For years, Apple has been ripped by 3rd party developers whenever they tried to release software that advanced their platform. "Waa! How can I sell my Asswriter shareware when MacWrite is free! Wah!"
It looks like Jobs finally looked over the fence at the Windows world, where Microsoft competes with almost everyone and they still all write software for Windows.
As others have pointed out Roxio essentially sells device drivers. They're doomed no matter what (PC OEMs don't want to pay the Roxio tax any more than Apple does.)
A much better example is Final Cut Pro which competes directly with Adobe Premiere. Apple realized if they were going to stay promienant in that particular submarket, they were going to have to be proactive and not sit on thier hands and let a cross-platform 3rd party product do the heavy lifting. The result is that Final Cut has sold more Macs than Premiere ever did.
Excel supported VBA going back to version 5.0 (Office 4.2). It was later expanded to the other products in the suite.
There's also other vendors like Corel WordPerfect that have licenced VBA from Microsoft. It's unclear if this is a problem in the VBA runtime or the Excel/PowerPoint fileformats though.
Hey -- The first computer I had with a Forward Delete key was a Mac SE! (I sprung the extra $100 or whatever for the extended keyboard, which I used until 1997.).
I don't really give a crap if you think two delete keys are too hard -- I'm used to it, and my fingers have managed to touchtype the correct key on numerous wierd PC laptops. As long as I'm using different machines, it will be a bitch. But thanks anyway for the shift delete tip.
The 'official' workaround for right-click is to long-click (click-n-hold for a few seconds).
I was a bit annoyed at the lack of a 2nd button my new-to-me G3 laptop, but the long-click is easy enough to get used to.
What's really pissing me off is the lack of a proper keyboard -- you need to hold down Fn to get at PageUp/PageDown, which means that have to use both hands to browse the web. There's also no Forward Delete key, which frankly just sucks in this day and age.
Add to this the fact that it GETS RESULTS and RUNS LINUX... you've got a perfect engine.
Hey - It's not so much that they RUN Linux it's that they do an exceptional job of INDEXING Linux- (and Perl- and so on) pages. In fact this seemed to be one of their primary focuses during the beta period.
Result -- lots of techies got on board early and spread the word to non techies. Genius marketing.
MS Kerberos interoperates with MIT Kerberos for authentication purposes (who you are, such as your user name). This seems to be the sell of Passport/Hailstorm.
Microsoft's extention was to add a NT UID (or UUID or whatever it's called), which effectively determines your authorization (what you can do). They used a field specifically designed for this purpose.
This eleminates the need for a local/etc/passwd type (or in MS terms 'SID') mapping of user name ("root") to UID (0). If you've ever worked in an NDS or other directory environment, you'd know that the primary point of a DS is to centralize security admin, so you can see why this was a necessary step.
Now, how this works out in Hailstorm probably depends on how you use it. For a message board or online shopping, the provider would probably just need the authentication and handle the authorization themselves (ie MS wouldn't provide the information that "CmdrTaco" is the admin of Slashdot, but would verify that CmdrTaco is who he says he is.)
BUT.. It could be that you could 'outsource' your PDC to Microsoft and set up LAN security using Hailstorm IDs. Sound retarded, but recall that the current crop of small shop MCSEs is having difficulty groking AD, and LanMan/NT4 is going away eventually. The next step would be move Exchange (or more likely "Small Business Server") off-site and make that a service also. You can see the possiblities.
I would agree with the points in that article (arcades weren't professionally managed or promoted), but I think it goes a little deeper than that. In the early 80s, every upstanding suburban mall had an arcade (see Fast Times at Ridgemont High, where it's a central motif.) These were gradually shutdown because they attracted the 'wrong element' (kids who caused trouble and shoplifted) and arcades were pretty much pushed into the dark corners of the inner city to be managed by the trolls. This put them out of reach from the average suburban kid with a good allowance.
What really finished off the arcades though, was the fact that everyone put their eggs in the fighting game basket. Those things sucked quarters like mad but developed a tournament mentatilty where only the people who memorized the button combos could survive, discouraging the casual player. No longer could you just drop a quarter and see if you could survive on your instincts and hand-eye coordination - some kid would put the super-kill move on you and that would be it for your 75 cents.
When the fighting game fad ended, the arcade owners were stuck with a bunch of huge, expensive machines that had no players. Maybe "Super Excellent Street Warrior IV Deluxe Edition" will get them back! -- Right... At that point nearly all the creativity had been sucked dry from the industry and it pretty much folded. The innercity arcade hole I stop buy occasionally seems to survive on it's pool tables and some well worn classics like Ms Pac Man and Centipede.
Well, the main problem with the cable industry is that building it out in the 70s and 80s proved to be a very expensive proposition. Smaller companies financed it with junk bonds and so on, and when they found that adoption rates were lower than expected, they sold out to bigger fish, who sold out to bigger fish. And so on, until AT+T and TimeWarner and a couple others ended up holding the bag of lots of customers and lots of debt (AT+T is billions in the hole).
AT+T's theory was that they were the bluest of the blue chips and they would be able to carry all that debt without a problem. When the dotcom bubble deflated, it turned out they were very wrong.
Cable Internet is probably not a super-profitable business, but it does get the TV subscription rates up. And once you have 'digital cable' installed, you are more likely to do the things that make them the real money -- premium channels, PPV, etc. (And eventually local telephone services, maybe.)
The upshot is that it you are probably right -- it will take very large diversified companies to effectively pay off the infrastructure costs without being swamped. The real moral of the story is that running a wire into everyone's house for a luxury service is a bad idea -- which is why you'll never see fiber-to-the-curb. Expect the next great build-outs to wireless only.
This bill, if passed, would make it a crime to use Linux, since it does not (and probably won't) incorporate the copy protection standards set by the government
You know, that's actually the best reason that this bill won't pass (at least as-is).
All the opponnents need to do is call Linus Torvalds to testify. He can describe his wonderful free software that's used by thousands of US businesses and the government, and how under this law, his software would be considered contraband, Linus and his fellow developers would have to leave the US and stop paying taxes.
That will at least narrow the scope of this law from all (desktop?) computers to "personal media terminals", which is probably Hollywoods ultimate intent anyway.
For MS, removal of DOS was a good thing because it removed the command line
Back in the late 80s and early 90s, there was a huge PC user war between the commandline gurus and their Novell login scripts (FIRE PHASERS) and batch files, and the GUI-friendly userbase that liked the Macintosh and even saw some value in Windows or OS/2. The DOS&Novell guys ruled the roost and routinely berated GUIs as "cartoony" and "toy computers" and did everything in their power to supress Windows 3.
Then businesses realized that GUI environments would save them $Money and the DOS/Novell fanatics were purged in a huge pogrom. They either were reeducated to think "GUI is great! Windows is great! Down with the command line!" or they retreated to using UNIX.
The unspoken legacy of this war is the main reason that CMD.EXE is downplayed in the MS crowd. Microsoft of course is of two minds about the subject - on one hand shipping many advanced admin tools as commandline only, and on the other not shipping what should be standard commandline stuff on the OS media but instead forcing you to install the Resoure Kit.
The "killer" feature for the iBook is that it has ZERO PCMCIA slots, which kills my willingness to go buy one.
It's true that built-in 1394/USB/Ethernet/Modem/802.11b does eliminate most of expansion problem, but I refuse to box myself because Apple decided to remove a standard feature in favor of brand engineering. (For one, I have lots of SCSI stuff around.)
The NEC has 1 PCMCIA slot, but all NEC laptops are poorly engineered turd chunks, so we'll forget that.
I think HP and Digital (and SGI) both made the same fundemental mistake. They both (correctly) figured that the longterm future would not be proprietary systems but instead it would be commodity operating systems (uhh, Windows) on a commodity platform (Intel 32 or 64-bit). They just sorely mis-estimated how long it would take to get this long-term future.
When the Pentium Pro briefly became the fastest chip available in the early 90s, it sent shockwaves through the Unix guys. They suddenly realized that there was no way they'd be able to keep up with WinTel, and would either have to get on the boat or sink.
What they forgot was that they had customers *today* that were worried more about shortterm roadmaps and support and so on. As soon as DEC and HP showed the slightest wavering around their commitment to UNIX on RISC, the Sun salesmen were there to steal accounts from them. Then, NT5 and IA64 were delayed, just long enough so that the dot com sales explosion was primarily focused good old-fashioned Unix - a market where HP and Compaq/DEC had almost entirely fallen off the radar for the non-instituational customer, and Sun and IBM cleaned up big time.
Another thing that these dinosaurs missed is the Andresson statement that the Internet makes the OS irrelevant. Sun and IBM figured it out and made a massive middleware ('eBusiness') push with Java. Microsoft figured the same with their own proprietary stuff. With DEC's software experience, they *should* have been on the stick, but by then I think they'd resigned themselves to the Compaq boxpushing mentality.
All of this doesn't change the final outcome tho -- the long term future is still commodity Windows|Linux on commodity Intel. Knowing this doesn't help, because you still have to support those VMS/Tandem/MPIX/HPUX/Tru64 customers for the next 10 years before it gets there. And when they do decide to jump ship, they're not necesarilly going to be calling the HP salesman.
The DMCA defines what they mean by "effective" copy-protection. It doesn't require strong encryption, just that the user effectively wouldn't be able to copy the content without using a device to bypass the protection.
For example, it was silly when MS tried to raise the DMCA against slashdot because they had password-protected a zip file. Turns out that asking for a password is an optional client feature of zip, not effective content protection. The same is not true for eBooks.
I should clarify - this was OS/2 Server 3 and 4 mixed with the regular client OS being used in a server role. Forgot that the server and client version numbers didn't line up.
I admined a large number of OS/2 boxes back in the 2.x and early 3.0 days, and that shit was far less reliable under load than even NT 3.51 (which is what we migrated to.)
Back in about 1999, Infoworld published a Microsoft study determining the causes of NT4 failures. "Internal Failure" ranked almost as high as "Third Party Driver". As someone who did production NT admin from 3.5 to about NT4 SP4, this definately lines up with experience that the early NT4 releases had LOTS of problems.
1) In direct postal mail a response rate of.5% is considered good. It could be that with essentially 0 cost to the sender that a response rate of.01% makes it all worthwhile.
2) Spammers don't make any money from spam. Instead they make their money scamming would-be spammers by selling 'consulting services', address lists (often containing the known complainers), and harvesting or mailing software (uhh, free dimitry!). It's a big pyramid scheme, which is why they talk themselves up like Amway salesmen. And it's natural because people think "I get so much spam, SOMEONE must be buying this stuff".
Of course, once the sucker has taken the hook and realized that spamming isn't a get-rich-quick scheme, the dim bulb goes off in their head and they immedeately start running trying to hook other would-be spammers. The cycle continues.
Oh there's some other Moore's Law that deals with that boring transitor stuff, but "Moore's Law of cpu marketing <--L@@K" is all about the Mhz, baby. Be sure that Intel worries more about the marketing law than the transistor one.
Sure, Apple was selling 68040 machines using the conventional clock measurement. For example, if you purchased a Quadra 950, both the manual and the CPU itself indicated that it was a 040-33.
Then Intel started selling lots of i486/DX2-66 machines, which were 66Mhz CPUs running on a 33Mhz bus. At this point, I believe that Apple had already introduced PowerPC machines, but still had a large low-end lineup of 040 machines.
On the pretense that some interal component of the chip ran at double the clock speed, Apple relabeled all of their 040 machines to make them sound like they had twice the clockspeed. Motorola's published specs for the CPU never changed, mind you - Apple did this by themselves. They also updated the spec sheets on their Internet site (gopher at the time) so that if you looked up your discontinued Quadra it would tell you that you now magically had a "040-66/33" CPU. (Also see my other reply.)
Traditionally, Microsoft has taken existing, proven technology, cloned it, and commoditized it by underselling their rivals. That's why they've gained so much support in the "nerd" community over the years (thinking specifically about the Windows versus OS/2 battle, or the fact WinNT was $300 when a Unix licence was $Thousands), which has translated into broad adoption of their products.
Now with recent changes in their pricing strategy this might not be true anymore. If anything the Open Source business methodology is interesting because it's the first strategy that Microsoft can't just undercut.
Well the reason that most hardware sites are useless is that their methodology sucks. "Ummm, OK, this got X frames in Quake and a Y score in winbench and the bios has a nice menu and I overclocked it and Windows 98 seemed OK for the 10 minutes I tried it. 5 STARS!!"
The fact that they somewhat accidentally found out the real quality of some product is astounding. God forbid that they would put every motherboard in as a production webserver for a week or so to see how it did. Or plug in a video capture board or 1024MB of RAM or any of the other things which might be useful to someone specing out a box that isn't meant to be this month's disposable game machine.
Another example is the terrible 2D quality of Nvidia cards. These sites are wall-to-wall GeForce reviews, but I've seen maybe two that actually even tagentially made a subjective judgement about the 2D quality. Maybe in their world they play 3D games 90% of the time, but not in mine.
MSNBC have proven themselves to be pretty damn impartial.
A couple years ago/. was arguing over whether MSNBC was being used for nefarious purposes by Microsoft, and I went to their site and searched for "General Electric" and "GE" and got zero hits. Considering they're one of the worlds largest corporations and own 50% of MSNBC, that seemed a little strange.
Currently, tons of hits come up, including an article about whether the GE chairman influenced NBC's election coverage.
Oh, you could be right, but I would caution that there's no guarantee that the next killer app will run on a PC at all.
The lynchpin of the success of the personal computer is as a desktop business machine, in that it's only improved and replaced the typewriter, adding machine, copy machine, telephone, overhead projector, and fax machine that were sitting there before.
Bill Gates was wrong when he supposedly said that because spreadsheet users (the killer app of it's day) were already hitting the memory limit. What limits are people hitting today that won't be resolved in the mythical 18 months? Give it a couple more gigahertz and you've got real-time software HDTV editing beat, not that that matters to business customers.
Then what? My guess is that the market fractures into far more specialized units, and lots of PC OEMs thinking that they could push a beige box on the Mhz number alone start going under.
or designing a processor that gets less done per clock and then jacking the clock up
You guys seem to forget that this is *exactly* what the Athlon did to the P3 -- put in a long pipeline to get the magic Mhz number up up up. (Is there any other way?)
The fact that Intel did the same shouldn't be a suprise to anyone here or AMD. After all they do have "Moore's Law (of cpu marketing)" which states that they WILL double the Mhz every 18 months or so.
AMD just got out-engineered AND out-marketed on this one. When they released their 1Ghz chip before Intel did they knew exactly how long they had to release one that was twice as fast to keep their lead. (I'm assuming that these P numbers are an act of desparation here, forgive me if someone over at AMD actually thought this was a bright idea.)
Slashdot Mirror
For years, Apple has been ripped by 3rd party developers whenever they tried to release software that advanced their platform. "Waa! How can I sell my Asswriter shareware when MacWrite is free! Wah!"
It looks like Jobs finally looked over the fence at the Windows world, where Microsoft competes with almost everyone and they still all write software for Windows.
As others have pointed out Roxio essentially sells device drivers. They're doomed no matter what (PC OEMs don't want to pay the Roxio tax any more than Apple does.)
A much better example is Final Cut Pro which competes directly with Adobe Premiere. Apple realized if they were going to stay promienant in that particular submarket, they were going to have to be proactive and not sit on thier hands and let a cross-platform 3rd party product do the heavy lifting. The result is that Final Cut has sold more Macs than Premiere ever did.
Excel supported VBA going back to version 5.0 (Office 4.2). It was later expanded to the other products in the suite.
There's also other vendors like Corel WordPerfect that have licenced VBA from Microsoft. It's unclear if this is a problem in the VBA runtime or the Excel/PowerPoint fileformats though.
Hey -- The first computer I had with a Forward Delete key was a Mac SE! (I sprung the extra $100 or whatever for the extended keyboard, which I used until 1997.).
I don't really give a crap if you think two delete keys are too hard -- I'm used to it, and my fingers have managed to touchtype the correct key on numerous wierd PC laptops. As long as I'm using different machines, it will be a bitch. But thanks anyway for the shift delete tip.
The 'official' workaround for right-click is to long-click (click-n-hold for a few seconds).
I was a bit annoyed at the lack of a 2nd button my new-to-me G3 laptop, but the long-click is easy enough to get used to.
What's really pissing me off is the lack of a proper keyboard -- you need to hold down Fn to get at PageUp/PageDown, which means that have to use both hands to browse the web. There's also no Forward Delete key, which frankly just sucks in this day and age.
Add to this the fact that it GETS RESULTS and RUNS LINUX... you've got a perfect engine.
Hey - It's not so much that they RUN Linux it's that they do an exceptional job of INDEXING Linux- (and Perl- and so on) pages. In fact this seemed to be one of their primary focuses during the beta period.
Result -- lots of techies got on board early and spread the word to non techies. Genius marketing.
MS Kerberos interoperates with MIT Kerberos for authentication purposes (who you are, such as your user name). This seems to be the sell of Passport/Hailstorm.
/etc/passwd type (or in MS terms 'SID') mapping of user name ("root") to UID (0). If you've ever worked in an NDS or other directory environment, you'd know that the primary point of a DS is to centralize security admin, so you can see why this was a necessary step.
.. It could be that you could 'outsource' your PDC to Microsoft and set up LAN security using Hailstorm IDs. Sound retarded, but recall that the current crop of small shop MCSEs is having difficulty groking AD, and LanMan/NT4 is going away eventually. The next step would be move Exchange (or more likely "Small Business Server") off-site and make that a service also. You can see the possiblities.
Microsoft's extention was to add a NT UID (or UUID or whatever it's called), which effectively determines your authorization (what you can do). They used a field specifically designed for this purpose.
This eleminates the need for a local
Now, how this works out in Hailstorm probably depends on how you use it. For a message board or online shopping, the provider would probably just need the authentication and handle the authorization themselves (ie MS wouldn't provide the information that "CmdrTaco" is the admin of Slashdot, but would verify that CmdrTaco is who he says he is.)
BUT
I would agree with the points in that article (arcades weren't professionally managed or promoted), but I think it goes a little deeper than that. In the early 80s, every upstanding suburban mall had an arcade (see Fast Times at Ridgemont High, where it's a central motif.) These were gradually shutdown because they attracted the 'wrong element' (kids who caused trouble and shoplifted) and arcades were pretty much pushed into the dark corners of the inner city to be managed by the trolls. This put them out of reach from the average suburban kid with a good allowance.
What really finished off the arcades though, was the fact that everyone put their eggs in the fighting game basket. Those things sucked quarters like mad but developed a tournament mentatilty where only the people who memorized the button combos could survive, discouraging the casual player. No longer could you just drop a quarter and see if you could survive on your instincts and hand-eye coordination - some kid would put the super-kill move on you and that would be it for your 75 cents.
When the fighting game fad ended, the arcade owners were stuck with a bunch of huge, expensive machines that had no players. Maybe "Super Excellent Street Warrior IV Deluxe Edition" will get them back! -- Right... At that point nearly all the creativity had been sucked dry from the industry and it pretty much folded. The innercity arcade hole I stop buy occasionally seems to survive on it's pool tables and some well worn classics like Ms Pac Man and Centipede.
Well, the main problem with the cable industry is that building it out in the 70s and 80s proved to be a very expensive proposition. Smaller companies financed it with junk bonds and so on, and when they found that adoption rates were lower than expected, they sold out to bigger fish, who sold out to bigger fish. And so on, until AT+T and TimeWarner and a couple others ended up holding the bag of lots of customers and lots of debt (AT+T is billions in the hole).
AT+T's theory was that they were the bluest of the blue chips and they would be able to carry all that debt without a problem. When the dotcom bubble deflated, it turned out they were very wrong.
Cable Internet is probably not a super-profitable business, but it does get the TV subscription rates up. And once you have 'digital cable' installed, you are more likely to do the things that make them the real money -- premium channels, PPV, etc. (And eventually local telephone services, maybe.)
The upshot is that it you are probably right -- it will take very large diversified companies to effectively pay off the infrastructure costs without being swamped. The real moral of the story is that running a wire into everyone's house for a luxury service is a bad idea -- which is why you'll never see fiber-to-the-curb. Expect the next great build-outs to wireless only.
This bill, if passed, would make it a crime to use Linux, since it does not (and probably won't) incorporate the copy protection standards set by the government
You know, that's actually the best reason that this bill won't pass (at least as-is).
All the opponnents need to do is call Linus Torvalds to testify. He can describe his wonderful free software that's used by thousands of US businesses and the government, and how under this law, his software would be considered contraband, Linus and his fellow developers would have to leave the US and stop paying taxes.
That will at least narrow the scope of this law from all (desktop?) computers to "personal media terminals", which is probably Hollywoods ultimate intent anyway.
For MS, removal of DOS was a good thing because it removed the command line
Back in the late 80s and early 90s, there was a huge PC user war between the commandline gurus and their Novell login scripts (FIRE PHASERS) and batch files, and the GUI-friendly userbase that liked the Macintosh and even saw some value in Windows or OS/2. The DOS&Novell guys ruled the roost and routinely berated GUIs as "cartoony" and "toy computers" and did everything in their power to supress Windows 3.
Then businesses realized that GUI environments would save them $Money and the DOS/Novell fanatics were purged in a huge pogrom. They either were reeducated to think "GUI is great! Windows is great! Down with the command line!" or they retreated to using UNIX.
The unspoken legacy of this war is the main reason that CMD.EXE is downplayed in the MS crowd. Microsoft of course is of two minds about the subject - on one hand shipping many advanced admin tools as commandline only, and on the other not shipping what should be standard commandline stuff on the OS media but instead forcing you to install the Resoure Kit.
The "killer" feature for the iBook is that it has ZERO PCMCIA slots, which kills my willingness to go buy one.
It's true that built-in 1394/USB/Ethernet/Modem/802.11b does eliminate most of expansion problem, but I refuse to box myself because Apple decided to remove a standard feature in favor of brand engineering. (For one, I have lots of SCSI stuff around.)
The NEC has 1 PCMCIA slot, but all NEC laptops are poorly engineered turd chunks, so we'll forget that.
I think HP and Digital (and SGI) both made the same fundemental mistake. They both (correctly) figured that the longterm future would not be proprietary systems but instead it would be commodity operating systems (uhh, Windows) on a commodity platform (Intel 32 or 64-bit). They just sorely mis-estimated how long it would take to get this long-term future.
When the Pentium Pro briefly became the fastest chip available in the early 90s, it sent shockwaves through the Unix guys. They suddenly realized that there was no way they'd be able to keep up with WinTel, and would either have to get on the boat or sink.
What they forgot was that they had customers *today* that were worried more about shortterm roadmaps and support and so on. As soon as DEC and HP showed the slightest wavering around their commitment to UNIX on RISC, the Sun salesmen were there to steal accounts from them. Then, NT5 and IA64 were delayed, just long enough so that the dot com sales explosion was primarily focused good old-fashioned Unix - a market where HP and Compaq/DEC had almost entirely fallen off the radar for the non-instituational customer, and Sun and IBM cleaned up big time.
Another thing that these dinosaurs missed is the Andresson statement that the Internet makes the OS irrelevant. Sun and IBM figured it out and made a massive middleware ('eBusiness') push with Java. Microsoft figured the same with their own proprietary stuff. With DEC's software experience, they *should* have been on the stick, but by then I think they'd resigned themselves to the Compaq boxpushing mentality.
All of this doesn't change the final outcome tho -- the long term future is still commodity Windows|Linux on commodity Intel. Knowing this doesn't help, because you still have to support those VMS/Tandem/MPIX/HPUX/Tru64 customers for the next 10 years before it gets there. And when they do decide to jump ship, they're not necesarilly going to be calling the HP salesman.
The DMCA defines what they mean by "effective" copy-protection. It doesn't require strong encryption, just that the user effectively wouldn't be able to copy the content without using a device to bypass the protection.
For example, it was silly when MS tried to raise the DMCA against slashdot because they had password-protected a zip file. Turns out that asking for a password is an optional client feature of zip, not effective content protection. The same is not true for eBooks.
I should clarify - this was OS/2 Server 3 and 4 mixed with the regular client OS being used in a server role. Forgot that the server and client version numbers didn't line up.
I admined a large number of OS/2 boxes back in the 2.x and early 3.0 days, and that shit was far less reliable under load than even NT 3.51 (which is what we migrated to.)
Apparently all BSOD's are NT's fault
Back in about 1999, Infoworld published a Microsoft study determining the causes of NT4 failures. "Internal Failure" ranked almost as high as "Third Party Driver". As someone who did production NT admin from 3.5 to about NT4 SP4, this definately lines up with experience that the early NT4 releases had LOTS of problems.
Two theories:
.5% is considered good. It could be that with essentially 0 cost to the sender that a response rate of .01% makes it all worthwhile.
1) In direct postal mail a response rate of
2) Spammers don't make any money from spam. Instead they make their money scamming would-be spammers by selling 'consulting services', address lists (often containing the known complainers), and harvesting or mailing software (uhh, free dimitry!). It's a big pyramid scheme, which is why they talk themselves up like Amway salesmen. And it's natural because people think "I get so much spam, SOMEONE must be buying this stuff".
Of course, once the sucker has taken the hook and realized that spamming isn't a get-rich-quick scheme, the dim bulb goes off in their head and they immedeately start running trying to hook other would-be spammers. The cycle continues.
Oh there's some other Moore's Law that deals with that boring transitor stuff, but "Moore's Law of cpu marketing <--L@@K" is all about the Mhz, baby. Be sure that Intel worries more about the marketing law than the transistor one.
Sure, Apple was selling 68040 machines using the conventional clock measurement. For example, if you purchased a Quadra 950, both the manual and the CPU itself indicated that it was a 040-33.
Then Intel started selling lots of i486/DX2-66 machines, which were 66Mhz CPUs running on a 33Mhz bus. At this point, I believe that Apple had already introduced PowerPC machines, but still had a large low-end lineup of 040 machines.
On the pretense that some interal component of the chip ran at double the clock speed, Apple relabeled all of their 040 machines to make them sound like they had twice the clockspeed. Motorola's published specs for the CPU never changed, mind you - Apple did this by themselves. They also updated the spec sheets on their Internet site (gopher at the time) so that if you looked up your discontinued Quadra it would tell you that you now magically had a "040-66/33" CPU. (Also see my other reply.)
You are simplifying what they did. By all conventional terms it was 33Mhz chip. See here.
Traditionally, Microsoft has taken existing, proven technology, cloned it, and commoditized it by underselling their rivals. That's why they've gained so much support in the "nerd" community over the years (thinking specifically about the Windows versus OS/2 battle, or the fact WinNT was $300 when a Unix licence was $Thousands), which has translated into broad adoption of their products.
Now with recent changes in their pricing strategy this might not be true anymore. If anything the Open Source business methodology is interesting because it's the first strategy that Microsoft can't just undercut.
Well the reason that most hardware sites are useless is that their methodology sucks. "Ummm, OK, this got X frames in Quake and a Y score in winbench and the bios has a nice menu and I overclocked it and Windows 98 seemed OK for the 10 minutes I tried it. 5 STARS!!"
The fact that they somewhat accidentally found out the real quality of some product is astounding. God forbid that they would put every motherboard in as a production webserver for a week or so to see how it did. Or plug in a video capture board or 1024MB of RAM or any of the other things which might be useful to someone specing out a box that isn't meant to be this month's disposable game machine.
Another example is the terrible 2D quality of Nvidia cards. These sites are wall-to-wall GeForce reviews, but I've seen maybe two that actually even tagentially made a subjective judgement about the 2D quality. Maybe in their world they play 3D games 90% of the time, but not in mine.
MSNBC have proven themselves to be pretty damn impartial.
/. was arguing over whether MSNBC was being used for nefarious purposes by Microsoft, and I went to their site and searched for "General Electric" and "GE" and got zero hits. Considering they're one of the worlds largest corporations and own 50% of MSNBC, that seemed a little strange.
A couple years ago
Currently, tons of hits come up, including an article about whether the GE chairman influenced NBC's election coverage.
Oh, you could be right, but I would caution that there's no guarantee that the next killer app will run on a PC at all.
The lynchpin of the success of the personal computer is as a desktop business machine, in that it's only improved and replaced the typewriter, adding machine, copy machine, telephone, overhead projector, and fax machine that were sitting there before.
Bill Gates was wrong when he supposedly said that because spreadsheet users (the killer app of it's day) were already hitting the memory limit. What limits are people hitting today that won't be resolved in the mythical 18 months? Give it a couple more gigahertz and you've got real-time software HDTV editing beat, not that that matters to business customers.
Then what? My guess is that the market fractures into far more specialized units, and lots of PC OEMs thinking that they could push a beige box on the Mhz number alone start going under.
or designing a processor that gets less done per clock and then jacking the clock up
You guys seem to forget that this is *exactly* what the Athlon did to the P3 -- put in a long pipeline to get the magic Mhz number up up up. (Is there any other way?)
The fact that Intel did the same shouldn't be a suprise to anyone here or AMD. After all they do have "Moore's Law (of cpu marketing)" which states that they WILL double the Mhz every 18 months or so.
AMD just got out-engineered AND out-marketed on this one. When they released their 1Ghz chip before Intel did they knew exactly how long they had to release one that was twice as fast to keep their lead. (I'm assuming that these P numbers are an act of desparation here, forgive me if someone over at AMD actually thought this was a bright idea.)