I'm not totally convinced by this argument. After all what does an "iPod" do? Does a "Ford Focus" give you a very sharp river crossing? What on earth has "Google" got to do with searching?
If I see a Dodge Ram, should I dodge it, or ram it?
Google keeps track of IP addresses and blocks which are doing an unusually high number of searches and disables requests from them.
How do I know? Because a friend of mine decided to find out how common all TLAs are (three-letter acronyms) by counting Google hits on each TLA. This was before the Google API, so he did it with good old fashioned HTTP/HTML. It didn't take long for Google to flag him as evil and block access from his IP block.
Sure, Microsoft could find some way around this-- using different enough IP addresses to conceal the source-- but that's more trouble than it's worse. Worse yet, it sets up a cat-and-mouse game and keeps M$ dependent on Google-- when their stated goal is to beat Google at its own game.
I've got a simpler explaination for what the author is seeing. His evidence is based on the fact that some pages being requested exist only in Google's cache. Well, spiders are supposed to do breadth-first searches so they don't hit the same site too often. Microsoft is probably going against data it collected a few weeks ago but hasn't put on its public servers yet. (Why not? Could be lots of things. Maybe they haven't put enough hardware on the front end to support the amount of data they have on the back end. Or maybe they're just slow.)
As much as I'd like to bash M$, there's nothing here that really looks suspicious to me.
Microsoft has the right idea there - ship an XY model first and stick the box / grid etc models in later.
I've been fighting with apps written on an XY model for far too long to let this pass. Sure, they work if you get the fonts, font sizes, non-scrollable window sizes, etc. right the first time for everybody and for ever and ever. But that never happens. You have:
Font substitution when you don't happen to have the requested font, should you be using a non-system font. This is especially important on X-Windows.
Users who change the default fonts and font sizes, should you be using a system font. Usually making the fonts huge to reduce eyestrain. Especially as you get higher and higer resolution monitors.
OS vendors who change system fonts and font sizes because pixels are getting smaller. There's a reason why Microsoft wants to move to Avalon.
Portability. Ever used a Java applet on a Mac? the Mac uses a different default point-to-pixel conversion than Windows, and its widgets are bigger than in Windows. As a result, a lot of applets end up with forms where half the buttons are missing because the frame is too small--and it can't be resized.
Resizing. A few jobs back I was developing Java servlets on Linux, Windows NT, and Solaris. We had to set environment variables (e.g. CLASSPATH). In Windows, this involved typing a 1000-character string into a 20-character field while navigating through about 10 tabs in a window with room for about 4 tabs. You couldn't resize the window, and everything was much smaller than it should have been. Granted, you can do resizing with an XY model, but people rarely bother. With other models, you get usable default resizing for free.
Frankly, unless you're doing an embedded system, XY models are wrong for just about anything. And even with embedded systems you need to make sure you have the right resolution display for ever and ever-- Palm's transition to higher resolution displays isn't as clean as I would like.
The software they're releasing is probably a project they've given up on (since they have the much more developed ViaVoice engine). Instead of letting it rot in a closet like most companies would, they give it away and score an immense amount of geek points in the process.
Actually, no. Speech rec is a pretty big business in call centers, as in "Speak the name of the city which you want to fly to..." That sort of thing is hard to do with touch-tone, and expensive to do with live agents. IBM's has a reputation for being worse than its competitors, but still adequate.
What this does is hit Microsoft's SALT before it gets traction. Right now the big debate in the industry is SALT vs. VoiceXML. SALT is being used for a number of prototypes and toy projects, but hasn't yet gotten beyond the "real servers run Unix" attitude in the industry. (Much like Microsoft's DNA thing several years ago... you couldn't tell from reading the advertisements and trade magazines that M$ wasn't getting any traction.)
In short, from the conversations I've had lately (I was at a speech conference this past week) this is an attack on Microsoft. But the real loosers are likely to be the major speech recognition telephony companies (not M$, which is a niche player) which now have to prove that their tool set is enough better to command the big bucks they've been charging.
As for ViaVoice, that's a much different product. It can handle much more flexible grammars, but it also trains on specific voices. When you're making airline reservations over the phone you don't want to spend five or ten minutes training the speech recognizer. Very different technologies.
Look, Ive used macs since the orignal - that just had 1Mb of ram! - and I always will. I just hope that Jobs starts putting comercial realities ahead of his personal ipod manic agenda and starts putting the boot into Gates at long last.
Just to nitpick, the original Mac had 128k. Six months later or so, they introduced the "Fat Mac" with 512k.
If you're upgrading, look for at least 128 MB of video memory in a card with Direct X 9.0 capability that installs into an AGP slot.
I thought Carmack was a big OpenGL fan. (Maybe the last one in the video game industry.) Why would you need DirectX for Doom? Maybe that's just shorthand for certain shared requirements, such as programmable GPU capabilities.
All of MS's own software is written & tuned for the NT kernel, so switching to a different kernel would mean a rewrite of MS-SQL and so on.
MS SQL Server is based on Sybase, which runs on a variety of Unixes. The GUI portions of SQL Server are Windows-specific, of course, but the basic database is (or at least forked from Sybase as) platform agnostic. And as others have pointed out, that part is Win32-based. Win32 is an API which runs on NT and Windows95/MS-DOS. They probably haven't done a lot of low-level performance tuning on it.
...they could come up with a lamp that would actually last. I've gone through at least 6 InFocus projectors in the last 1.5 years due mainly to bulb failure.
I had a similar problem with my projector (an older SharpVision) destroying my bulbs. Not just burning them out, but warping them. Fortunately, I had the extended warranty and they eventually replaced the main board (which apparently includes the power supply and LCD arrays, since the power plug was new, and the dead pixels had moved.) No problems yet. (Fingers crossed...)
The bulbs on all of these projectors have an estimated lifespan of 1500-2500 hours. That's many, many years the way I watch TV. Unfortunately, the bulbs are expensive and not covered by the warranty-- even if the problem is bulb destruction!
In other words, you got a dud. Just like I did. You need to get it repaired, which will probably cost less than a single bulb, even if they gut it and replace it.
Learn Java. The development tools, documentation, and tutorials are all free online. And it's perhaps the most popular programming language after VB and its ilk. Last I checked, it's more popular than C++ or C#.
One reason to do dual-core rather than better hyperthreading is clock speed. The clock signal needs to propagate across the entire core to keep the circuitry synchronized.
There's a fundamental limitation (the speed of light) and a more practical limitation: the propagation speed of the electrical signal. To boost the propagation speed, you need to boost the voltage, which in turn heats up the chip. And you can't go faster than light.
Let's see if I remember my high school physics. At 3x10^8 m/s (the speed of light), a 1-cm radius circular chip is limited to 1/(2*3x10^6) second propagation (signal travelling from the center and back), or 6*10^6 herz. That's six gigaherz-- and that's assuming a straight line from clock to the farthest point on the chip! I'm not sure how big the chips really are, but we're already halfway to a fundamental physical limit. (A more realistic geometry is to assume the chip is any shape, but 2cm between the farthest points.)
With a multi-cored design, each core can have a separate clock, so the propagation speed is less. Assume we replace your 3Ghz processor with four Pentium II cores in the same space. (Assume it's 1cm diagonal for simplicity.) If you make the (unrealistic) assumption that die size is the only thing that affects clock speed, we can twice the clock speed in each core: 6Ghz, with a 0.5 cm core.
In short, smaller cores allows for faster processors. If they continue to make sprawling, single-core chips, they can't boost the clock speed significantly because the universe won't let them.
The upward swing in the stock may be from too many people shorting it. Short sales are a bet that the stock will go down, but they involve a stock purchase, which is reflected in an increase in the stock price.
He wants to run a CPU-intensive background process, performing a totally straightforward set of calculations, and nothing else. No GUI (beyond a few simple controls to make it play nice), and nothing server-side - sounds like a perfect candidate for anything but Java.
Except that what makes Java slow is its do-it-yourself GUI and run-time object management. Those disadvantages mostly go away if you program Java as if it were C: use and re-use arrays of primative types in preference to short-lived objects, and cluster things together in memory (via arrays) when they are used together. The just-in-time compiler will convert the bytecode into machine instructions on the fly, using information that's not available at compile time.
Personally, I'd say this even sounds like a good candidate for hand-tuned assembly.
Okay, I'll bite. I've got two macs and to Linux boxes here. Each has a different processor architecture (Celeron, Athalon, G3, G4.) I don't know much about hand-tuning for x86, but on the G4 you shove your floats into 128-bit vectors and do your ray tracing in chunks of four floats. The G3 lacks the vector coprocessor, so you'd optimize it in a very different way.
Besides, by writing in Java and not hand-written assembly, he gets to run it on machines like mine, which probably weren't the original target platform. The name of the game in distributed processing is to use as many spare processors as possible, even if some are slow. An hour of work to promote his project is more likely to pick up 10% more (or 1000% more, with \.) CPUs than an hour of hand-tuning assembly is likely to get 10% more speed out of a particular processor.
and quite frankly, i don't want there to be. if we're going to start messing with the names of linux stuff, i vote we put an 'n' in umount and an 'e' in resolv.conf first.
You know, I really want to edit my new eResolv.conf(TM)-- it sounds so internet-friendly and professional!
Nobody in their right mind wants to build a cluster out of machines in desktop/deskside chasses. We've done it once, with the first generation Itanium systems where there was no rackmount option for a 2-way box, and we'll never do it again -- remote management of those machines was and is actively painful.
Like most Unix workstations I've worked with, Macs let you do just about everything from SSH. They also have software support for temperature monitoring and the like. What's painful to do remotely that's specific to desktops?
... its most likely not Microsoft... they tried this game once. When NT 4 was new, they were the proud owners of SoftImage... it didnt work too well for them then... cant see them trying again.
To the contrary, it worked great. Up until that point nobody would take 3D rendering on PCs seriously. Between licensing OpenGL and buying SoftImage, they were able to move a huge portion of the market to NT-- a market which was owned exclusively by SGI. At the time, 3D graphics cards were virtually unheard of, and it was thought that card manufacturers couldn't compete with SGI because (a) all the really smart people already worked for SGI, and (b) you coudn't do fast 3D with a comodity motherboard. (SGI used high-speed switching instead of a lowly PCI bus.)
Admittedly, SoftImage didn't help them put 3D into their low-end offerings, nor did it sell well enough for them to keep it, but it did break up the SGI monopoly and send more people to their cash cow: Windows.
First of all, this TechNote was last updated 12/26/2003. It probably only resurfaced today because someone mentioned it in a \. thread yesterday.
Secondly, you can get 90% of the effect in any JavaScript-enabled web browser by using a mouseover in the status bar. That's not as bad as spoofing in the URL bar, as IE does, but it would likely fool far more geeks than would care to believe it.
You see, humans have lazy eyes and creative brains. The eye can only focus on a small area (which is why eye tracking allows psychologists to tell what word someone is reading) and yet we think we can see everything all at once. Peripheral vision is very good at detecting motion, which compensates quite well in the natural world. However, when a GUI element changes in a predictable way (e.g. the URL changing in the URL bar), our brains tend to be lazy at fact-checking and just fill in the blanks. Thus, even geeks like myself who use the URL bar extensively won't look when we think we know what's there.
There was an interesting usability study once regarding how often people use the status bar in Office-type programs. During the test, at random intervals, a message showed up in the status bar which said something like "There is a $20 bill on the bottom of your chair. If you see this message, you can take the bill." Not a single one of the test subjects took the money.
It's not that he had a blind spot-- he had a price target. The Lisa, introduced a year before the Mac, did have a hard disk, and cost roughly $10k. The whole point of the Mac wasn't a brand-new GUI-- that was introduced with the Lisa-- it was the Lisa GUI in a price range that people would accept. They stripped down everything they could-- display, RAM, storage-- until they got something in the same league as an IBM PC. Then they blew an incredible wad on marketing it. After the Super Bowl commercial, everyone had heard of the Macintosh, but only computer nerds had heard of the Lisa.
That doesn't work if you work freelance or for a small business. If you need potential customers to be able to contact you, you need to do things like post a legible email address in a public place and even have it be guessable, e.g. sales@mydomain.com.
You probably haven't had your domain discovered by spammers yet. They do things like send to a@foo.com, b@foo.com, c@foo.com... zzzz@foo.com. They might keep track of which ones bounce, but I'm not sure. One thing I am sure of is that I've been spammed at addresses I've never posted publicly, and I still get spam from an address that I stopped using around 1997.
What tools do you use to help you decide which movies to see?
Personally, I use MovieLens, a University of Minnesota computer science research project. Mind you, I was one of the people doing research with it back before recommendation systems were popularized (by Amazon, among others.)
The site has been improved a lot since I was at the U, and I still use it to tell me whether or not I should see a new movie.
2. If they haven't seen it on the Internet, they'll learn about it at school. And most likely do it after school. Make sure they know about condoms and safe sex. Practice your "Way to go, son!" speech just in case they get lucky
I'm sorry, but I've gotten spam for stuff I'm too young to learn about-- and I'm 31. Either that or the "pee party" spam was about juvenile potty humor, but I don't think so.
I'm all for open and explicit sex education (it worked for me, and everyone else I know who had it) -- but the sort of fetishes that come in my inbox just because I have a public email addresses are not conducive to healthy sexual development.
Decline, or just higher standards?
on
Human Accomplishment
·
· Score: 4, Insightful
For example, we have 65 playwrights alive today for every one in Elizabethan England. Yet do we have dozens of Shakespeares? The picture is even more stark when the 12,000 members of the screen Writers Guild are taken into account.
Simple reason: increased competition leads to decreased fame. If you have dozens of Shakespeares, no single one of them is going to be as famous as if you have just one. Since the measure of success is necessarily some form of popularity contest, you cannot fairly compare Shakespeare's talent to modern talent. It's like saying Perrier is better than tap water when you tasted the former at noon on a hot day in Death Valley and the latter on a cold and rainy October day in Minnesota.
Nor can you say "I saw a Shakespeare play today and a Tom Stoppard play yesterday, and X is better." We have a culture that primes you for Shakespeare starting before Kindergarden, with all of those "Romeo and Juliet" references, among other things. Stoppard, on the other hand, writes for a post-Shakespeare audience, and builds on Shakespeare. You can't isolate one from the other, nor either from its environment.
Slashdot Mirror
Google keeps track of IP addresses and blocks which are doing an unusually high number of searches and disables requests from them.
How do I know? Because a friend of mine decided to find out how common all TLAs are (three-letter acronyms) by counting Google hits on each TLA. This was before the Google API, so he did it with good old fashioned HTTP/HTML. It didn't take long for Google to flag him as evil and block access from his IP block.
Sure, Microsoft could find some way around this-- using different enough IP addresses to conceal the source-- but that's more trouble than it's worse. Worse yet, it sets up a cat-and-mouse game and keeps M$ dependent on Google-- when their stated goal is to beat Google at its own game.
I've got a simpler explaination for what the author is seeing. His evidence is based on the fact that some pages being requested exist only in Google's cache. Well, spiders are supposed to do breadth-first searches so they don't hit the same site too often. Microsoft is probably going against data it collected a few weeks ago but hasn't put on its public servers yet. (Why not? Could be lots of things. Maybe they haven't put enough hardware on the front end to support the amount of data they have on the back end. Or maybe they're just slow.)
As much as I'd like to bash M$, there's nothing here that really looks suspicious to me.
I've been fighting with apps written on an XY model for far too long to let this pass. Sure, they work if you get the fonts, font sizes, non-scrollable window sizes, etc. right the first time for everybody and for ever and ever. But that never happens. You have:
There's a reason why Microsoft wants to move to Avalon.
Frankly, unless you're doing an embedded system, XY models are wrong for just about anything. And even with embedded systems you need to make sure you have the right resolution display for ever and ever-- Palm's transition to higher resolution displays isn't as clean as I would like.
Wow. I didn't even know they included self-destruct hardware!
Actually, no. Speech rec is a pretty big business in call centers, as in "Speak the name of the city which you want to fly to..." That sort of thing is hard to do with touch-tone, and expensive to do with live agents. IBM's has a reputation for being worse than its competitors, but still adequate.
What this does is hit Microsoft's SALT before it gets traction. Right now the big debate in the industry is SALT vs. VoiceXML. SALT is being used for a number of prototypes and toy projects, but hasn't yet gotten beyond the "real servers run Unix" attitude in the industry. (Much like Microsoft's DNA thing several years ago... you couldn't tell from reading the advertisements and trade magazines that M$ wasn't getting any traction.)
In short, from the conversations I've had lately (I was at a speech conference this past week) this is an attack on Microsoft. But the real loosers are likely to be the major speech recognition telephony companies (not M$, which is a niche player) which now have to prove that their tool set is enough better to command the big bucks they've been charging.
As for ViaVoice, that's a much different product. It can handle much more flexible grammars, but it also trains on specific voices. When you're making airline reservations over the phone you don't want to spend five or ten minutes training the speech recognizer. Very different technologies.
Imagining... imagining... nope, can't do it. Must not be possible.
Just to nitpick, the original Mac had 128k. Six months later or so, they introduced the "Fat Mac" with 512k.
MS SQL Server is based on Sybase, which runs on a variety of Unixes. The GUI portions of SQL Server are Windows-specific, of course, but the basic database is (or at least forked from Sybase as) platform agnostic. And as others have pointed out, that part is Win32-based. Win32 is an API which runs on NT and Windows95/MS-DOS. They probably haven't done a lot of low-level performance tuning on it.
The bulbs on all of these projectors have an estimated lifespan of 1500-2500 hours. That's many, many years the way I watch TV. Unfortunately, the bulbs are expensive and not covered by the warranty-- even if the problem is bulb destruction!
In other words, you got a dud. Just like I did. You need to get it repaired, which will probably cost less than a single bulb, even if they gut it and replace it.
Learn Java. The development tools, documentation, and tutorials are all free online. And it's perhaps the most popular programming language after VB and its ilk. Last I checked, it's more popular than C++ or C#.
One reason to do dual-core rather than better hyperthreading is clock speed. The clock signal needs to propagate across the entire core to keep the circuitry synchronized.
There's a fundamental limitation (the speed of light) and a more practical limitation: the propagation speed of the electrical signal. To boost the propagation speed, you need to boost the voltage, which in turn heats up the chip. And you can't go faster than light.
Let's see if I remember my high school physics. At 3x10^8 m/s (the speed of light), a 1-cm radius circular chip is limited to 1/(2*3x10^6) second propagation (signal travelling from the center and back), or 6*10^6 herz. That's six gigaherz-- and that's assuming a straight line from clock to the farthest point on the chip! I'm not sure how big the chips really are, but we're already halfway to a fundamental physical limit. (A more realistic geometry is to assume the chip is any shape, but 2cm between the farthest points.)
With a multi-cored design, each core can have a separate clock, so the propagation speed is less. Assume we replace your 3Ghz processor with four Pentium II cores in the same space. (Assume it's 1cm diagonal for simplicity.) If you make the (unrealistic) assumption that die size is the only thing that affects clock speed, we can twice the clock speed in each core: 6Ghz, with a 0.5 cm core.
In short, smaller cores allows for faster processors. If they continue to make sprawling, single-core chips, they can't boost the clock speed significantly because the universe won't let them.
So it sounds like every modern car has this problem, but people notice it most with hybrids.
The upward swing in the stock may be from too many people shorting it. Short sales are a bet that the stock will go down, but they involve a stock purchase, which is reflected in an increase in the stock price.
Besides, by writing in Java and not hand-written assembly, he gets to run it on machines like mine, which probably weren't the original target platform. The name of the game in distributed processing is to use as many spare processors as possible, even if some are slow. An hour of work to promote his project is more likely to pick up 10% more (or 1000% more, with \.) CPUs than an hour of hand-tuning assembly is likely to get 10% more speed out of a particular processor.
eLinux-eServer# vi /eTc/eResolv.conf\(TM\)
Like most Unix workstations I've worked with, Macs let you do just about everything from SSH. They also have software support for temperature monitoring and the like. What's painful to do remotely that's specific to desktops?
To the contrary, it worked great. Up until that point nobody would take 3D rendering on PCs seriously. Between licensing OpenGL and buying SoftImage, they were able to move a huge portion of the market to NT-- a market which was owned exclusively by SGI. At the time, 3D graphics cards were virtually unheard of, and it was thought that card manufacturers couldn't compete with SGI because (a) all the really smart people already worked for SGI, and (b) you coudn't do fast 3D with a comodity motherboard. (SGI used high-speed switching instead of a lowly PCI bus.)
Admittedly, SoftImage didn't help them put 3D into their low-end offerings, nor did it sell well enough for them to keep it, but it did break up the SGI monopoly and send more people to their cash cow: Windows.
Secondly, you can get 90% of the effect in any JavaScript-enabled web browser by using a mouseover in the status bar. That's not as bad as spoofing in the URL bar, as IE does, but it would likely fool far more geeks than would care to believe it.
You see, humans have lazy eyes and creative brains. The eye can only focus on a small area (which is why eye tracking allows psychologists to tell what word someone is reading) and yet we think we can see everything all at once. Peripheral vision is very good at detecting motion, which compensates quite well in the natural world. However, when a GUI element changes in a predictable way (e.g. the URL changing in the URL bar), our brains tend to be lazy at fact-checking and just fill in the blanks. Thus, even geeks like myself who use the URL bar extensively won't look when we think we know what's there.
There was an interesting usability study once regarding how often people use the status bar in Office-type programs. During the test, at random intervals, a message showed up in the status bar which said something like "There is a $20 bill on the bottom of your chair. If you see this message, you can take the bill." Not a single one of the test subjects took the money.
--
Friendster has a new direction.
It's not that he had a blind spot-- he had a price target. The Lisa, introduced a year before the Mac, did have a hard disk, and cost roughly $10k. The whole point of the Mac wasn't a brand-new GUI-- that was introduced with the Lisa-- it was the Lisa GUI in a price range that people would accept. They stripped down everything they could-- display, RAM, storage-- until they got something in the same league as an IBM PC. Then they blew an incredible wad on marketing it. After the Super Bowl commercial, everyone had heard of the Macintosh, but only computer nerds had heard of the Lisa.
Hard disks were *expensive* at the time.
That doesn't work if you work freelance or for a small business. If you need potential customers to be able to contact you, you need to do things like post a legible email address in a public place and even have it be guessable, e.g. sales@mydomain.com.
You probably haven't had your domain discovered by spammers yet. They do things like send to a@foo.com, b@foo.com, c@foo.com... zzzz@foo.com. They might keep track of which ones bounce, but I'm not sure. One thing I am sure of is that I've been spammed at addresses I've never posted publicly, and I still get spam from an address that I stopped using around 1997.
The site has been improved a lot since I was at the U, and I still use it to tell me whether or not I should see a new movie.
I'm all for open and explicit sex education (it worked for me, and everyone else I know who had it) -- but the sort of fetishes that come in my inbox just because I have a public email addresses are not conducive to healthy sexual development.
Simple reason: increased competition leads to decreased fame. If you have dozens of Shakespeares, no single one of them is going to be as famous as if you have just one. Since the measure of success is necessarily some form of popularity contest, you cannot fairly compare Shakespeare's talent to modern talent. It's like saying Perrier is better than tap water when you tasted the former at noon on a hot day in Death Valley and the latter on a cold and rainy October day in Minnesota.
Nor can you say "I saw a Shakespeare play today and a Tom Stoppard play yesterday, and X is better." We have a culture that primes you for Shakespeare starting before Kindergarden, with all of those "Romeo and Juliet" references, among other things. Stoppard, on the other hand, writes for a post-Shakespeare audience, and builds on Shakespeare. You can't isolate one from the other, nor either from its environment.