"Most people just lack that "this is bullshit" detector."
Evolution is a slow process and e-mail is a new thing - it just takes time to adapt. In a couple million years every human being will be able to tell a 419 from miles away.
NetWare 3 was painful to the extreme. It would require you to set up a dedicated server just for sharing files and, IIRC, the machines that shared the printers took a very heavy performance hit. That and the NLM way of deploying server software made NW3 the worst thing I ever saw in server software. Of course, these were because of the compromises derived from running it on crippled computers. The catch is that it was too complex and expensive for small offices that only wanted to work smarter.
WfW mirrored the simple networking of the Macs of the time, but I am not sure it could be called a copy - it used some of the same code and protocols LAN Manager used. Besides that, it allowed people not to fork out the big bucks it would cost to migrate to Macs.
Architecturally, LAN Manager was vastly superior to NetWare 3 because if offered an environment with real multitasking, memory protection and central authentication. Windows NT grew out of this structure. It's useful to remember that OS/2 and NT performed better and were much less expensive than comparable Unix solutions running on the same hardware.
It's then unfair to say that WfW only succeeded because it sucked less. It was a good product for its time and filled a niche (the "we need a network without throwing away our current computers and programs" one) that needed filling badly.
One could say that since 2006 AMD lost 80% of its value while Intel lost about 10%. OTOH, if you look to 2008 alone, AMD lost about 8% and Intel 25%.
Numbers and stock valuations can be tricky.
The only way you can take on a market leader is by changing the rules. AMD did that with the Opteron, introducing a 64 bit chip people really wanted, and Intel had to play catch up for the following years and still has to feed Itanic for a couple years. But catch up they did and the current AMD lineup is less promising than the one Intel has. And I wouldn't count on Intel falling for the same trick twice.
Now AMD has a really pissed off giant on their backs, which is a very poor choice of where to put one.
They sure put a lot of effort into Vista and its SP1. On the other hand, it's painfully obvious that much effort did not translate into a timely delivery of a solid improvement over XP. This leads to the inescapable conclusion they put their effort in the wrong places.
Had they really been paying attention to their customers, they would have released Vista three years ago with a subset of its features. It would be twice as big as XP and run just as fast. Instead, they delivered what they thought the market wanted, years later than the market wanted and with less improvements than they expected and, surprise, the market does not want it at all.
Vista's failure was well deserved. Of course it will be sold by the millions, thanks to the OEM deals. But it is still a failure.
Microsoft has been, for a lot of its history, very customer focused. They would not be able to achieve their current market position without being customer focused. It wasn't until they have secured their monopoly on Windows through OEM deals that they became the evil company they are today.
I can remember a couple brilliant examples where they outsmarted their competition by paying attention to what the market really wanted:
- Windows for Workgroups: They realized people did not want file and servers - they wanted to share files and printers and do e-mail. WfW, for all its failures, was a bright example of simplicity. With this, they more or less took the low-end of the NetWare business from Novell. This foothold allowed them to claim the rest of NetWare's share with NT.
- Visual Basic: People wanted an easy to use language to develop for Windows. The C/C++ tools they had were hideously expensive and painful to use (they more or less still are - C++ on Windows helps create the ugliest C++ I ever saw). VB surpassed all other development environments for Windows for flexibility, ease of use and productivity. It was the Ruby on Rails of its time. Sadly, it pretty much caused massive brain damage to a generation of programmers that never quite recovered.
Windows 3: People wanted GUIs but couldn't care less about bulletproof multitasking. OS/2 was great, but Windows 3 hit the sweet spot. 3.11 hit it even better with its TrueType rendering.
Windows 95: The last major overhaul of their consumer OS. Gave a nice (for the time) and easy to use GUI overhaul to the tired Windows 3 desktop. Thanks to the problems with the 68K to PPC transition, it was even stable compared to Macs - a first for MS.
"Because Google would/never/ considering doing anything in terms of an operating system. That's just silly!"
I think they should pursue this line (more or less the gOS the Everex Clousdbook uses) not in the same game as Microsoft, but to further commoditize the OS. Having an OS that's inexpensive, reliable and able to easily integrate online app offerings would inflict another painful wound on Microsoft.
Take their OS monopoly away and they will crumble in no time.
"Actually, you would find that you wouldn't be arrested by jackbooted thugs for doing so. Many people pirate Microsoft products without any repercussions at all."
Small-scale piracy (when you copy someone's disk or download and burn an image) is tolerated by MS because these effectively "free as in beer" copies compete with cheaper software alternatives that may or may not be free (as in beer). They don't compete with Microsoft's product line because they have nothing in this price range and, as a bonus, they increase the market share and thus add value to their products because of the network effect.
Microsoft goes aggressively against counterfeiters because their products compete with the boxes they sell in stores. When someone sets out to buy a copy of, say, Windows or Office, they are doing so because they are unwilling to use pirated software. Every sale of counterfeit software is a sale Microsoft lost.
For very similar reasons, MS goes against the hardware makers who bundle pirated versions of Windows with their systems. For Microsoft, the more important part of the OEM relationship is being able to strong-arm OEMs into bundling (or not) whatever they want, so, they are able to shape the software market to their needs.
MS will never go RIAA-style against the small guy because small scale piracy is actually helping them.
Shouldn't the branch predictor be different since the instruction set is different? Or it does its magic on the micro-instruction level (if there is one in P6 and z6)?
BTW, the first message was kind of a joke. It is possible to make more or less the same chip look like a few different ones by changing just a few small parts.
IIRC, there is a company being sued by IBM for making custom-microcode-Itanic-based servers that look too much like IBM mainframes.
"This mainframe uses the the z6 CPU, which is effectively a POWER6 with a different instruction decoder and MMU"
It probably also has a different memory interface and a different register file. What's left from the POWER6 when you take out the instruction decoder, MMU, register file and memory interface? True, they share about 90% in mass - probably the same chip carrier;-)
At some point in time, the engineers had to give up (assuming, of course, they ever really tried to do it) and build something else marketing could call "WinFS".
After intoxicating half its team from the smoke and have to deal with the severe mutilations due to broken mirrors (not to mention the colossal bad-luck caused by them), Microsoft gave up and decided they will never again fall for their own propaganda and try to implement the vaporware.
Vaporware is designed _NOT_ to be implemented, after all.
Spotlight is nice, but it's mostly a search engine that lets applications help it. This is much less than what MS has been promising about WinFS in its various incarnations, even the most modest ones.
WinFS is the ultimate object-oriented filestore with every bell and whistle that fits the age it's announced. In its first concept, it allowed integration and nesting of different data from any application installed in the system. In its last version, it was infinitely searchable. One can only wonder what will it look when it resurfaces again.
It's vaporware that's resurrected every once and then (ever since the early NT vs. IBM's OS/2 times), designed to make Microsoft look like it has some flashy technology pointy-haired-bosses will not be able to tell it's a Really Bad Idea. And they won't because it will never, ever ship.
I am not sure about the others, but I suspect billg is motivated, primaril, by his urge to compensate for every time he was abused by his colleagues at school. His bullies had a lasting and unintended influence in the world...
So while, in a sense, getting a patent should be much harder (it really has to be original, new, useful, non-obvious), perhaps, keeping a patent should be cheaper, so the small guy can protect it.
It's not the whole system that is broken, but the part that allows people to patent the idea of doing something exceedingly vague.
Would you like to share what patents were those and who was that other company? That could be some useful material to defend your position.
"Megacorp : We might be but we will drag you through the courts for the next three years to prove we are not, but meanwhile you are infringing 256 of ours so pay up now...."
The safe route out of this is to make no real product while you litigate - separate your IP licensing business from your thing-making business. If the claims are reasonable - you have a solid patent - Megacorp may drag you through the courts for a couple years but you should be able to secure funding because you will win. If claims are good enough and you are shielded from infringing, Megacorp should quickly seek an agreement out of court to avoid injunctions and damages.
If their lawyers are competent, I mean.
In the meantime, people may call you a patent troll.
You may have a common PC with blue LEDs all over the transparent neon-lit case, with it's multi-color fans and copper coolers that look like jet-engines and you may have been able to violate the license agreement and even hack OSX into it, but don't deceive yourself. It's not a Macintosh.
It's like putting a Romanée-Conti label on a US$5 wine and pretending it's the real thing.
If you can't tell the difference, it's not for you.
Slashdot Mirror
"Most people just lack that "this is bullshit" detector."
Evolution is a slow process and e-mail is a new thing - it just takes time to adapt. In a couple million years every human being will be able to tell a 419 from miles away.
+1 insightful. Too bad I already posted.
"If anything, take whatever he says and do the opposite."
That would be too easy to manipulate.
I would advise to listen to his complaints, disregard his solutions and continue to do whatever makes sense.
NetWare 3 was painful to the extreme. It would require you to set up a dedicated server just for sharing files and, IIRC, the machines that shared the printers took a very heavy performance hit. That and the NLM way of deploying server software made NW3 the worst thing I ever saw in server software. Of course, these were because of the compromises derived from running it on crippled computers. The catch is that it was too complex and expensive for small offices that only wanted to work smarter.
WfW mirrored the simple networking of the Macs of the time, but I am not sure it could be called a copy - it used some of the same code and protocols LAN Manager used. Besides that, it allowed people not to fork out the big bucks it would cost to migrate to Macs.
Architecturally, LAN Manager was vastly superior to NetWare 3 because if offered an environment with real multitasking, memory protection and central authentication. Windows NT grew out of this structure. It's useful to remember that OS/2 and NT performed better and were much less expensive than comparable Unix solutions running on the same hardware.
It's then unfair to say that WfW only succeeded because it sucked less. It was a good product for its time and filled a niche (the "we need a network without throwing away our current computers and programs" one) that needed filling badly.
One could say that since 2006 AMD lost 80% of its value while Intel lost about 10%. OTOH, if you look to 2008 alone, AMD lost about 8% and Intel 25%.
Numbers and stock valuations can be tricky.
The only way you can take on a market leader is by changing the rules. AMD did that with the Opteron, introducing a 64 bit chip people really wanted, and Intel had to play catch up for the following years and still has to feed Itanic for a couple years. But catch up they did and the current AMD lineup is less promising than the one Intel has. And I wouldn't count on Intel falling for the same trick twice.
Now AMD has a really pissed off giant on their backs, which is a very poor choice of where to put one.
Maybe because it would be fun to watch Microsoft burn with either of the two purchases.
Anyway, I too agree SAP is a much better deal than Yahoo.
But it would still be hell.
And immensely fun to watch.
They sure put a lot of effort into Vista and its SP1. On the other hand, it's painfully obvious that much effort did not translate into a timely delivery of a solid improvement over XP. This leads to the inescapable conclusion they put their effort in the wrong places.
Had they really been paying attention to their customers, they would have released Vista three years ago with a subset of its features. It would be twice as big as XP and run just as fast. Instead, they delivered what they thought the market wanted, years later than the market wanted and with less improvements than they expected and, surprise, the market does not want it at all.
Vista's failure was well deserved. Of course it will be sold by the millions, thanks to the OEM deals. But it is still a failure.
That's simply not true.
Microsoft has been, for a lot of its history, very customer focused. They would not be able to achieve their current market position without being customer focused. It wasn't until they have secured their monopoly on Windows through OEM deals that they became the evil company they are today.
I can remember a couple brilliant examples where they outsmarted their competition by paying attention to what the market really wanted:
- Windows for Workgroups: They realized people did not want file and servers - they wanted to share files and printers and do e-mail. WfW, for all its failures, was a bright example of simplicity. With this, they more or less took the low-end of the NetWare business from Novell. This foothold allowed them to claim the rest of NetWare's share with NT.
- Visual Basic: People wanted an easy to use language to develop for Windows. The C/C++ tools they had were hideously expensive and painful to use (they more or less still are - C++ on Windows helps create the ugliest C++ I ever saw). VB surpassed all other development environments for Windows for flexibility, ease of use and productivity. It was the Ruby on Rails of its time. Sadly, it pretty much caused massive brain damage to a generation of programmers that never quite recovered.
Windows 3: People wanted GUIs but couldn't care less about bulletproof multitasking. OS/2 was great, but Windows 3 hit the sweet spot. 3.11 hit it even better with its TrueType rendering.
Windows 95: The last major overhaul of their consumer OS. Gave a nice (for the time) and easy to use GUI overhaul to the tired Windows 3 desktop. Thanks to the problems with the 68K to PPC transition, it was even stable compared to Macs - a first for MS.
"Because Google would /never/ considering doing anything in terms of an operating system. That's just silly!"
I think they should pursue this line (more or less the gOS the Everex Clousdbook uses) not in the same game as Microsoft, but to further commoditize the OS. Having an OS that's inexpensive, reliable and able to easily integrate online app offerings would inflict another painful wound on Microsoft.
Take their OS monopoly away and they will crumble in no time.
But Google _is_ evil.
It's just the lesser of the two.
"Actually, you would find that you wouldn't be arrested by jackbooted thugs for doing so. Many people pirate Microsoft products without any repercussions at all."
Small-scale piracy (when you copy someone's disk or download and burn an image) is tolerated by MS because these effectively "free as in beer" copies compete with cheaper software alternatives that may or may not be free (as in beer). They don't compete with Microsoft's product line because they have nothing in this price range and, as a bonus, they increase the market share and thus add value to their products because of the network effect.
Microsoft goes aggressively against counterfeiters because their products compete with the boxes they sell in stores. When someone sets out to buy a copy of, say, Windows or Office, they are doing so because they are unwilling to use pirated software. Every sale of counterfeit software is a sale Microsoft lost.
For very similar reasons, MS goes against the hardware makers who bundle pirated versions of Windows with their systems. For Microsoft, the more important part of the OEM relationship is being able to strong-arm OEMs into bundling (or not) whatever they want, so, they are able to shape the software market to their needs.
MS will never go RIAA-style against the small guy because small scale piracy is actually helping them.
Hope that clears it up.
Shouldn't the branch predictor be different since the instruction set is different? Or it does its magic on the micro-instruction level (if there is one in P6 and z6)?
BTW, the first message was kind of a joke. It is possible to make more or less the same chip look like a few different ones by changing just a few small parts.
IIRC, there is a company being sued by IBM for making custom-microcode-Itanic-based servers that look too much like IBM mainframes.
Wish they could make it available as a single PDF. I don't have enough free time to write a program to download them all and merge them.
"This mainframe uses the the z6 CPU, which is effectively a POWER6 with a different instruction decoder and MMU"
;-)
It probably also has a different memory interface and a different register file. What's left from the POWER6 when you take out the instruction decoder, MMU, register file and memory interface? True, they share about 90% in mass - probably the same chip carrier
At some point in time, the engineers had to give up (assuming, of course, they ever really tried to do it) and build something else marketing could call "WinFS".
After intoxicating half its team from the smoke and have to deal with the severe mutilations due to broken mirrors (not to mention the colossal bad-luck caused by them), Microsoft gave up and decided they will never again fall for their own propaganda and try to implement the vaporware.
Vaporware is designed _NOT_ to be implemented, after all.
Spotlight is nice, but it's mostly a search engine that lets applications help it. This is much less than what MS has been promising about WinFS in its various incarnations, even the most modest ones.
WinFS is the ultimate object-oriented filestore with every bell and whistle that fits the age it's announced. In its first concept, it allowed integration and nesting of different data from any application installed in the system. In its last version, it was infinitely searchable. One can only wonder what will it look when it resurfaces again.
WinFS is not and never will be a file system.
In fact, I doubt it will ever be a real product.
It's vaporware that's resurrected every once and then (ever since the early NT vs. IBM's OS/2 times), designed to make Microsoft look like it has some flashy technology pointy-haired-bosses will not be able to tell it's a Really Bad Idea. And they won't because it will never, ever ship.
WinFS is not real.
"I don't get how people can say someone is making too much money."
It seems he did a fine job with Vista. In that regard, it's apparent MS did pay him way too much money compared to what he was worth.
"...who dumped the still not-ready-for-prime-time OS into RC1 status as he bolted for a new gig at Amazon."
From what I read, Vista SP1 is not ready for prime time either...
"Do they? which sheep?"
The black ones. Always.
It's like "hit Microsoft and run and make a whole bunch of money in the process".
Sweet.
I am not sure about the others, but I suspect billg is motivated, primaril, by his urge to compensate for every time he was abused by his colleagues at school. His bullies had a lasting and unintended influence in the world...
US$ 640 million should be enogh for anyone
So while, in a sense, getting a patent should be much harder (it really has to be original, new, useful, non-obvious), perhaps, keeping a patent should be cheaper, so the small guy can protect it.
It's not the whole system that is broken, but the part that allows people to patent the idea of doing something exceedingly vague.
Would you like to share what patents were those and who was that other company? That could be some useful material to defend your position.
"Megacorp : We might be but we will drag you through the courts for the next three years to prove we are not, but meanwhile you are infringing 256 of ours so pay up now ...."
The safe route out of this is to make no real product while you litigate - separate your IP licensing business from your thing-making business. If the claims are reasonable - you have a solid patent - Megacorp may drag you through the courts for a couple years but you should be able to secure funding because you will win. If claims are good enough and you are shielded from infringing, Megacorp should quickly seek an agreement out of court to avoid injunctions and damages.
If their lawyers are competent, I mean.
In the meantime, people may call you a patent troll.
"Now I'm on Macintosh/Hackintosh,"
"(Hint : Intel chipset + Intel CPU + nVidia GPU = Mac)"
Not any more than Volkswagen == Lamborghini
You may have a common PC with blue LEDs all over the transparent neon-lit case, with it's multi-color fans and copper coolers that look like jet-engines and you may have been able to violate the license agreement and even hack OSX into it, but don't deceive yourself. It's not a Macintosh.
It's like putting a Romanée-Conti label on a US$5 wine and pretending it's the real thing.
If you can't tell the difference, it's not for you.