Bank of America's site, including online banking, statement printing, and bill paying, works just fine with Mozilla/Firefox.
Bank of America has one of the best online bill payment systems around. No fees. They'll transfer money to major companies with which they have an arrangement, and mail checks to anyone else you select.
Bank of America is the biggest bank in the United States.
Point this out to any bank that wants you to use IE.
First off, the Nielsen/KRNIC figures for Internet penetration show 62% of the South Korean population using the Internet. The corresponding figure for the US is 69%. The highest value is Sweden, at 77%. Note that this is individuals, not households.
Second, US broadband penetration, as a fraction of Internet users, passed 50% some time this month.
It was at 49.48% last month. That number is climbing steadily at 10-15% per year. Within two years it will pass cable TV penetration, stuck at 66% for years. This number is projected to pass 80% in 2006. We don't need to do anything to encourage broadband; it's already happened.
The US still has a huge population of dialup users. Remember, the US has flat-rate wireline local phone calls; most of the world charges per minute. So dialup is very cost-effective in the US. And 56Kb/s isn't bad for web surfing.
10% of users are still using modems slower than 56Kb/s. Around 4% are still at 14.4Kb/s, a number that's held steady for five years. So there's a small customer base that doesn't feel the need for speed at all. And a big customer base that doesn't want to pay $50 per month for broadband.
Dial-up access starts, after all, under $5 per month.
So there is no "broadband penetration problem".
It's over.
Where the US is behind is in bandwidth for "broadband" users. A sizable fraction of US broadband users have sub-megabit speeds, and very few have enough bandwidth for HDTV.
One of the big features of NT 3.51 is that OpenGL would play together with the windowing system. Microsoft bought Softimage and moved that product over to NT. That was the beginning of the end for SGI workstations. 1998 was the year of the big switch. In 1997, if you went to a major animation shop, everything was SGI. By 1998, the newer machines were NT, and by 2000, the SGI machines were gathering dust.
The X client/server approach has never meshed well with hardware 3D graphics. Retrofitting 3D into Linux has been something of a hack, and it's still not well integrated.
It's an impressive machine. 640 nodes, each of which is an 8-CPU shared memory multiprocessor. The unusual feature is that the machines are interconnected not by a network, but by a brute-force 640 x 640 crossbar switch. Each data path has 12.3GB/s, and 640x2 paths can be going at once. 83,000 separate cables. Aggregate bandwidth about 8TB/s.
It's a reasonable way to make a one-off machine, but a dead end as a commercial product.
I've known quite a number of very good programmers, and worked with some of them. By this I mean people who wrote entire operating systems, major CAD packages, game physics engines, key parts of TCP/IP, and such. I mean people from the original Xerox PARC crowd, Stanford, MIT and CMU. I don't mean people who think they're l33t because they've memorized most of the UNIX command line options.
Few of the really good ones are like the stereotype of the "geek" mentioned here. First, top programmers write well, and have demonstrated this by writing for publication. Second, they have strong theoretical backgrounds. Some are self taught, but are comfortable filling a white board with math. Third, they're not overly attached to a single programming language or operating system. Fourth, they tend to have a sense of aesthetics, and can articulate why something is ugly in an engineering sense, rather than merely grumbling about it.
NT was way ahead of anything else at the time, say, circa 1997. UNIX and graphics were not yet playing well together at that point.
The Mac was still stuck with an unprotected single-thread OS underneath. Sure, you could run Apache on a Linux server, but realistically, that's all you wanted to run on Linux at the time.
What would stop the Federal government from coming in and hiring their staffs to work in a nationally-owned and operated supercomputer center?
We already have a whole chain of federally funded supercomputer centers. San Diego CA, Huntsville AL, Pittsburgh PA, Fairbanks AK., etc. San Diego is generally considered to do serious number-crunching work. Huntsville is installing firewalls for grade schools.
Sounds more like "If we don't do this, nobody will make Inanium motherboards".
Now that Intel has cloned AMD's 64-bit architecture, I'm surprised they're still pushing the Inanium at all.
Actually, the real question is how much longer Microsoft will support the Itanium. Remember when NT supported MIPS, Alpha, PowerPC, and x86? Actually, Microsoft only supports the Itanium in a very limited way. The OS, and a few server side apps, run native. But that's it. The desktop apps all run in emulation, as far as I can find out.
Of course it's really fast when you have the whole band to yourself. You could get 10Mb/s over analog cell phones if you could tie up all 860 channels.
Big deal.
After fifty years of effort, it's clear that chemically-powered launchers are a dead end. Chemical fuels will never get any better. Weight reduction has gone about as far as it can go.
Our launchers are terribly fragile, and not getting any better.
If you could build a spacecraft with the weight budget of a commercial airliner, space travel would be straightforward, craft would be reliable, and the technology would be useful. But chemical fuels are just too weak to do the job.
Chemically fueled rockets are the Zeppelins of the space age. They're big, fragile, and have too little load capacity. They work just well enough that you can delude yourself into thinking the technology can become widely useful. But it just can't happen.
This was well known fifty years ago. NASA, and Apollo, led us down a techological blind alley, trying to improve Kennedy's poll ratings.
Until we get something better than chemical rockets, we should stick to unmanned flights.
Nuclear rockets are quite possible; prototype engines were tested in the 1950s. But a crash would be a major disaster. We still can't do fusion. Nor can we create antimatter efficiently.
But, fundamentally, we have to harness a better power source or we're not going anywhere.
If it had a two-way link with everything and could figure out the configuration of all your gear, that would be something. As it is, you'll spend forever setting it up, and probably won't get it right.
There's a future for a home master control, but slapping a GUI on a one-way remote is a dead end.
If appliances used two-way IrDA links with some sane protocol, you could do more. The remote could talk to anything you point it at,find out what it was and provide a suitable GUI, and get info back from the device. Now that would actually be useful.
$700 and you don't even get the channel guide on the controller?
The tube guys had a handle on that problem. The big UNIVAC tube machines had a "high margin" mode, which increased the voltages on the tubes by about 15%.
Each morning, after powerup, the machine was run in "high margin" for about ten minutes. Any tubes that failed were replaced. Lights on the cabinets indicated tube failures, so replacement was quick.
After that, tube failures during normal operation were almost nonexistent.
The people who run a large archival disk farm (petabytes) tell me that about 7% of the drives fail each year. Detailed field MTBF reports from other sites with thousands of drives would be a valuable asset. Hosting services need that kind of info.
Does anybody track this?
If you have 25,000 disk drives, one of them fails every five hours.
Plutonium is made by transmutation in a nuclear reactor.
Transmutation produces not just plutonium, but a whole range of transuranic elements. These are then separated chemically.
Both uranium and plutonium extraction are very messy processes from a chemical engineering standpoint. They involve highly corrosive materials, including fluorine and acids. During the chemical processing, the corrosives become mixed with radioactive byproducts. So you get liquid mixtures which are both corrosive and radioactive.
Worse, early on, Hanford went in for diluting these liquids with water. "Dilution is the solution" was an early phrase. This gets the corrosiveness down, but now you have huge tanks of low-level radioactive solutions. Some of the cleanup effort today focuses on re-concentrating those solutions so the more dangerous components can be vitrified in glass for long term storage.
And, yes, they did put this stuff in metal drums, which were then buried. And they rusted, then leaked. Now they're being dug up.
Looking at today's SCOX chart, we can see an attempt to support the price at $4.00. Dropped to $3.90, and a buy comes in to bring it back up to $4.
Later, dropped to $3.98, and a quick buy brought it back up to $4. Look at the chart. A clear pattern of support.
Remember that SCOX has an announced stock buyback plan. So the support is, we have to assume, by SCO itself, spending its own money to keep the stock price up. Unsuccessfully. Support at $4 will probably be about as successful as the previous support attempts at $15, $10, and $5.
Report bug on bug reporting system on SourceForge.
Wait a few days.
Explore messages on project message board.
Discover that the developers don't read the bug reporting system. Find appropriate Yahoo group they actually do read. Repost bug.
Wait a few days.
Get reply on message board: "Have you tried this in the beta release?"
Set up CVS to talk to SourceForge. Get sources.
Try to build program. Discover dependencies on specific versions of other projects. Get them.
Wash, rinse, repeat.
Try original problem in latest source. Verify problem.
Reply to "Have you tried this in the beta release"
with "yes".
Wait a few days.
Nothing happens.
Wait some more.
Nothing happens.
Dig into code. Find defect. Fix defect. Verify that bug is gone.
Run regression tests. Discover that regression tests show regression test errors. Run regression tests on released version. See same regression test errors. Read CVS comments to discover that regression tests haven't been updated to match source.
Report fix on message board.
Wait a few days.
Nothing happens.
Write on message board asking for source check-in permission.
Get message that a major rewrite of that section is underway and the developers don't want changes to the old code in that area right now.
Point out that developers haven't done a check-in on that section of code in three years.
Get check-in permission.
Check in fix. Rebuild. Rerun regression tests. Update README. Put message on message board about fix.
Receive bug report from other user who was relying on the broken behavior.
This is why you don't fix bugs in the programs of others.
There's a anti-nicotine vaccine in this family, too. Like the cocaine vaccine, it's a modification of an older vaccine against cholera. It binds to nicotine and prevents it from crossing the blood/brain barrier.
The anti-smoking vaccine is only in phase I testing (safety), while the anti-cocaine vaccine is in phase II (effectiveness). But it looks like both are going to work.
In some ways, the anti-smoking vaccine is more promising. The cocaine vaccine might well be overwhelmed by injected overdoses. That's unlikely to happen with smoking.
That actually exists. There was a networked version of the NES used for onboard airline entertainment systems. Games were downloaded from a server somewhere in the aircraft. It's on some older Virgin aircraft. The user-end unit is in a box under the seat, reducing legroom.
There's a community of people who think markets can solve any problem. It's sort of a libertarian/extropian axis. They keep trying to hang micropayment schemes on everything. Nobody is interested.
The first generation of these schemes included DigiCash, CyberCash, and CyberCoin. Remember?
You don't stretch images; you condense them. The source material must always be higher resolution than the output. Final pass is a very slight Gaussian blur, then an image resize down to the desired scale. This guy is detecting stretched images, which isn't hard.
It's a good robotics problem. As with most industrial robot applications, they've had to develop an end-effector specific to the task.
(The term "end effector" covers more than grippers; an end effector may be a socket wrench or a paint sprayer.) This one really is a gripper, one composed of two flat plates, one longer than the other. The problem is structured enough that a relatively simple gripper will work, but requires good force feedback to pull out one book without disturbing others.
Putting books back is tougher, and will probably
require a more complex system, with one manipulator to open up a space for the book and
another to put the book into place.
Slashdot Mirror
Actually, a user manual for telephones would be useful. A full list of all the # and * functions is hard to find.
Bank of America has one of the best online bill payment systems around. No fees. They'll transfer money to major companies with which they have an arrangement, and mail checks to anyone else you select.
Bank of America is the biggest bank in the United States.
Point this out to any bank that wants you to use IE.
Second, US broadband penetration, as a fraction of Internet users, passed 50% some time this month. It was at 49.48% last month. That number is climbing steadily at 10-15% per year. Within two years it will pass cable TV penetration, stuck at 66% for years. This number is projected to pass 80% in 2006. We don't need to do anything to encourage broadband; it's already happened.
The US still has a huge population of dialup users. Remember, the US has flat-rate wireline local phone calls; most of the world charges per minute. So dialup is very cost-effective in the US. And 56Kb/s isn't bad for web surfing. 10% of users are still using modems slower than 56Kb/s. Around 4% are still at 14.4Kb/s, a number that's held steady for five years. So there's a small customer base that doesn't feel the need for speed at all. And a big customer base that doesn't want to pay $50 per month for broadband. Dial-up access starts, after all, under $5 per month.
So there is no "broadband penetration problem". It's over.
Where the US is behind is in bandwidth for "broadband" users. A sizable fraction of US broadband users have sub-megabit speeds, and very few have enough bandwidth for HDTV.
The X client/server approach has never meshed well with hardware 3D graphics. Retrofitting 3D into Linux has been something of a hack, and it's still not well integrated.
It's an impressive machine. 640 nodes, each of which is an 8-CPU shared memory multiprocessor. The unusual feature is that the machines are interconnected not by a network, but by a brute-force 640 x 640 crossbar switch. Each data path has 12.3GB/s, and 640x2 paths can be going at once. 83,000 separate cables. Aggregate bandwidth about 8TB/s.
It's a reasonable way to make a one-off machine, but a dead end as a commercial product.
Few of the really good ones are like the stereotype of the "geek" mentioned here. First, top programmers write well, and have demonstrated this by writing for publication. Second, they have strong theoretical backgrounds. Some are self taught, but are comfortable filling a white board with math. Third, they're not overly attached to a single programming language or operating system. Fourth, they tend to have a sense of aesthetics, and can articulate why something is ugly in an engineering sense, rather than merely grumbling about it.
NT was way ahead of anything else at the time, say, circa 1997. UNIX and graphics were not yet playing well together at that point. The Mac was still stuck with an unprotected single-thread OS underneath. Sure, you could run Apache on a Linux server, but realistically, that's all you wanted to run on Linux at the time.
We already have a whole chain of federally funded supercomputer centers. San Diego CA, Huntsville AL, Pittsburgh PA, Fairbanks AK., etc. San Diego is generally considered to do serious number-crunching work. Huntsville is installing firewalls for grade schools.
Actually, the real question is how much longer Microsoft will support the Itanium. Remember when NT supported MIPS, Alpha, PowerPC, and x86? Actually, Microsoft only supports the Itanium in a very limited way. The OS, and a few server side apps, run native. But that's it. The desktop apps all run in emulation, as far as I can find out.
Of course it's really fast when you have the whole band to yourself. You could get 10Mb/s over analog cell phones if you could tie up all 860 channels. Big deal.
After fifty years of effort, it's clear that chemically-powered launchers are a dead end. Chemical fuels will never get any better. Weight reduction has gone about as far as it can go. Our launchers are terribly fragile, and not getting any better.
If you could build a spacecraft with the weight budget of a commercial airliner, space travel would be straightforward, craft would be reliable, and the technology would be useful. But chemical fuels are just too weak to do the job.
Chemically fueled rockets are the Zeppelins of the space age. They're big, fragile, and have too little load capacity. They work just well enough that you can delude yourself into thinking the technology can become widely useful. But it just can't happen.
This was well known fifty years ago. NASA, and Apollo, led us down a techological blind alley, trying to improve Kennedy's poll ratings.
Until we get something better than chemical rockets, we should stick to unmanned flights.
NASA had a Breakthrough Propulsion Program from 1996 to 2002, but nobody got a solid, reproduceable result of any value.
Nuclear rockets are quite possible; prototype engines were tested in the 1950s. But a crash would be a major disaster. We still can't do fusion. Nor can we create antimatter efficiently. But, fundamentally, we have to harness a better power source or we're not going anywhere.
Actually, you can get Microsoft to fix bugs, if you buy the higher-priced support options.
If it had a two-way link with everything and could figure out the configuration of all your gear, that would be something. As it is, you'll spend forever setting it up, and probably won't get it right.
There's a future for a home master control, but slapping a GUI on a one-way remote is a dead end.
If appliances used two-way IrDA links with some sane protocol, you could do more. The remote could talk to anything you point it at,find out what it was and provide a suitable GUI, and get info back from the device. Now that would actually be useful.
$700 and you don't even get the channel guide on the controller?
You think I'm kidding? Call Sprint PCS information. Not only do they charge you, they run an ad "For movie listings, times, and more, call...".
And it's not even encrypted end to end.
The tube guys had a handle on that problem. The big UNIVAC tube machines had a "high margin" mode, which increased the voltages on the tubes by about 15%. Each morning, after powerup, the machine was run in "high margin" for about ten minutes. Any tubes that failed were replaced. Lights on the cabinets indicated tube failures, so replacement was quick. After that, tube failures during normal operation were almost nonexistent.
If you have 25,000 disk drives, one of them fails every five hours.
Both uranium and plutonium extraction are very messy processes from a chemical engineering standpoint. They involve highly corrosive materials, including fluorine and acids. During the chemical processing, the corrosives become mixed with radioactive byproducts. So you get liquid mixtures which are both corrosive and radioactive.
Worse, early on, Hanford went in for diluting these liquids with water. "Dilution is the solution" was an early phrase. This gets the corrosiveness down, but now you have huge tanks of low-level radioactive solutions. Some of the cleanup effort today focuses on re-concentrating those solutions so the more dangerous components can be vitrified in glass for long term storage.
And, yes, they did put this stuff in metal drums, which were then buried. And they rusted, then leaked. Now they're being dug up.
Remember that SCOX has an announced stock buyback plan. So the support is, we have to assume, by SCO itself, spending its own money to keep the stock price up. Unsuccessfully. Support at $4 will probably be about as successful as the previous support attempts at $15, $10, and $5.
This is why you don't fix bugs in the programs of others.
The anti-smoking vaccine is only in phase I testing (safety), while the anti-cocaine vaccine is in phase II (effectiveness). But it looks like both are going to work.
In some ways, the anti-smoking vaccine is more promising. The cocaine vaccine might well be overwhelmed by injected overdoses. That's unlikely to happen with smoking.
That actually exists. There was a networked version of the NES used for onboard airline entertainment systems. Games were downloaded from a server somewhere in the aircraft. It's on some older Virgin aircraft. The user-end unit is in a box under the seat, reducing legroom.
The first generation of these schemes included DigiCash, CyberCash, and CyberCoin. Remember?
The old Ford Bronco was popular for this application, because its wheel track is compatible with railroad tracks.
There's a long, long history of Wierd Intermodal Ideas.
Or, as artists say, "work big".
Putting books back is tougher, and will probably require a more complex system, with one manipulator to open up a space for the book and another to put the book into place.