Looking at the bigger picture, this whole episode may be a good thing for Linux no matter what the outcome. Today, many people assume that Linux is worthless or a toy because it is cheap or free.
SCO is working very hard to convince people that even the small subset of Linux that they claim was copied is worth $billions. PHBs everywhere will begin to figure that if there's such a ruckus being raised over this OS, it must be worth using.
In fact most computers sold today come with Office pre-installed. There is no reason for these people to pirate the software nor install workalikes like OpenOffice.
You're saying that as if it's free or something. Check out a typical midrange Dell desktop. Base price with WordPerfect = $1239. Cheap MS Office adds $150. MS Office Pro adds $350.
How many potential pirates or Open Office users are going to add $150 or $350 to the price of their new computer so they can get a bundled copy of Microsoft Office? Not many. That's why Microsoft realizes that the most that they can get in this situation is to promote customer lock-in with a freebie.
I think that these base-2 units are confusing because their names look too much like the base-10 units and because their magnitudes are too close.
A better way would be to invent an all new imperial-style system for measuring computer storage. That way, there would be no chance for confusion with any base-10 system. For example:
This system easily covers storage capacities up to today's confusingly named "petabyte". Plus, there's no ambiguity about what you're measuring. Any of these units implies bytes of storage, which is a much cleaner solution.
The computer I'm using now has 71+29/32 watzes of system memory and 44+10/16 spoffs of disk space. There's no confusion about fuzzy definitions of "mega" with that measurement.
Flush shit down toilet -> let shit mellow at sewage plant -> strain shit residue out of bottom of sewage vat -> haul to field -> spread on grass -> grass grows -> cow eats grass -> pull cow's udder, direct milk into bucket -> ferment milk to cheese -> shred cheese -> spread on dough -> Pizza!
Their alternative if they can't sustain massive growth is increasing salaries and bonuses, and that will cut dramatically into their profit margins, which certainly will further damage their share price.
Yeah, if they had to give everyone a huge raise, the profit margins might nosedive below 82%, maybe even below 79% *shudder*. If it ever came to that, they'd be better off just throwing in the towel.
the fusion splicer is a cheap $35,000.00US and can be destroyed easily.
But most of that cost is for the flux capacitor, which is usually salvageable. Anyway, if you use reasonable care selecting fusion fuel, you probably won't ruin your splicer.
Re:OSI Papers notwithstanding...
on
OSI vs SCO
·
· Score: 1
Their line...go with the smart, non-litigated choice...Windows XP. Now with Software Assurance!
Even if everyone on Earth were ready to migrate to IPv6, it still wouldn't happen. As we found out in Independence Day, IPv4 is an entrenched *intergalactic* standard. There are just too many star systems involved to be able to roll out this upgrade in the forseeable future.
The "synchronization is slow" myth is a very dangerous one, because it motivates programmers to compromise the thread-safety of their programs to avoid a perceived performance hazard.
However, the article perpetuates another myth: "Synchronization should be easy. The more things you synchronize, the better off you are."
My hard experience says otherwise. First off, making multithreaded programs work correctly is very hard. Therefore, multiple threads should be avoided if at all possible. You can avoid a lot of these problems in many cases if you use a function like "select()" in a single-threaded program (which, IIRC, Java unfortunately doesn't support). Even though it looks harder to program, it ends up being easier to debug.
However, sometimes you just can't avoid threads. IMHO, adding "synchronize" as a language keword and encouraging easy creation of threads was a mistake. That doesn't begin to solve your problems. For example, it does nothing to help you avoid deadlocks. In fact, sprinkling synchronized blocks around your program is a recipe for deadlocks and unexpected timing-dependent buggy behavior.
If you must use multiple threads, there should be one main thread that runs almost all of the program's logic, and a set of highly constrained, carefully controlled worker threads. These threads should not interact with any other (mutable) data structures in the program. Ideally, there should be at most two synchronization points in the program: a work queue and a results queue. The elements of these queues should package up all of the state needed for a worker thread to solve a piece of a problem or deliver its results.
With an approach like this that has minimal synchronization, there's no need to add a keyword to the language or put synchronization into many library container classes. And of course, performance is hardly an issue at all when you only synchronize twice per worker thread run.
They should just license the Soyuz from the Russians and launch them from Florida. They cost something like $20 million per launch, and they are probably the most reliable and cost-effective launch system in existence. The article mentions a $6 billion low-end limit on developing a new shuttle. That would pay for 300 Soyuz launches without even factoring in a per-launch cost for the shuttle. But of course, this will never happen due to the NIH factor.
the 10-year old BMW 520i model that "suffered a simple electronic failure".
That reminds me of the time I had circa 1991. I was out shopping for cars, and I thought I would have a look at the BMW 5 series. They had one in a color I liked, and I asked if I could look inside. This one had just come in, and its battery was dead. Turns out, the car had some kind of all-electronic door locks, and there was no mechanical way to unlock the car. The battery would have to be charged first; there was a plug under the bumper to do that in just this situation. Here we were at the BMW dealer, and they couldn't get inside their own car. Not good.
The rising cost of programming, especially rights fees that networks pay sports leagues to broadcast games, means that networks lose money by putting their shows on broadcast stations instead of cable, the networks say. "Sports content will be the first to go to cable," Karmazin warned, noting that CBS paid $6 billion to broadcast the NCAA men's basketball tournament for 11 years. "Then other [programming] will follow."
So what. Maybe they should think about not paying people $5 million per year to play a kid's game or $1 million per episode to act in a sitcom. Get to the root of the "rising cost of programming" problem rather than begging the government to let them consolidate programming into even more homogenized crap than it already is.
This is how it's supposed to work. If overpaid actors, sports players and coke-sniffing media executives are making the industry unprofitable, then companies go out of business, supply of overpaid job openings dwindle, and costs fall back into line. We don't need to put all of a scarce public resource in control of a single private party just to prop up a bad business model.
I only wasted about 5 months of my life
on
AI Going Nowhere?
·
· Score: 1
Back in the 80's I studied AI in grad school. It was only a few short months before I found out that the emperor had not clothes. Basically, the program could have been more accurately called "how to hack around the limitations of Lisp so that you can get it to do things Pascal does easily". They focused on Lisp minutia because it turns out they had zero insight on what actually constitutes "intelligence".
I quickly figured out that you're never going to approach the power of the human brain (which is a trillion-element semi-analog fully associative storage system) by shoveling a few digital bytes at a time through a memory bus and an accumulator register. I believe that assesment is still true today.
I dropped out the first year and never looked back. It was one of the best decisions I ever made.
Indeed; let's make a comparison with of storage technology pricing. I can buy a 512MB RAM module for around $50. This solid state storage supports random access to any location within nanoseconds, hundreds of megabytes per second bandwidth, and can be rewritten an unlimited number of times.
I can buy a 700MB Top-40 CD for $16.99. While the cost per byte is only 25% of that for DRAM, the CD has many offsetting disadvantages. It is an optical disk that requires a dedicated hardware drive to access. Random access time is many milliseconds (10 or more seconds for the first access), and only supports a couple of megabytes per second bandwidth. It can't be written to at all. But the worst part is that this storage is indelibly encoded with an hour of unlistenable audio crap.
In short, compared to the DRAM, the CD is utterly useless. Yet somehow it commands a price per byte almost 1/4 of the DRAM. It's also intersting to note that about a dozen years ago, the DRAM would have cost about $500000 (a 10000:1 price drop), but the CD was still about $16.99. Somehow, the CD is immune to cost reduction. This is real price fixing.
IE is already loaded in memory, hence the slow startup time. Mozilla doesn't have this feature on linux I think.
That feature rocks. I recently got SuSE 8.2, and I figure KDE must now be pulling the same trick as IE. Konqueror always used to be about twice as fast launching as Mozilla, but now it blows it away. On my Athlon 1800+, Mozilla takes about 2.5 seconds to pull up (the second time, with its pages already swapped into memory). Konqueror launched from a terminal is a little over 1 second, but Konqueror launched from any KDE GUI control is less than 0.1 seconds; I can't even perceive the delay. I've got a system-wide hotkey to pull up the browser, and it's there before I can pull my finger off the key.
The LCD sub-pixel font rendering also works in Konqueror. If Mozilla has it, I can't find where to turn it on. The bar keeps getting raised; Mozilla used to look jaggy, now it looks fuzzy.
I'm also a total drinker of the KDE integrated file manager-browser koolaid.
IMHO, the only advantage Mozilla has is that some web designers actually test against it, so it comes in handy on the 3% of websites that barf on Konqueror.
That's not usually the problem. The problem is that Un*x refuses to unmount a drive with open files. It's compounded by the tendency of things like KDE to hold on to open handles for a while after a program is "closed". Or maybe the cdrom was the current directory when you started some unrelated program in a terminal hours ago.
You're going to have to teach the secretary how to use "lsof/dev/cdrom" to track down which processes won't let go of the CD.
This is one case where Windows does it right. It just unmounts the CD, and further attempts to access it by programs result in an I/O error. This behaviour is much more intuitive and convenient. Usually, if an I/O error occurs, a dialog pops up and you can figure out that you need to reinsert the CD in question if you really need that program to keep running.
Silenty disabling the physical eject button on the cdrom drive is one of those things that can make users want to smash their system with a sledgehammer. It is not a way to encourage people to migrate to Linux.
While it may make programming more consistent, it is a usability bug and should be fixed.
PETA would never put there name on something to harm the animals, especially the fuzzy ones.
That's why they're calling this a "petawatt laser". They're trying to get some buy-in from PETA on this scheme. The animal rights group is holding out for the more explicit moniker "PETA/Laser", though.
I, personally, would consider not crashing to Earth a rather important intrinsic value of those miles...
It's an important value, but it is an extrinsic value. Not crashing into the earth is the primary goal; going around in circles isn't the primary goal.
You went places and did things at those places on each car trip. For most shuttle missions, there is no point in going round and round the world. If it were physically possible to hover in zero-g at a stationary point 200 miles above the earth, they would do that instead. It would require 80% less energy to get there than to go into orbit.
Travelling millions of miles in orbits at 18,000 mph is just a dangerous energy-wasting implementation detail that we unfortunately have to put up with due to physics. It is not the main point of the vast majority of missions. (Even if your mission involves scanning large parts of the planet, it would be better to move at a much more leisurely pace when doing so.)
It just doesn't make sense to count the miles traveled in orbit as having some kind of intrinsic value. (Other than maybe that time-dilation experiment they did with the atomic clock on board.)
Re:The price of exploration
on
Shuttle Politics
·
· Score: 3, Insightful
Actually, broken down on passenger miles, it's the safest way to travel, on or off this planet...
Only if you count the useless miles that the shuttle takes on its circuitous route. If you cancel those out, you're left with 200 miles up from the Florida launch pad, and 200 miles back down to the Florida landing strip. For this 400-mile round trip, the odds are pretty dismal, even compared to medieval seafarers.
Yes, but why was the particular key-combination of Alt+Ctrl+Delete chosen instead of a more sandwich friendly (and more user-with-disability-friendly) sequence such as plain-old SysRq? Obviously because someone in Redmond thought that they would outsmart people with DOS-based bootable trojans.
They thought that the risk of DOS-based programs was so great that it was worth retraining every user to use what used to be an obscure last-resort system reset hack as the first command every time they sit down at their system.
Slashdot Mirror
SCO is working very hard to convince people that even the small subset of Linux that they claim was copied is worth $billions. PHBs everywhere will begin to figure that if there's such a ruckus being raised over this OS, it must be worth using.
Here's a simple script that you can pipe your network traffic through to solve most problems:
s:<HTML><HEAD><TITLE>Slashdot.* </HTML>: <HTML><BODY>GET BACK TO WORK, NOW!!!!</BODY></HTML>
No. Given that they're in India, they'll probably be able to do research and development rather cheaply.
You're saying that as if it's free or something. Check out a typical midrange Dell desktop. Base price with WordPerfect = $1239. Cheap MS Office adds $150. MS Office Pro adds $350.
How many potential pirates or Open Office users are going to add $150 or $350 to the price of their new computer so they can get a bundled copy of Microsoft Office? Not many. That's why Microsoft realizes that the most that they can get in this situation is to promote customer lock-in with a freebie.
A better way would be to invent an all new imperial-style system for measuring computer storage. That way, there would be no chance for confusion with any base-10 system. For example:
korb = 3 bytes
fleb = 12 korbs
splin = 20 fleeb
fnit = 6 splins
Fnit = 6000 splins
frush = 48 fnits
watz = 18 frushes (19.5 frushes in the U.K.)
spoff = 480 watzen
nurm = 320 spoffs
long nurm = 80 nurm
munnel = 24 long nurm
This system easily covers storage capacities up to today's confusingly named "petabyte". Plus, there's no ambiguity about what you're measuring. Any of these units implies bytes of storage, which is a much cleaner solution.
The computer I'm using now has 71+29/32 watzes of system memory and 44+10/16 spoffs of disk space. There's no confusion about fuzzy definitions of "mega" with that measurement.
Here's how:
Flush shit down toilet -> let shit mellow at sewage plant -> strain shit residue out of bottom of sewage vat -> haul to field -> spread on grass -> grass grows -> cow eats grass -> pull cow's udder, direct milk into bucket -> ferment milk to cheese -> shred cheese -> spread on dough -> Pizza!
Hey, didn't you "research" that verse from Tom Lehrer?
Yeah, if they had to give everyone a huge raise, the profit margins might nosedive below 82%, maybe even below 79% *shudder*. If it ever came to that, they'd be better off just throwing in the towel.
But most of that cost is for the flux capacitor, which is usually salvageable. Anyway, if you use reasonable care selecting fusion fuel, you probably won't ruin your splicer.
In BSA audited enterprise, vendor litigates YOU!
Even if everyone on Earth were ready to migrate to IPv6, it still wouldn't happen. As we found out in Independence Day, IPv4 is an entrenched *intergalactic* standard. There are just too many star systems involved to be able to roll out this upgrade in the forseeable future.
However, the article perpetuates another myth: "Synchronization should be easy. The more things you synchronize, the better off you are."
My hard experience says otherwise. First off, making multithreaded programs work correctly is very hard. Therefore, multiple threads should be avoided if at all possible. You can avoid a lot of these problems in many cases if you use a function like "select()" in a single-threaded program (which, IIRC, Java unfortunately doesn't support). Even though it looks harder to program, it ends up being easier to debug.
However, sometimes you just can't avoid threads. IMHO, adding "synchronize" as a language keword and encouraging easy creation of threads was a mistake. That doesn't begin to solve your problems. For example, it does nothing to help you avoid deadlocks. In fact, sprinkling synchronized blocks around your program is a recipe for deadlocks and unexpected timing-dependent buggy behavior.
If you must use multiple threads, there should be one main thread that runs almost all of the program's logic, and a set of highly constrained, carefully controlled worker threads. These threads should not interact with any other (mutable) data structures in the program. Ideally, there should be at most two synchronization points in the program: a work queue and a results queue. The elements of these queues should package up all of the state needed for a worker thread to solve a piece of a problem or deliver its results.
With an approach like this that has minimal synchronization, there's no need to add a keyword to the language or put synchronization into many library container classes. And of course, performance is hardly an issue at all when you only synchronize twice per worker thread run.
Bambi's mother gets shot by a hunter! WTF was up with that?
Acually, I haven't seen the new Matrix movie yet, but I was feeling left out because I hadn't written a spoiler warning yet. I feel better now.
They should just license the Soyuz from the Russians and launch them from Florida. They cost something like $20 million per launch, and they are probably the most reliable and cost-effective launch system in existence. The article mentions a $6 billion low-end limit on developing a new shuttle. That would pay for 300 Soyuz launches without even factoring in a per-launch cost for the shuttle. But of course, this will never happen due to the NIH factor.
That reminds me of the time I had circa 1991. I was out shopping for cars, and I thought I would have a look at the BMW 5 series. They had one in a color I liked, and I asked if I could look inside. This one had just come in, and its battery was dead. Turns out, the car had some kind of all-electronic door locks, and there was no mechanical way to unlock the car. The battery would have to be charged first; there was a plug under the bumper to do that in just this situation. Here we were at the BMW dealer, and they couldn't get inside their own car. Not good.
So what. Maybe they should think about not paying people $5 million per year to play a kid's game or $1 million per episode to act in a sitcom. Get to the root of the "rising cost of programming" problem rather than begging the government to let them consolidate programming into even more homogenized crap than it already is.
This is how it's supposed to work. If overpaid actors, sports players and coke-sniffing media executives are making the industry unprofitable, then companies go out of business, supply of overpaid job openings dwindle, and costs fall back into line. We don't need to put all of a scarce public resource in control of a single private party just to prop up a bad business model.
I quickly figured out that you're never going to approach the power of the human brain (which is a trillion-element semi-analog fully associative storage system) by shoveling a few digital bytes at a time through a memory bus and an accumulator register. I believe that assesment is still true today.
I dropped out the first year and never looked back. It was one of the best decisions I ever made.
I can buy a 700MB Top-40 CD for $16.99. While the cost per byte is only 25% of that for DRAM, the CD has many offsetting disadvantages. It is an optical disk that requires a dedicated hardware drive to access. Random access time is many milliseconds (10 or more seconds for the first access), and only supports a couple of megabytes per second bandwidth. It can't be written to at all. But the worst part is that this storage is indelibly encoded with an hour of unlistenable audio crap.
In short, compared to the DRAM, the CD is utterly useless. Yet somehow it commands a price per byte almost 1/4 of the DRAM. It's also intersting to note that about a dozen years ago, the DRAM would have cost about $500000 (a 10000:1 price drop), but the CD was still about $16.99. Somehow, the CD is immune to cost reduction. This is real price fixing.
That feature rocks. I recently got SuSE 8.2, and I figure KDE must now be pulling the same trick as IE. Konqueror always used to be about twice as fast launching as Mozilla, but now it blows it away. On my Athlon 1800+, Mozilla takes about 2.5 seconds to pull up (the second time, with its pages already swapped into memory). Konqueror launched from a terminal is a little over 1 second, but Konqueror launched from any KDE GUI control is less than 0.1 seconds; I can't even perceive the delay. I've got a system-wide hotkey to pull up the browser, and it's there before I can pull my finger off the key.
The LCD sub-pixel font rendering also works in Konqueror. If Mozilla has it, I can't find where to turn it on. The bar keeps getting raised; Mozilla used to look jaggy, now it looks fuzzy.
I'm also a total drinker of the KDE integrated file manager-browser koolaid.
IMHO, the only advantage Mozilla has is that some web designers actually test against it, so it comes in handy on the 3% of websites that barf on Konqueror.
You're going to have to teach the secretary how to use "lsof /dev/cdrom" to track down which processes won't let go of the CD.
This is one case where Windows does it right. It just unmounts the CD, and further attempts to access it by programs result in an I/O error. This behaviour is much more intuitive and convenient. Usually, if an I/O error occurs, a dialog pops up and you can figure out that you need to reinsert the CD in question if you really need that program to keep running.
Silenty disabling the physical eject button on the cdrom drive is one of those things that can make users want to smash their system with a sledgehammer. It is not a way to encourage people to migrate to Linux.
While it may make programming more consistent, it is a usability bug and should be fixed.
That's why they're calling this a "petawatt laser". They're trying to get some buy-in from PETA on this scheme. The animal rights group is holding out for the more explicit moniker "PETA/Laser", though.
It's an important value, but it is an extrinsic value. Not crashing into the earth is the primary goal; going around in circles isn't the primary goal.
Travelling millions of miles in orbits at 18,000 mph is just a dangerous energy-wasting implementation detail that we unfortunately have to put up with due to physics. It is not the main point of the vast majority of missions. (Even if your mission involves scanning large parts of the planet, it would be better to move at a much more leisurely pace when doing so.)
It just doesn't make sense to count the miles traveled in orbit as having some kind of intrinsic value. (Other than maybe that time-dilation experiment they did with the atomic clock on board.)
Only if you count the useless miles that the shuttle takes on its circuitous route. If you cancel those out, you're left with 200 miles up from the Florida launch pad, and 200 miles back down to the Florida landing strip. For this 400-mile round trip, the odds are pretty dismal, even compared to medieval seafarers.
They thought that the risk of DOS-based programs was so great that it was worth retraining every user to use what used to be an obscure last-resort system reset hack as the first command every time they sit down at their system.