Very few DOS apps ran correctly on win95. Many win95 apps didn't run on Win98. Microsoft changed their Word format between Office 95 and Office 98 (or 2000, whatever) in a gratuitous manner which prevented old office from reading new office documents. Microsoft repeatedly changed the Win32 spec in trivial ways to introduce incompatibilities with OS/2's Win32 implementation. Visual basic programmers have face repeated forward and backward compatibility problems (some noted VB book authors even quit recommending VB because of this).
Microsoft has no qualms about pissing off their (locked-in) developer community. They've repeatedly broke compatibility in every possible way. Why anyone trusts Microsoft, I'll never know. Microsoft's history (the real history, not Bill Gate's rewritten version) should scare anyone away. I'm guessing that you're either extremely young, extrememly naive, extremely forgetful, or paid by Microsoft (the last one was a joke).
I don't at all understand why you write "an RMS autocracy." RMS is not a dictator. He is, however, a very careful and concerned member of the community. He's the sole parent of the Free Software movement, which explains why he's concerened about what happens in all the projects.
GNU is about certain types of freedom, as carefully laid out in the various manifestos and licenses. I believe your are grossly, erringly oversimplifying the meaning of freedom. The GNU project is not about anarchy, which is what I think you're suggesting.
Fortunately, for my own pride, the "simple piece of software" referred to a small client app on the user's desktop -- which is complicated enough, but simple in the realm of end-user software. I'm happy to leave the "hard" part up to you.
Thanks for the assesment of difficulty. I don't know that I agree with that your scenario above is "abuse". You may charge more for bandwidth than a flat rate provider, but a saavy customer would walk away with a lower bill for the level of service they received.
At any rate, I think I understand better why people give away free radios.;-)
Would it be more reasonable to have the customer dynamically *choose* their bandwidth, and by the act of choosing, generate a bill?
Imagine that there is a rate limiter upstream of my connection. I, as a customer, have a simple piece of software which does two things: 1) sets the limit for my traffic, and 2) generates an "invoice". It would be akin to long distance phone calls -- I know that I'll be billed differently depending on which numbers *I* dial.
That would relieve some of the complication in the billing and enforcement end of things.
Comparing DSL to a T1 soley in terms of bandwidth seems misleading. I believe a T1 comes with availability guarantees, whereas DSL (typically?) does not. I expect that reliability is a non-trivial part of T1 pricing.
On the other hand, if you're buying new equipment, it's no problem to stick to linux-supported equipment. That's how I buy stuff for me, personally, and that's how I buy stuff for the lab I'm working in.
While the term frankenPC makes me laugh due to intimate familiarity with such beasts (esp. with family tech support), linux does run just about any hardware you might want to use for computation and simulation. It's in the "user-interface" areas (i.e. "multimedia") that linux drivers are lacking. Things like sound, 3D graphics, and tv tuners and capture cards are the items for which you must be most careful. Linux has (hard-fought) support for some hardware in each of these areas, just like the Mac. Come to think of it, linux might have as many or more 3D vid card drivers as the Mac does, which seems really strange.
If you pay for MacOS and it doesn't work right because of hardware incompatibilities, you complain to Apple. If you pick up GNU/Linux for free and it doesn't work right because of hardware incompatibilities, you complain to yourself for not having done the research ahead of time. If you buy a machine preloaded with GNU/Linux from Company X and it doesn't work right due to hard incompatibilities, you complain to Company X.
Not to knock MacOS clusters, but I don't think ease of use is a big priority with cash-strapped physicists. In particular, I'm thinking of astronomers. All of the astronomers I've met have a surprising knowledge of copmuter hardware and system administration. I expect this is because they always have to do everything themselves, as they can't afford to hire someone to do it for them.
As a result, the astronomers I've met put together their own machines from parts unless they've recently been given a big grant. I know one who was buliding a beowulf cluster. He didn't go to Microway an ask them to build one, because that would be two expensive. Instead, he looked for the cheapest rack-mountable x86 boxes he could find.
I don't know that all physicists are like astronomers in this area, but I do know that applied physicists as a group are 1) smart, 2) have a basic knowledge of electronic circuitry, 3) build specialized test equipment because nobody provides what they need, and 4) are having a heck of a time finding money these days. This just doesn't sound like an Apple market to me. Especially if the physicist is at a small school.
Speaking of Beowulf expertise, for some reason I seem to think that Donald Becker has a degree in physics.
In fact, I can't think of anyone who really needs a cluster, that doesn't have at least one linux-loving longhair around. That includes the animation studios.
> Care to suggest a "stable" kernel that will run on any machine you throw at it?"
Care to show me a Mac OS kernel that runs on any machine you throw at it? How about one that runs on some particular Alpha config? Or a Power IV config? Or a StrongArm config (thought not for a cluster;-)?
Okay folks. Many people are claiming that nobody would be stupid enough to fall for the (known) SEC fakes. Especially if their fake status becomes "common knowledge". Well, I've got some news for you.
There are people who forward chain letters because they're afraid they'll die if they don't. There are people who believe Amway will make them rich. There are people who think Scientology makes sense. There are people who killed themselves in late 1999 (IIRC) to meet on the backside of a comet.
There are people who believe that Bill Gates will pay them for being part of an email tracking experiment. There are people who believe computer viruses can make your ice cream go all melty (well, maybe that's stretching things...). Some people gave money, of their own free will, to Donald Trump.
And there are people who believed GW Bush had a "clear policy for the Middle East" during the presidential debates.
You could show the mcwhortle site to these people, tell them it's a fake, and they'd still fall for it. While we like to claim that those who write "first post" are the least intelligent creatures on earth, they're already head-and-shoulders above the people the SEC is trying to help. Many of these people are retired, and hoping to get something back from the society they gave 40+ years of work to (like my father). Some of these people abandon all reason when it comes to this hope (unlike my father, thankfully).
The SEC is trying to reduce the number of fraud-related tragedies among these people, and I think it's a good thing. In fact, I think this is one of the coolest things I've seen our government do for the public, ever. The SEC seems to have a clue about real life and real people, unlike the Whitehouse and Congress (no matter who is in residence at the time).
Spec isn't everything. In fact, many changes to Spec have been made over the years at the request of various processor companies. Many of these changes are responsible for making x86 look compareable to RISC architectures.
SPEC suggested that our 21264 EV6 Alphas should perform about 3 times faster on floating point code, and 1.5 times faster on integer code, per clock against Intel P-III. My mixed fp and int codes, however, seem to run at least 5 times faster; sometimes more. Among the reasons is memory bandwidth and architecture.
We bought a dual Alpha '264 machine, using a 264DP motherboard, in early 1999 IIRC. Each cpu has it's own 2.6GP/sec point-to-point link to memory. I'm not sure when the 264DP was debuted, but it's taken the x86 world at least 2.5 years to catch up. Furthermore, that bandwidth was available through highly interleaved PC100 SDRAM in a special form-factor (not sure how many bits per transfer, and hence didn't have the latency problems that plague Rambus. And the caches are huge, fast, and amazing (number of simultaneous outstanding cache-line fills, etc). The x86 world, at least AMD anyway, are starting to catch up.
SPEC just doesn't do a good job of predicting real world performance. Partially this is because of special compiler flags manufactures use when testing; flags which in all honesty should be named "--for-SPEC" (this information comes from compiler developers for a large CPU company). There are many other differences between how the Spec programs are run for spec, and how these programs would be run if doing real work.
To the point: the MIPS core is very good at fp, and I don't give a damn what SPEC says. If you want to consider an SGI, talk to an SGI salesman and have them run *your* code on their machines. Other RISC manufacturers (are there any left? Damn you Mike Capellas!) will do the same (Intel might if you're interested in the Itanic). Then discuss price -- while there's no negotiating to do with Intel, SGI, IBM, and Compaq will negotiate if you take the time to push the price issue. Educational discounts with Compaq can be as high as 50% for certain configurations (I think you can get an ES-40 Model II with 1 cpu, 1GB, and a 36GB hd for $18,000 -- then buy 32 1GB sticks from Dataram for $1800 each;-). Once you've got a *real* price, compute price/performance. I guarantee that x86 will not come out on top all the time (because it never has for us).
Why does everyone watch Amazon.com and Egghead.com and... I've been happily shopping at Bookpool (they're limited to technical books, but I generally am these days, too;-) for 4 years now, with no problems to report *ever*. Computers4sure (which has recently been bought by Office Depot, it seems) has filled many of my computer orders during the last 3 years, again with no problems *ever*.
Both have intelligently designed websites, good prices, good availability and, perhaps above all, reasonably priced and quick shipping. Both of these companies have been in business nearly as long (if not as long -- I'm just using my memory) as Amazon, without resorting to the bizarre expansions and gimicks.
Why we should judge online retailing by Amazon.com is beyond me. I suppose one argument is that Amazon sells a wide range of books, whereas Bookpool specializes in a niche market. I bevelive this is a bogus argument: the internet is all about niche markets, and attempts to use it otherwise are stupid. Television is a way to force your message down every person's throats. The internet is a way for people to find what they *want*.
The niche-nature of the internet has been demonstrated many times, perhaps foremost by Katz when releasing "Run to the Mountain" -- if I remember correctly, he even made it into Amazon's top ten for while. Certainly this didn't happen because of anything Amazon did (except to be known as a book seller, which they're leaving behind). It happened because of Katz's niche marketing.
Who cares if Sears fails online? I'll drive to their store and talk to their staff if I don't know which tool I need, but I need it now or want to see it first. If I know what tool I want, or I don't need it now or don't need to see it first, I'll research tools on the web, and then look for an online vendor with good prices and policies. Is Sears likely to be this online vendor? No. When shopping for computer parts, is CompUSA's online service compelling? No -- their prices are sometimes good, but their shipping is horribly slow (their best is "usu. w/in 2 days", as opposed to "orders by 3PST will be shipped same day").
Is Amazon likely to have the best price on technical books while Bookpool is around? No. Amazon seems to be aiming for "online Walmart" status. However, Walmart succeeded because they could keep their prices low and provide a little of everything (though most of nothing) to geographic communities. Online, geography is irrelevant (except when you ship ground, and if warehouses are distributed then even ground shipping can be fast). As a result, finding a sizeable customer base is easier. That means niche retailers can buy bulk and keep their prices down. Furthermore, niche retailers are likely to carry higher quality goods than Walmart and know more about their products than Walmart.
So why would you go to Walmart, unless they're the only store with what you wanted (which is unlikely)? I expect that if Amazon is still around in 10 years, we'll think of them like Walmart or K-Mart, except less convenient than these stores.
Having read through the responses, it's clear I should have specified *which* Alpha architecure! Some people have had no problem with 2.4 on Alpha, but those using the 264DP claim that 2.4.17 has improved things.
Also, 2.2.21pre1 or pre2 seems to have fixed the compile bugs for Alpha. It appears that I would have gotten in over my head had I tried to fix them myself. =-)
Thanks for sharing your experiences! I just might force 2.4.17 down some poor unsuspecting user's throats soon (inlcuding mine)!;-)
Microsoft has a lot to overcome to stop being the Evil Empire. The problem is that there is nearly no good will, benefit of the doubt, or trust left for Microsoft. They've screwed everyone multiple times. That includes business partners, OEM customers, end-users, you-name-it.
Ballmer said they have a "popularity bug". It's no bug, it's by their own design. They've earned their place in the hall of shame. They want to win everyting, regardless of what's good for the people around them. Some people call that "hardball", but I call it antisocial.
The question, then, is why should we believe Microsoft is really going to change anything? Why isn't this just another publicity stunt? They've lied to everyone many times, including falsification of evidence in a US court of law. If Microsoft magically transfigured themselves into a perfect company today, it would still take many years before I would trust them.
In fact, the name "Unix" is something of a pun (Ritchie described it as a "somewhat treacherous pun";-) on Multics. They weren't supposed to build Unix, but Multics was just to darn big and complicated to be genuinely useful. So they went from Mul (bigness) to Un(i) (small). I think there were also political considerations when making an new operating system while you're supposed to be working on another.
And this guy appears to be talking about only x86 machines. My lab has had a horrible time with 2.4 on Alphas. In fact, we've moved back to 2.2.18 on some macines. (2.2.20 for Alpha didn't compile properly, and I didn't want to mess with it -- anyone know if which 2.2 kernel is best for number-crunching Alphas right now?). Oh, the pain. The lost time. "Kernel of Pain" is a fine description of our 2.4 experience on Alphas.
Some of the testimony in the DoJ and States v MS demonstrated that the price of retail copies of Windows did not drop with age in the same way other software (including OS) products did. Perhaps the overcharging suits will discuss the price paid for Win95 once Win98 was out, or similar.
Thanks for mentioning the Symphony. TA's music was terrific! I sometimes put the game disc in my stereo while I'm working.
One interesting bit of trivia -- the orchestra, the Seattle Symphony, doubled themselves for the recording -- that is, they played through the same passages twice (not necessarily playing the same notes), and mixed the results together. This is what provides the Mahler-esque proportion to the sound.
Slashdot Mirror
Very few DOS apps ran correctly on win95. Many win95 apps didn't run on Win98. Microsoft changed their Word format between Office 95 and Office 98 (or 2000, whatever) in a gratuitous manner which prevented old office from reading new office documents. Microsoft repeatedly changed the Win32 spec in trivial ways to introduce incompatibilities with OS/2's Win32 implementation. Visual basic programmers have face repeated forward and backward compatibility problems (some noted VB book authors even quit recommending VB because of this).
Microsoft has no qualms about pissing off their (locked-in) developer community. They've repeatedly broke compatibility in every possible way. Why anyone trusts Microsoft, I'll never know. Microsoft's history (the real history, not Bill Gate's rewritten version) should scare anyone away. I'm guessing that you're either extremely young, extrememly naive, extremely forgetful, or paid by Microsoft (the last one was a joke).
-Paul Komarek
I don't at all understand why you write "an RMS autocracy." RMS is not a dictator. He is, however, a very careful and concerned member of the community. He's the sole parent of the Free Software movement, which explains why he's concerened about what happens in all the projects.
GNU is about certain types of freedom, as carefully laid out in the various manifestos and licenses. I believe your are grossly, erringly oversimplifying the meaning of freedom. The GNU project is not about anarchy, which is what I think you're suggesting.
-Paul Komarek
He can. But it would be misleading to call it GNOME. Miguel doesn't suggest branching, he's suggesting a direction for GNOME itself.
-Paul Komarek
Fortunately, for my own pride, the "simple piece of software" referred to a small client app on the user's desktop -- which is complicated enough, but simple in the realm of end-user software. I'm happy to leave the "hard" part up to you.
;-)
Thanks for the assesment of difficulty. I don't know that I agree with that your scenario above is "abuse". You may charge more for bandwidth than a flat rate provider, but a saavy customer would walk away with a lower bill for the level of service they received.
At any rate, I think I understand better why people give away free radios.
-Paul Komarek
Would it be more reasonable to have the customer dynamically *choose* their bandwidth, and by the act of choosing, generate a bill?
Imagine that there is a rate limiter upstream of my connection. I, as a customer, have a simple piece of software which does two things: 1) sets the limit for my traffic, and 2) generates an "invoice". It would be akin to long distance phone calls -- I know that I'll be billed differently depending on which numbers *I* dial.
That would relieve some of the complication in the billing and enforcement end of things.
-Paul Komarek
Comparing DSL to a T1 soley in terms of bandwidth seems misleading. I believe a T1 comes with availability guarantees, whereas DSL (typically?) does not. I expect that reliability is a non-trivial part of T1 pricing.
-Paul Komarek
On the other hand, if you're buying new equipment, it's no problem to stick to linux-supported equipment. That's how I buy stuff for me, personally, and that's how I buy stuff for the lab I'm working in.
While the term frankenPC makes me laugh due to intimate familiarity with such beasts (esp. with family tech support), linux does run just about any hardware you might want to use for computation and simulation. It's in the "user-interface" areas (i.e. "multimedia") that linux drivers are lacking. Things like sound, 3D graphics, and tv tuners and capture cards are the items for which you must be most careful. Linux has (hard-fought) support for some hardware in each of these areas, just like the Mac. Come to think of it, linux might have as many or more 3D vid card drivers as the Mac does, which seems really strange.
If you pay for MacOS and it doesn't work right because of hardware incompatibilities, you complain to Apple. If you pick up GNU/Linux for free and it doesn't work right because of hardware incompatibilities, you complain to yourself for not having done the research ahead of time. If you buy a machine preloaded with GNU/Linux from Company X and it doesn't work right due to hard incompatibilities, you complain to Company X.
-Paul Komarek
Not to knock MacOS clusters, but I don't think ease of use is a big priority with cash-strapped physicists. In particular, I'm thinking of astronomers. All of the astronomers I've met have a surprising knowledge of copmuter hardware and system administration. I expect this is because they always have to do everything themselves, as they can't afford to hire someone to do it for them.
As a result, the astronomers I've met put together their own machines from parts unless they've recently been given a big grant. I know one who was buliding a beowulf cluster. He didn't go to Microway an ask them to build one, because that would be two expensive. Instead, he looked for the cheapest rack-mountable x86 boxes he could find.
I don't know that all physicists are like astronomers in this area, but I do know that applied physicists as a group are 1) smart, 2) have a basic knowledge of electronic circuitry, 3) build specialized test equipment because nobody provides what they need, and 4) are having a heck of a time finding money these days. This just doesn't sound like an Apple market to me. Especially if the physicist is at a small school.
Speaking of Beowulf expertise, for some reason I seem to think that Donald Becker has a degree in physics.
In fact, I can't think of anyone who really needs a cluster, that doesn't have at least one linux-loving longhair around. That includes the animation studios.
-Paul Komarek
> Care to suggest a "stable" kernel that will run on any machine you throw at it?"
;-)?
Care to show me a Mac OS kernel that runs on any machine you throw at it? How about one that runs on some particular Alpha config? Or a Power IV config? Or a StrongArm config (thought not for a cluster
-Paul Komarek
Okay folks. Many people are claiming that nobody would be stupid enough to fall for the (known) SEC fakes. Especially if their fake status becomes "common knowledge". Well, I've got some news for you.
There are people who forward chain letters because they're afraid they'll die if they don't. There are people who believe Amway will make them rich. There are people who think Scientology makes sense. There are people who killed themselves in late 1999 (IIRC) to meet on the backside of a comet.
There are people who believe that Bill Gates will pay them for being part of an email tracking experiment. There are people who believe computer viruses can make your ice cream go all melty (well, maybe that's stretching things...). Some people gave money, of their own free will, to Donald Trump.
And there are people who believed GW Bush had a "clear policy for the Middle East" during the presidential debates.
You could show the mcwhortle site to these people, tell them it's a fake, and they'd still fall for it. While we like to claim that those who write "first post" are the least intelligent creatures on earth, they're already head-and-shoulders above the people the SEC is trying to help. Many of these people are retired, and hoping to get something back from the society they gave 40+ years of work to (like my father). Some of these people abandon all reason when it comes to this hope (unlike my father, thankfully).
The SEC is trying to reduce the number of fraud-related tragedies among these people, and I think it's a good thing. In fact, I think this is one of the coolest things I've seen our government do for the public, ever. The SEC seems to have a clue about real life and real people, unlike the Whitehouse and Congress (no matter who is in residence at the time).
-Paul Komarek
"I've seen better technobabble on episodes of ST:TNG."
That's because ST is for geeks. Technical scams are for non-technical chumps.
-Paul Komarek
Well, fix it up if you know how, and submit the result! If you're an American, think of it as volunteering for your country.
-Paul Komarek
Screw the letter of the law. I wan't to see companies prosecuted for violating the spirit of the law, or at least for being antisocial.
-Paul Komarek
My first thought when reading this title (nothing to do with the article:
"Producing Hot Air with Quantum Mechanics."
-Paul Komarek
Spec isn't everything. In fact, many changes to Spec have been made over the years at the request of various processor companies. Many of these changes are responsible for making x86 look compareable to RISC architectures.
;-). Once you've got a *real* price, compute price/performance. I guarantee that x86 will not come out on top all the time (because it never has for us).
SPEC suggested that our 21264 EV6 Alphas should perform about 3 times faster on floating point code, and 1.5 times faster on integer code, per clock against Intel P-III. My mixed fp and int codes, however, seem to run at least 5 times faster; sometimes more. Among the reasons is memory bandwidth and architecture.
We bought a dual Alpha '264 machine, using a 264DP motherboard, in early 1999 IIRC. Each cpu has it's own 2.6GP/sec point-to-point link to memory. I'm not sure when the 264DP was debuted, but it's taken the x86 world at least 2.5 years to catch up. Furthermore, that bandwidth was available through highly interleaved PC100 SDRAM in a special form-factor (not sure how many bits per transfer, and hence didn't have the latency problems that plague Rambus. And the caches are huge, fast, and amazing (number of simultaneous outstanding cache-line fills, etc). The x86 world, at least AMD anyway, are starting to catch up.
SPEC just doesn't do a good job of predicting real world performance. Partially this is because of special compiler flags manufactures use when testing; flags which in all honesty should be named "--for-SPEC" (this information comes from compiler developers for a large CPU company). There are many other differences between how the Spec programs are run for spec, and how these programs would be run if doing real work.
To the point: the MIPS core is very good at fp, and I don't give a damn what SPEC says. If you want to consider an SGI, talk to an SGI salesman and have them run *your* code on their machines. Other RISC manufacturers (are there any left? Damn you Mike Capellas!) will do the same (Intel might if you're interested in the Itanic). Then discuss price -- while there's no negotiating to do with Intel, SGI, IBM, and Compaq will negotiate if you take the time to push the price issue. Educational discounts with Compaq can be as high as 50% for certain configurations (I think you can get an ES-40 Model II with 1 cpu, 1GB, and a 36GB hd for $18,000 -- then buy 32 1GB sticks from Dataram for $1800 each
-Paul Komarek
Why does everyone watch Amazon.com and Egghead.com and ... I've been happily shopping at Bookpool (they're limited to technical books, but I generally am these days, too ;-) for 4 years now, with no problems to report *ever*. Computers4sure (which has recently been bought by Office Depot, it seems) has filled many of my computer orders during the last 3 years, again with no problems *ever*.
Both have intelligently designed websites, good prices, good availability and, perhaps above all, reasonably priced and quick shipping. Both of these companies have been in business nearly as long (if not as long -- I'm just using my memory) as Amazon, without resorting to the bizarre expansions and gimicks.
Why we should judge online retailing by Amazon.com is beyond me. I suppose one argument is that Amazon sells a wide range of books, whereas Bookpool specializes in a niche market. I bevelive this is a bogus argument: the internet is all about niche markets, and attempts to use it otherwise are stupid. Television is a way to force your message down every person's throats. The internet is a way for people to find what they *want*.
The niche-nature of the internet has been demonstrated many times, perhaps foremost by Katz when releasing "Run to the Mountain" -- if I remember correctly, he even made it into Amazon's top ten for while. Certainly this didn't happen because of anything Amazon did (except to be known as a book seller, which they're leaving behind). It happened because of Katz's niche marketing.
Who cares if Sears fails online? I'll drive to their store and talk to their staff if I don't know which tool I need, but I need it now or want to see it first. If I know what tool I want, or I don't need it now or don't need to see it first, I'll research tools on the web, and then look for an online vendor with good prices and policies. Is Sears likely to be this online vendor? No. When shopping for computer parts, is CompUSA's online service compelling? No -- their prices are sometimes good, but their shipping is horribly slow (their best is "usu. w/in 2 days", as opposed to "orders by 3PST will be shipped same day").
Is Amazon likely to have the best price on technical books while Bookpool is around? No. Amazon seems to be aiming for "online Walmart" status. However, Walmart succeeded because they could keep their prices low and provide a little of everything (though most of nothing) to geographic communities. Online, geography is irrelevant (except when you ship ground, and if warehouses are distributed then even ground shipping can be fast). As a result, finding a sizeable customer base is easier. That means niche retailers can buy bulk and keep their prices down. Furthermore, niche retailers are likely to carry higher quality goods than Walmart and know more about their products than Walmart.
So why would you go to Walmart, unless they're the only store with what you wanted (which is unlikely)? I expect that if Amazon is still around in 10 years, we'll think of them like Walmart or K-Mart, except less convenient than these stores.
-Paul Komarek
Hmm, I guess I'd sort of hope that gcc supports the PDP-11. You'd think someone would have done it already.
-Paul Komarek
Having read through the responses, it's clear I should have specified *which* Alpha architecure! Some people have had no problem with 2.4 on Alpha, but those using the 264DP claim that 2.4.17 has improved things.
;-)
Also, 2.2.21pre1 or pre2 seems to have fixed the compile bugs for Alpha. It appears that I would have gotten in over my head had I tried to fix them myself. =-)
Thanks for sharing your experiences! I just might force 2.4.17 down some poor unsuspecting user's throats soon (inlcuding mine)!
-Paul Komarek
Microsoft has a lot to overcome to stop being the Evil Empire. The problem is that there is nearly no good will, benefit of the doubt, or trust left for Microsoft. They've screwed everyone multiple times. That includes business partners, OEM customers, end-users, you-name-it.
Ballmer said they have a "popularity bug". It's no bug, it's by their own design. They've earned their place in the hall of shame. They want to win everyting, regardless of what's good for the people around them. Some people call that "hardball", but I call it antisocial.
The question, then, is why should we believe Microsoft is really going to change anything? Why isn't this just another publicity stunt? They've lied to everyone many times, including falsification of evidence in a US court of law. If Microsoft magically transfigured themselves into a perfect company today, it would still take many years before I would trust them.
-Paul Komarek
In fact, the name "Unix" is something of a pun (Ritchie described it as a "somewhat treacherous pun" ;-) on Multics. They weren't supposed to build Unix, but Multics was just to darn big and complicated to be genuinely useful. So they went from Mul (bigness) to Un(i) (small). I think there were also political considerations when making an new operating system while you're supposed to be working on another.
-Paul Komarek
And this guy appears to be talking about only x86 machines. My lab has had a horrible time with 2.4 on Alphas. In fact, we've moved back to 2.2.18 on some macines. (2.2.20 for Alpha didn't compile properly, and I didn't want to mess with it -- anyone know if which 2.2 kernel is best for number-crunching Alphas right now?). Oh, the pain. The lost time. "Kernel of Pain" is a fine description of our 2.4 experience on Alphas.
-Paul Komarek
Me, on the other hand...I was really hoping someone was selling PDP-10 kits.
-Paul Komarek
Some of the testimony in the DoJ and States v MS demonstrated that the price of retail copies of Windows did not drop with age in the same way other software (including OS) products did. Perhaps the overcharging suits will discuss the price paid for Win95 once Win98 was out, or similar.
-Paul Komarek
What makes you think they'd install the software? I suspect they'd mail the CD and never be heard from again. ;-)
-Paul Komarek
Thanks for mentioning the Symphony. TA's music was terrific! I sometimes put the game disc in my stereo while I'm working.
One interesting bit of trivia -- the orchestra, the Seattle Symphony, doubled themselves for the recording -- that is, they played through the same passages twice (not necessarily playing the same notes), and mixed the results together. This is what provides the Mahler-esque proportion to the sound.
-Paul Komarek