The history of the IT industry is littered with the desiccated bones and rotting corpses of companies that trusted and partnered with Microsoft.
Yes and the funny thing is people keep falling for it again and again. This Nokia business is just one more example of it. Heck even the Simpsons made fun of it once.
Yes it's a flat rate where Nokia isn't selling cellphones anywhere near the amount they thought they would so the result is $60 paid per phone i.e. more than what the other Windows Phone manufacturers pay. I honestly cannot understand how you think this was a good deal for Nokia. It wasn't and anyone with half a brain could have seen it at the time.
I don't think the licensing costs are as low as you claim they are. I have heard from several sources that due to the way the licensing was structured Nokia is paying a lot more per unit than the other Windows Phone vendors. I have also heard that they have paid back in royalties to Microsoft any cash that Microsoft has provided to them. Which is not particularly hard to believe since that article claims they are paying $250 million per quarter on licensing costs alone.
DEC is more complicated to solve. They missed the dot com boom despite having perfectly fine hardware. They did not market Alpha servers as well as Sun advertised SPARC. They did not manage to keep their CPU manufacturing plant to full utilization unlike IBM who manufactures chips for other companies. Their PC server division could also be considered non existent. Their desktops were constantly more expensive and more obsolete than the competition. That is basically it.
What Nokia was on the verge of was a series of interlocking conflicting plans which were unachievable with no solid product pipeline to create the revenue to fix the mess.
1. Sell N9.
2. Profit.
3. Make cheaper versions of the N9 and migrate those to the low end market.
4. Profit.
5. Keep improving MeeGo or whatever its name is today.
6. Release new top end phone.
7. Go back to step 1.
Android has better licensing costs and allows more customization than Windows Phone. It has more apps. What reason was there to pick Windows Phone anyway? Besides Elop having come from Microsoft that is.
GlobalFoundries got AMD's old fabs at Dresden. Since AMD fabs their CPUs at GlobalFoundries they are basically still using the same factories. The GPUs are still fabbed at TSMC just like ATI used to do. They are supposedly going to fab an APU (CPU/GPU combo) at TSMC but I am unaware if this has happened yet or not.
No I meant exactly what I said. Microsoft is no stranger to doing this sort of thing. They also did a nice little FUD campaign to camouflage their actions. Not that the FUD went 100% like MS wanted but MS still managed to kill Linux on netbooks. And netbooks.
The same thing they did to the netbooks. They will try to get the manufacturers to pay a fee to Microsoft for every Chromebook they sell. The fee will of course be greater than the cost of a Windows ilcense. If the manufacturers balk at it they will just threaten to increase the cost of the Windows licenses for the Windows laptops these manufacturers already sell. Google better get some non Windows laptop manufacturers making these things fast.
Samsung has been following the ARM roadmap pretty much faithfully so they aren't copying anything in chip design from Apple. Their core design is just an ARM licensed core. Which BTW is as good or better than the A6X core (triple-issue out-of-order). The A7 core remains to be seen.
same chip is being produced (it's Apple's design), it's just a different manufacturer
Tell that to AMD. They have been trying to outsource CPU manufacturing for years and they kept failing along the way. First AMD wanted to outsource to Chartered which was supposedly using the same manufacturing process (developed jointly by IBM, AMD, Samsung, Chartered) but it turned out they couldn't just trivially port their design over. Then they considered switching to TSMC. Another fail. They also considered switching GPU manufacturing from TSMC to GlobalFoundries after purchasing ATI. Yet another fail. The more low level optimizations the chip has the harder it is to port it. You don't just hit a compile button and then the thing magically works. Each manufacturing process has its own little details you have to work around in order for the design to be manufactureable and hit the right performance and power consumption targets.
What the article does not say is that it usually takes 12 to 18 months for a processor to go from tapeout to final production. Just because they have a design ready for TSMC to manufacture does not mean the first run will be bug free not to mention that the latency between a wafer going in and coming out is larger than a lot of people realize.
The netbook was nothing but a quick bait and switch by manufacturers that wanted to make a quick buck off of the recession. The image of a business person using a netbook is just that. Users of netbooks were people with little money looking for a new toy, and nothing more.
The reason for the downfall was indeed Microsoft. The original EeePC came with a Celeron processor (not Atom) and an SSD. It had a longer battery life. Microsoft's hardware requirements made sure you couldn't use a cheap low capacity SSD but had to use an hard-disk drive with more capacity that would still be relatively cheap just so it could run their bloatware. Then there was the licensing cost Microsoft imposed on the netbook vendors which eliminated any margin the vendors were supposed to have.
You do not need a warp drive or propellantless drive or whatever if all you want is to get to Alpha Centauri or Tau Ceti.
Nuclear pulse propulsion or anti-matter catalyzed nuclear pulse propulsion is good enough. The problem is it takes a really long time to get there e.g. 100 years to Alpha Centauri. You have to spend time accelerating then you need to decelerate once you reach the target. However you could have waystations along the way to make each leg of the trip shorter: e.g. the Oort cloud is a spherical shell of icy bodies supposedly nearly a light year away from the Sun. The destination may have a similar cloud of its own. There may even be a rogue planet, brown dwarf or some other large mass en route.
Once a viable destination exists it is only a matter of time before a probe is sent. It will be a challenge to make it have the necessary longevity but it is certainly not impossible. If the probe detects anything worthwhile there I am pretty certain someone will find a way to get there even with current technology.
the open-source development model cannot provide guaranteed deadline compliance, will not respect NDAs covering proprietary electronics design, and provides no-one to sue in the event of a serious flaw. They'd have to set up their own in-house development team to work on coreboot customisations - a difficult and thus risky prospect, as well as expensive.
This can be solved by creating a company like Cygnus Solutions or CodeSourcery that adapts coreboot to your motherboard for a fee.
What Ubuntu did was very unsatisfactory. You still cannot easily compile your own kernel. What that ex-RedHat guy did was a lot better since you can load anything you want as long as you confirm your choice on boot.
Here is what RMS should be doing instead of this petition which is going to get nowhere:
1. Restart work on coreboot
2. Make coreboot work with Windows and Linux as is
3. Convince more motherboard manufacturers to support coreboot
4. Ask Linux users on install if they want to backup their old BIOS and install coreboot as their default BIOS
You seem to think that if the course was online things would be fine. However it is more complicated than that. Many universities have online classes you could follow (e.g. Stanford) but the dirty secret is a good student will be ahead of the material taught in classes anyway. This is done by reading the books *before* the teacher discusses the material. Then the classes are only used to hash out fine details you did not quite figure out. Teachers have after school schedules for taking questions so this does not even need to happen inside class. If you are in a CS course they will answer e-mails with questions as well. It is also a good idea to get someone else's notes from class and study those. Universities usually do not force you to attend classes. I skipped nearly all theoretical classes starting in my second year and only bothered attending labs and things like that then did the exams. I did this because it was really cumbersome for me to commute to college. Grades will suffer if you do this but it is possible to finish a degree like this. If you have the time and inclination to study by yourself. One good time estimate, in my experience, is take the time listed in the class schedule and double that.
It will be very difficult to find the time to study while you are working a full time job at the same time. The only people I know who managed to work and study at the same time had part-time jobs close to the university grounds and they lived nearby (their commutes took less than 30 minutes). If your company has 50k employees they probably have a lot of students in there and some working place close to university grounds. It may be possible also that they will allow you to take an extended leave of absence from work to study but I doubt it will be long enough to finish an undergraduate course.
If you really want an undergraduate degree just look for a college near work and stick to working part-time.
Slashdot Mirror
Yes and the funny thing is people keep falling for it again and again. This Nokia business is just one more example of it. Heck even the Simpsons made fun of it once.
Yes it's a flat rate where Nokia isn't selling cellphones anywhere near the amount they thought they would so the result is $60 paid per phone i.e. more than what the other Windows Phone manufacturers pay. I honestly cannot understand how you think this was a good deal for Nokia. It wasn't and anyone with half a brain could have seen it at the time.
I don't think the licensing costs are as low as you claim they are. I have heard from several sources that due to the way the licensing was structured Nokia is paying a lot more per unit than the other Windows Phone vendors. I have also heard that they have paid back in royalties to Microsoft any cash that Microsoft has provided to them. Which is not particularly hard to believe since that article claims they are paying $250 million per quarter on licensing costs alone.
Expensive, poor software catalog, poorly marketed. Next!
DEC is more complicated to solve. They missed the dot com boom despite having perfectly fine hardware. They did not market Alpha servers as well as Sun advertised SPARC. They did not manage to keep their CPU manufacturing plant to full utilization unlike IBM who manufactures chips for other companies. Their PC server division could also be considered non existent. Their desktops were constantly more expensive and more obsolete than the competition. That is basically it.
1. Sell N9.
2. Profit.
3. Make cheaper versions of the N9 and migrate those to the low end market.
4. Profit.
5. Keep improving MeeGo or whatever its name is today.
6. Release new top end phone.
7. Go back to step 1.
Yeah really difficult.
Android has better licensing costs and allows more customization than Windows Phone. It has more apps. What reason was there to pick Windows Phone anyway? Besides Elop having come from Microsoft that is.
Actually Qualcomm designs their own CPU core (Krait) and GPU (Adreno).
GlobalFoundries got AMD's old fabs at Dresden. Since AMD fabs their CPUs at GlobalFoundries they are basically still using the same factories. The GPUs are still fabbed at TSMC just like ATI used to do. They are supposedly going to fab an APU (CPU/GPU combo) at TSMC but I am unaware if this has happened yet or not.
No I meant exactly what I said. Microsoft is no stranger to doing this sort of thing. They also did a nice little FUD campaign to camouflage their actions. Not that the FUD went 100% like MS wanted but MS still managed to kill Linux on netbooks. And netbooks.
The same thing they did to the netbooks. They will try to get the manufacturers to pay a fee to Microsoft for every Chromebook they sell. The fee will of course be greater than the cost of a Windows ilcense. If the manufacturers balk at it they will just threaten to increase the cost of the Windows licenses for the Windows laptops these manufacturers already sell. Google better get some non Windows laptop manufacturers making these things fast.
Samsung has been following the ARM roadmap pretty much faithfully so they aren't copying anything in chip design from Apple. Their core design is just an ARM licensed core. Which BTW is as good or better than the A6X core (triple-issue out-of-order). The A7 core remains to be seen.
same chip is being produced (it's Apple's design), it's just a different manufacturer
Tell that to AMD. They have been trying to outsource CPU manufacturing for years and they kept failing along the way. First AMD wanted to outsource to Chartered which was supposedly using the same manufacturing process (developed jointly by IBM, AMD, Samsung, Chartered) but it turned out they couldn't just trivially port their design over. Then they considered switching to TSMC. Another fail. They also considered switching GPU manufacturing from TSMC to GlobalFoundries after purchasing ATI. Yet another fail. The more low level optimizations the chip has the harder it is to port it. You don't just hit a compile button and then the thing magically works. Each manufacturing process has its own little details you have to work around in order for the design to be manufactureable and hit the right performance and power consumption targets.
What the article does not say is that it usually takes 12 to 18 months for a processor to go from tapeout to final production. Just because they have a design ready for TSMC to manufacture does not mean the first run will be bug free not to mention that the latency between a wafer going in and coming out is larger than a lot of people realize.
Not to mention the whole C issue where pointers to something and arrays of something are sort of the same but not really.
This is the problem you get when your strings don't know their allocated size like in that ghastly language Pascal.
The netbook was nothing but a quick bait and switch by manufacturers that wanted to make a quick buck off of the recession. The image of a business person using a netbook is just that. Users of netbooks were people with little money looking for a new toy, and nothing more.
Try reading the whole thread dear rabbid downmoderators.
Moore's law means the capacity would rise eventually to 16GB or 32GB which is what a tablet has today.
So in your mind 7" tablets are a bait and switch as well.
The reason for the downfall was indeed Microsoft. The original EeePC came with a Celeron processor (not Atom) and an SSD. It had a longer battery life. Microsoft's hardware requirements made sure you couldn't use a cheap low capacity SSD but had to use an hard-disk drive with more capacity that would still be relatively cheap just so it could run their bloatware. Then there was the licensing cost Microsoft imposed on the netbook vendors which eliminated any margin the vendors were supposed to have.
You do not need a warp drive or propellantless drive or whatever if all you want is to get to Alpha Centauri or Tau Ceti.
Nuclear pulse propulsion or anti-matter catalyzed nuclear pulse propulsion is good enough. The problem is it takes a really long time to get there e.g. 100 years to Alpha Centauri. You have to spend time accelerating then you need to decelerate once you reach the target. However you could have waystations along the way to make each leg of the trip shorter: e.g. the Oort cloud is a spherical shell of icy bodies supposedly nearly a light year away from the Sun. The destination may have a similar cloud of its own. There may even be a rogue planet, brown dwarf or some other large mass en route.
Once a viable destination exists it is only a matter of time before a probe is sent. It will be a challenge to make it have the necessary longevity but it is certainly not impossible. If the probe detects anything worthwhile there I am pretty certain someone will find a way to get there even with current technology.
the open-source development model cannot provide guaranteed deadline compliance, will not respect NDAs covering proprietary electronics design, and provides no-one to sue in the event of a serious flaw. They'd have to set up their own in-house development team to work on coreboot customisations - a difficult and thus risky prospect, as well as expensive.
This can be solved by creating a company like Cygnus Solutions or CodeSourcery that adapts coreboot to your motherboard for a fee.
What Ubuntu did was very unsatisfactory. You still cannot easily compile your own kernel. What that ex-RedHat guy did was a lot better since you can load anything you want as long as you confirm your choice on boot.
Here is what RMS should be doing instead of this petition which is going to get nowhere:
1. Restart work on coreboot
2. Make coreboot work with Windows and Linux as is
3. Convince more motherboard manufacturers to support coreboot
4. Ask Linux users on install if they want to backup their old BIOS and install coreboot as their default BIOS
You seem to think that if the course was online things would be fine. However it is more complicated than that. Many universities have online classes you could follow (e.g. Stanford) but the dirty secret is a good student will be ahead of the material taught in classes anyway. This is done by reading the books *before* the teacher discusses the material. Then the classes are only used to hash out fine details you did not quite figure out. Teachers have after school schedules for taking questions so this does not even need to happen inside class. If you are in a CS course they will answer e-mails with questions as well. It is also a good idea to get someone else's notes from class and study those. Universities usually do not force you to attend classes. I skipped nearly all theoretical classes starting in my second year and only bothered attending labs and things like that then did the exams. I did this because it was really cumbersome for me to commute to college. Grades will suffer if you do this but it is possible to finish a degree like this. If you have the time and inclination to study by yourself. One good time estimate, in my experience, is take the time listed in the class schedule and double that.
It will be very difficult to find the time to study while you are working a full time job at the same time. The only people I know who managed to work and study at the same time had part-time jobs close to the university grounds and they lived nearby (their commutes took less than 30 minutes). If your company has 50k employees they probably have a lot of students in there and some working place close to university grounds. It may be possible also that they will allow you to take an extended leave of absence from work to study but I doubt it will be long enough to finish an undergraduate course.
If you really want an undergraduate degree just look for a college near work and stick to working part-time.