about a 20% return on the endowment, and they got carried away. Then when the market tanked, they started hitting on us alumni for more money.
I don't see that in the stated policies.
There are real issues, but the article doesn't address them.
Tree-based databases are thirty years old. See MUMPS. Explictly linked databases are also thirty years old. See the CODASYL DBMS. XML database enthusiasts need to read up on those old systems to avoid making the same mistakes.
Relational databases aren't enough, either. When you find yourself putting columns of serial numbers in tables so you can link tables together, the relational model isn't fitting the problem.
These issues are not being addressed all that well.
MetroCard stores the data both on the card and in a database. They're crosschecked every time a card is used. The data on the card is encrypted, of course. Cards are checked by the station computer and by a central database. Cards are read both at entry and exit from the subway, and if entries and exits don't match, it's noticed. There are protections against fraud by insiders. This system was intended to be Mafia-proof.
So far, so good. No big frauds so far.
And it doesn't bother them.
I've talked with teenagers about what it means when their cell phone has GPS. They're not bothered by having their location reported. They like the idea of knowing where all their friends are. Then they'd know who's nearby, and could hook up. It's a feature.
Lois Bujold's best work is the Lord Vorkosigan series. She won a Hugo for one of those in 1992. But she has to pay the rent, so she cranks out those fantasy novels. The Vorkosigan series is too complex and unsettling for many readers.
Wrong. Once experiment confirmed special relativity, it was widely accepted. Rather quickly, too. Einstein published the key paper in 1905, and by 1910, the physics community accepted it. Definitely not "slow and progressive". Even though Einstein was a nobody at the time. He was still working at the Patent Office.
In the hard sciences, progress is made by experimental confirmation. In areas where theoretical change is "slow and progressive", there's not much progress. Change, yes, progress, no. Consider psychiatry.
There's definitely a "geek factor" here.
Miami looks OK; the storm is hitting further north.
NASA KSC appears to be offline.
At 1333, the cops have the beaches cleared, and there's just one emergency vehicle. A few good waves, but not much happening.
This, incidentally, is one of the problems with paper ballots. In Chicago, the way it worked was that the ward heeler stood outside the polling place. He'd voted early, but didn't put his ballot in the box. Instead, he kept it and marked it. As he paid off voters, he'd give them a pre-marked ballot to deposit. When they came back with a blank one, they were paid.
This sort of thing is rare today. Those old-line regular party organizations are dead. Nobody has enough people in place any more to pull that off. Effort today goes into sucking up to donors and buying ads.
The best camera is the Miami Beach ultra high resolution panoramic webcam. 8000 x 2320 pixels.
The American Locomotive Company built three packaged nuclear plants for the US Army between 1958 and 1962. One was installed at Fort Greely, AK, one at Camp Century in Greenland, and one in the US as a test and training unit.
They worked, but weren't useful enough to be worth the trouble.
USS Thresher and USS Scorpion were lost at sea. USS Guitarro sank alongside a pier during construction for reasons that can only be described as Really Dumb, but was refloated and repaired.
No US subs have been lost since the 1970s, though.
That's a violation of the "Certified for Windows XP" logo rules. The vendor should be denied the use of the Windows logo.
Why bother? DVD-sized media with more capacity are coming, and compression good enough for HDTV on existing media is already available.
Besides, hard disk replication would be a pain. We'd be back to where things were in the early days of VHS replication. Huge farms of consumer grade VCRs slaved off one good digital tape player. A bunch of people running around changing media. (Later, better ways to duplicate tape were developed.) CD/DVD replication is a high-volume process that's incredibly cheap in volume.
This may affect the DRM issues surrounding inkjet cartridge refilling. Anybody know?
They must really hate to lose the Labor Day weekend revenue.
First, in the US broadband passed modems last month. The trend is steady and that number should pass 80% within two years.
Second, because the US has free local calling, good line quality, and plenty of telco switch capacity, dialup works well in the US. In many countries, dialup involves per-minute costs, and you can't stay on all day. It the US, it's been flat-rate monthly for years. And dial-up is really cheap.
Third, more people in the US have Internet access than buy books or subscribe to newspapers. The literate fraction of the population is already on line. If you can't read, even AOL isn't useful.
What's the problem?
A real one-time pad key generator using current technology would have a hardware device like a noise diode that generated random bits at a high rate and created two identical DVDs. That would be useful. With 4GB of random bits, you could encrypt about 5000 minutes of phone-grade voice.
SCOX has been in the $3.50 to $4 range all week, with light trading. Six months ago, it was at $14.
Next court date: September 15, Judge Kimball. Motions in both the Novell vs. SCO and SCO vs. IBM cases will be heard. SCO has been stalling to put off this date, but it looks like the judge won't tolerate any more delay and some major issues will be decided in two weeks.
-
Most players are made in China.
-
The Chinese government wants to reduce dependencies on foreign technology that requires royalties.
-
With players selling for as little as $29.95, paying royalties to high-wage countries is no longer competitive.
-
The top-grossing movie this week is
Hero.
It's from China.
It no longer matters what Microsoft or Hollywood wants. EVD players will be in Wal-Mart.A good engine programmer should have working familiarity with a great many algorithms--so many that attempting to list them here would be silly. The most necessary algorithms perform tasks like spatial partitioning, clustering, and intersection and clipping of geometric primitives. Most algorithms will be mainly focused on one task area, like rendering or physics, but these algorithms are often very deep and take a while to master. For years we have been mining academic research to find and modify appropriate algorithms. However, a game engine must meet soft realtime requirements, and most academic work in the relevant subject areas is geared toward batch computation. (Most of the past research in graphics has applied to offline cinematic rendering. Most physics algorithms are unstable and can fail outright, which is solved in a batch setting by tweaking the initial conditions and trying again. These algorithms do not adapt successfully to a soft realtime setting.) As games are now starting to be taken seriously by the academic community, this is beginning to change, but most academic research is still pointed in directions that don't do us much good. So, creating a technically ambitious game engine will often require a substantial amount of original research.
Hacking isn't enough any more. You need theory. Lots of theory.
Game development has some really smart people. I've met some of them. The best theoretical work today is being done in game development. Game AI is better than academic AI, and it's gaining.