After reading this article I think that history is repeating itself. I've been scoffing at the P4, but now I think that Intel may be laughing at the end.
If you remember then the Pentium Pro came out, people (including me) dissed it because it was years behind schedule, huge, expensive and hot. Actually, its architecture was just ahead of the process technology curve. With a few tweaks, the same CPU core came to dominate the world with the P-II and the P-III.
Looking at the radical changes in the P4, including storing only uOPs in the instruction cache and reserving (currently useless) pipeline stages for speed-of-light cross chip delays, they are planning ahead for future realities. We can think of the current P4 as being like the Pentium Pro, just a short-lived beta release.
The more interesting question is which approach to driving uOPs will win out: P4, Transmeta or Itanium. P4 and Transmeta convert legacy x86 opcodes to internal wide architecture on-the-fly (P4 in hardware, transmeta in software); Itanium makes the compiler generate wide architecture directly. Note that the original pre-translated instruction format (CISC, RISC, Java bytecodes, whatever) is now largely irrelevant.
My view is that in the abstract, Transmeta has the best approach, followed by P4 and Itanium last. This is because the software approach is the most flexible and can even be upgraded in the field. In theory, it could detect and store the individual performance characteristics of each program on a user's machine. Granted, they currently focus on low-power, but if they retargeted their technology at high speed, it could be interesting.
The P4 approach is hardwired, but at least it can adapt to local code characteristics and translate them to the current internal architecture version.
The Itanium exposes low-level chip details to the compiler, and the decisions are cast in concrete from there on out. It doesn't seem very future-proof to me; if the IA64 architecture changes in the future, today's compiled code will suffer.
That's the sort of money Microsoft is asking, is it not?
Except that a RedHat distribution comes with features that are more comparable to an MSDN subscription (which was running something like $2K per year a few years ago; I haven't checked lately). It's too bad that the features are skewed more towards a developer than your typical app user, however. (IIRC, MSDN came with a handy copy of MS Office that could be used for "testing";-)
The good news is that RedHat software probably won't disable itself when it 'expires'.
Personally, I'd probably pay the $19, because I'm too busy/lazy to keep up-to-date on every individual package I use.
TPJ has only 2-3 pages of blah-blah and news of the world. The rest is 100% listings.
This is one of the reasons the TPJ financial difficulties. They had to invest large amounts of money in custom typesetting machines that were outfitted with extra $ and @ characters.
Yes, the Earth is warming in some areas, e.g. Siberia. But, this is totally expected if you look on a geological timescale, vis-a-vis the Ice Age cycle.
It's warming right here in the USA. I don't need a scientist to tell me this; when I step outside the front door it feels hotter over the last 10 years.
10 years is not a geological timescale. On a true geological timescale, a century looks like an instant.
It's likely that 100 million years from now, the effects of the human race plotted against time are going to look indistinguishable from the effects of the comet that wiped out the dinasaurs. Given human nature, I doubt that there's even anything that can be done to avoid that.
At the very least, we all get to witness one of the biggest events in the earth's history, like being a passenger in a huge slow-motion train wreck.
Linux was actually invented in China over 4000 years ago. They feel that they are justified in violating the GPL today because the original ancient Chinese authors/scholars were not properly attributed for their work.
(The original punched-parchment scrolls of Linux were illicitly smuggled out of China by spice traders, and centuries later ended up in the archives of an obscure Finnish museum. It's not clear what happened after that.)
... The
Which brings up a point- although I was an engineering major, I punished myself by taking Business and Technical Writing as one of my humanities electives. I was the only geek in the class, and it turned out to be one of the hardest classes I took.
It was graded by a perfectionist on a non-curved scale, and we were forced to write letters to respond to impossible customer service nightmares. (My favorite: Try to convince a doctor to keep using your baby food brand after he found half of a dead cockroach in it. I'm not kidding.)
Despite this, B&TW was just about the single most valuable course I took. A large part of success in the real world depends on how well you communicate.
Not even in retrospect... I knew it was flawed when I was watching it during its first season. My favorites:
A nuclear waste dummp blows up and sends the moon out of earth orbit at what must be relativistic velocities. The moon is an intact sphere (not blown to plasma), the people weren't crushed by the acceleration, and miraculously, it was aimed exaclty in the direction of numerous interesting planets and aliens.
They are critically short of space shuttles, but manage to lose one per episode. Somehow, next week, n-1 == n again.
Nevertheless, I'd really like to see this series again. Of all of the shows I saw as a kid, it really stood out as one of my favorites.
My Timex/Sinclair ZX81 has made zero noise since 1982. Moreover, the flat membrane keyboard makes less noise than any keyboard on the market today. It still works, too.
Does anyone have a URL for the NetBSD port?
Misleading marketing campaign
on
Review: A.I.
·
· Score: 2
After being pummelled for the last couple of weeks by radio ads for this movie and seeing the trailer a couple of times, I thought that this was supposed to be a sappy, sentimental chick flick. Apparently, that's not the case. Maybe I'll go see it after all.
It looks like the studio spent enormous piles of cash trying to convince me not to see the movie, but they may have been foiled.
A good amount of high frequency changes in the image and it won't even detect your face anymore. The systems are appearance based and appearance based systems have trouble with multiclored high frequeny changing images.
OK, everyone, its time to stock up on KISS and Darth Maul makeup kits!
Let's say you want your users to be able to script your application with the C language.
If you're a pretender, you might write a wimpy interpreted version of C.
If you're a real software architect, however, you'd #include the gcc source tree right into your app, compile the user's code on the fly, pipe the output into a memory-mapped file, and long jump straight into the results. Damn the torpedoes.
Make no compromises on the performance of your users' macros. If they're worthy of your program, they need the speed. If they're not worthy, they'll write some lame pointer bug and dump core; I say good riddance to those lusers.
I've been wondering if using non-root installs is really totally effective, especially on personal machines. For example, what's to stop a user-mode trojan from reading all of my personal user-mode files and then transmitting my valuable data to a server in Elbonia via a web request?
The kernel is untouched, but my security has still been breached.
Why even have "sector"-based addressing when the hard drive is just going to munge the addresses into some other physical layout anyway? It's been sector/cylinder/head compatibility hell for the last twenty years.
Maybe just once they should make the painful switch to a simple flat 128-bit address space and be done with it.
That's enough money to buy a slightly used aircraft carrier, all spent on shiny pastel brochures, magazine ads, etc (with some reserved for campaign contributions), all targeted at clueless PHBs and other decision makers.
Slashdot legions could scream there heads off for centuries and still not get as much exposure as MS is buying.
In cases like this, the decades-long trial can achieve the goals of the prosecutors.
I worked for IBM briefly in the 80's, and I could see that their anti-trust trial was putting a real
drag on their operations. For example, every scribble we jotted down onto a scrap of paper had to be copied
and saved on microfilm in case it was needed as evidence. Things like
this tend to put a damper on productivity. They were also very timid about
enforcing patents during this time, allowing competitors like Compaq to spring up out of nowhere.
Over time, the trial adds enough friction to the business to allow others in the market to catch up.
The real question is, with the new administration, will they keep the trial going.
It looks like they'll go for an early settlement instead, thus
greatly reducing the 'punishment'.
There's nothing I like better than doing things dirt cheap; I still run a 486 for certain tasks. However, your point works best in the third world.
If you install a bunch of 486s running Linux in the local library in the US, its just going to give Linux a bad name and turn people off, especially if see they ever see of a modern computer running Windows. These people won't have the technical context to make a fair comparison.
Likewise, the best way to remotely use a good computer is another good computer; not some POS 486 that should be used as a firewall. (Along with its 16-bit ISA 512K 800x600 graphics card and a fuzzy dim monitor.)
I use a laptop that I got cheap from the Dell factory outlet for this purpose. It is big and heavy with a huge, bright display and a relatively wimpy Celeron processor. I stuck a wavelan card in it and loaded most everything from the Red Hat 7.1 distro on it. No special installation steps needed.
I run web browsing and similar tasks locally (with snappy performance), and I use it as an X-terminal for editing and other tasks that need to use the files on my main machine. It works untethered anywhere in the house, too, due to the wireless network. (Don't forget to use ssh to start X sessions since the 802.11 encryption is "questionable".)
Also, slashcode will soon be released under a shared-source licensing.
Actually, the new release of slashcode is being delayed
pending the port from Perl on MySQL to VisualBasic.NET on IIS.
One notable change will be that all users will log in to slashdot
using their Passport accounts. This will allow everyone
to surf seamlessly between slashdot, hotmail and the MSDN site without
needing to retype their passwords.
Unfortunately, slashcode 2001 will only allow users to post via a special ActiveX rich
text edit control. However, this new control will enable users to embed great streaming multimedia content in their flames and trolls. The port of this control to non-IE6 platforms may be
done by a third party at a to-be-determined date. In the meantime,
Windows XP will now be the/. user's platform of choice.
I don't care if it's legal in Canada or not, it's still stealing!
Maybe you have a point there. I have an idea that will turn this into a win-win situation for everyone:
I've got a dual LNB DirecTV setup and two boxes. Even though it seemed like a good idea at the time, it turns out that i never use the second box. I still pay $4.99/month to enable it, though, mainly because I sunk so much money into the 2nd box and I'd feel like a schmuck if I admitted that it was a mistake (DirecTV doesn't subsidize your 2nd box cost, BTW).
So, in the spirit of cooperation and friendship with our neighbors to the north, I hereby bequeath my wasted $4.99/month towards the amount owed by some 3l337 C4n4dl4n H4X0r who is using this illicit system. If only small fraction of the DirecTV subscribers who underutilize their service will pitch in as well, it would easily cover the fees that should have been paid by the small cadre of pioneering Linux (ab)users who actually implement this setup.:-)
Re:Vertical markets with nice profit margins
on
Compaq Shifts Focus
·
· Score: 2
charges $15,000 for an interfacing PC, and said PC is nothing more than a $700 clone PC.
Maybe it could help to break the total cost into components:
Clone PC: $700...
Interface Logic: $1200...
Doctor uses PC to control medical equipment and save dying patient:
Priceless. Err... I mean $13,100
Anyway, the cost doesn't look to outrageous compared to
an ordinary doctor's office visit, where you pay up to $200
to talk to some guy in a room for 5 minutes. What makes this racket worth the
money is that he's (hopefully) going to tell you that contrary to your worst fears, your symptoms are not in fact life-threatening.
... And on the third day, He didst create the molds and the spores and anaerobic slime amongst all of the spheres throughout the heavens. And He saw that it was good...
Slashdot Mirror
Here's a technical reason: Murphy's law.
Just becuase these two are currently the cheapest doesn't mean that you shouldn't work towards better technoligies for the future.
If you remember then the Pentium Pro came out, people (including me) dissed it because it was years behind schedule, huge, expensive and hot. Actually, its architecture was just ahead of the process technology curve. With a few tweaks, the same CPU core came to dominate the world with the P-II and the P-III.
Looking at the radical changes in the P4, including storing only uOPs in the instruction cache and reserving (currently useless) pipeline stages for speed-of-light cross chip delays, they are planning ahead for future realities. We can think of the current P4 as being like the Pentium Pro, just a short-lived beta release.
The more interesting question is which approach to driving uOPs will win out: P4, Transmeta or Itanium. P4 and Transmeta convert legacy x86 opcodes to internal wide architecture on-the-fly (P4 in hardware, transmeta in software); Itanium makes the compiler generate wide architecture directly. Note that the original pre-translated instruction format (CISC, RISC, Java bytecodes, whatever) is now largely irrelevant.
My view is that in the abstract, Transmeta has the best approach, followed by P4 and Itanium last. This is because the software approach is the most flexible and can even be upgraded in the field. In theory, it could detect and store the individual performance characteristics of each program on a user's machine. Granted, they currently focus on low-power, but if they retargeted their technology at high speed, it could be interesting.
The P4 approach is hardwired, but at least it can adapt to local code characteristics and translate them to the current internal architecture version.
The Itanium exposes low-level chip details to the compiler, and the decisions are cast in concrete from there on out. It doesn't seem very future-proof to me; if the IA64 architecture changes in the future, today's compiled code will suffer.
Except that a RedHat distribution comes with features that are more comparable to an MSDN subscription (which was running something like $2K per year a few years ago; I haven't checked lately). It's too bad that the features are skewed more towards a developer than your typical app user, however. (IIRC, MSDN came with a handy copy of MS Office that could be used for "testing" ;-)
The good news is that RedHat software probably won't disable itself when it 'expires'.
Personally, I'd probably pay the $19, because I'm too busy/lazy to keep up-to-date on every individual package I use.
This is one of the reasons the TPJ financial difficulties. They had to invest large amounts of money in custom typesetting machines that were outfitted with extra $ and @ characters.
It's warming right here in the USA. I don't need a scientist to tell me this; when I step outside the front door it feels hotter over the last 10 years.
10 years is not a geological timescale. On a true geological timescale, a century looks like an instant.
It's likely that 100 million years from now, the effects of the human race plotted against time are going to look indistinguishable from the effects of the comet that wiped out the dinasaurs. Given human nature, I doubt that there's even anything that can be done to avoid that.
At the very least, we all get to witness one of the biggest events in the earth's history, like being a passenger in a huge slow-motion train wreck.
(The original punched-parchment scrolls of Linux were illicitly smuggled out of China by spice traders, and centuries later ended up in the archives of an obscure Finnish museum. It's not clear what happened after that.)
They make the patient sign a contract that forbids them from travelling to California.
It was graded by a perfectionist on a non-curved scale, and we were forced to write letters to respond to impossible customer service nightmares. (My favorite: Try to convince a doctor to keep using your baby food brand after he found half of a dead cockroach in it. I'm not kidding.)
Despite this, B&TW was just about the single most valuable course I took. A large part of success in the real world depends on how well you communicate.
Not even in retrospect... I knew it was flawed when I was watching it during its first season. My favorites:
-
A nuclear waste dummp blows up and sends the moon out of earth orbit at what must be relativistic velocities. The moon is an intact sphere (not blown to plasma), the people weren't crushed by the acceleration, and miraculously, it was aimed exaclty in the direction of numerous interesting planets and aliens.
-
They are critically short of space shuttles, but manage to lose one per episode. Somehow, next week, n-1 == n again.
Nevertheless, I'd really like to see this series again. Of all of the shows I saw as a kid, it really stood out as one of my favorites.Does anyone have a URL for the NetBSD port?
It looks like the studio spent enormous piles of cash trying to convince me not to see the movie, but they may have been foiled.
OK, everyone, its time to stock up on KISS and Darth Maul makeup kits!
If you're a pretender, you might write a wimpy interpreted version of C.
If you're a real software architect, however, you'd #include the gcc source tree right into your app, compile the user's code on the fly, pipe the output into a memory-mapped file, and long jump straight into the results. Damn the torpedoes.
Make no compromises on the performance of your users' macros. If they're worthy of your program, they need the speed. If they're not worthy, they'll write some lame pointer bug and dump core; I say good riddance to those lusers.
The kernel is untouched, but my security has still been breached.
Maybe just once they should make the painful switch to a simple flat 128-bit address space and be done with it.
Here's the difference between /. and MS:
-
Slashdot budget: a couple of $million max
- MS FUD machine budget: Win XP gets $1Billion marketing blitz
That's enough money to buy a slightly used aircraft carrier, all spent on shiny pastel brochures, magazine ads, etc (with some reserved for campaign contributions), all targeted at clueless PHBs and other decision makers.Slashdot legions could scream there heads off for centuries and still not get as much exposure as MS is buying.
I worked for IBM briefly in the 80's, and I could see that their anti-trust trial was putting a real drag on their operations. For example, every scribble we jotted down onto a scrap of paper had to be copied and saved on microfilm in case it was needed as evidence. Things like this tend to put a damper on productivity. They were also very timid about enforcing patents during this time, allowing competitors like Compaq to spring up out of nowhere.
Over time, the trial adds enough friction to the business to allow others in the market to catch up.
The real question is, with the new administration, will they keep the trial going. It looks like they'll go for an early settlement instead, thus greatly reducing the 'punishment'.
If you install a bunch of 486s running Linux in the local library in the US, its just going to give Linux a bad name and turn people off, especially if see they ever see of a modern computer running Windows. These people won't have the technical context to make a fair comparison.
I know that's true, but I think the original quote was made by some general well before helicopters were a big deal. It just sounded cool.
Anyway, a laptop is actually more like a helicopter than an ancient 486 is. :)
Likewise, the best way to remotely use a good computer is another good computer; not some POS 486 that should be used as a firewall. (Along with its 16-bit ISA 512K 800x600 graphics card and a fuzzy dim monitor.)
I use a laptop that I got cheap from the Dell factory outlet for this purpose. It is big and heavy with a huge, bright display and a relatively wimpy Celeron processor. I stuck a wavelan card in it and loaded most everything from the Red Hat 7.1 distro on it. No special installation steps needed.
I run web browsing and similar tasks locally (with snappy performance), and I use it as an X-terminal for editing and other tasks that need to use the files on my main machine. It works untethered anywhere in the house, too, due to the wireless network. (Don't forget to use ssh to start X sessions since the 802.11 encryption is "questionable".)
Actually, the new release of slashcode is being delayed pending the port from Perl on MySQL to VisualBasic.NET on IIS.
One notable change will be that all users will log in to slashdot using their Passport accounts. This will allow everyone to surf seamlessly between slashdot, hotmail and the MSDN site without needing to retype their passwords.
Unfortunately, slashcode 2001 will only allow users to post via a special ActiveX rich text edit control. However, this new control will enable users to embed great streaming multimedia content in their flames and trolls. The port of this control to non-IE6 platforms may be done by a third party at a to-be-determined date. In the meantime, Windows XP will now be the /. user's platform of choice.
Maybe you have a point there. I have an idea that will turn this into a win-win situation for everyone:
I've got a dual LNB DirecTV setup and two boxes. Even though it seemed like a good idea at the time, it turns out that i never use the second box. I still pay $4.99/month to enable it, though, mainly because I sunk so much money into the 2nd box and I'd feel like a schmuck if I admitted that it was a mistake (DirecTV doesn't subsidize your 2nd box cost, BTW).
So, in the spirit of cooperation and friendship with our neighbors to the north, I hereby bequeath my wasted $4.99/month towards the amount owed by some 3l337 C4n4dl4n H4X0r who is using this illicit system. If only small fraction of the DirecTV subscribers who underutilize their service will pitch in as well, it would easily cover the fees that should have been paid by the small cadre of pioneering Linux (ab)users who actually implement this setup. :-)
Maybe it could help to break the total cost into components:
- Clone PC: $700...
- Interface Logic: $1200...
- Doctor uses PC to control medical equipment and save dying patient:
Priceless. Err... I mean $13,100
Anyway, the cost doesn't look to outrageous compared to an ordinary doctor's office visit, where you pay up to $200 to talk to some guy in a room for 5 minutes. What makes this racket worth the money is that he's (hopefully) going to tell you that contrary to your worst fears, your symptoms are not in fact life-threatening.