There are people who are reluctant to put effort into a project, which then may not be usable by them in some future situation. And, looking at the number of active contributors to LLVM, compared to GCC, I'd say there are more of these people than people who'd rather contribute to a GPL'd project. Sticking with the BSDL might improve chances of the project staying alive.
See? It works both ways. Oh, and for the record I get paid to write BSDL code fairly often, but I've never been paid to write GPL'd code (although I do sometimes get paid to write LGPL'd code).
Which doesn't contradict anything that the grandparent said. MIT and BSD licenses are GPL compatible. Any license that is more permissive than the GPL is GPL compatible, because the GPL only states that code may not have more restrictions placed on it than are present in the GPL (thus GPLv3 is incompatible with GPLv2, because it includes restrictions not present in GPLv2). You can include a BSDL component in a GPL'd project. There are a few BSDL projects hosted on GNU Savannah, although not very many.
The BSD license does not grant you the right to change the license of the code. It does, however, allow you to create new derived works that incorporate the existing work and are released under a new license, as long as you follow a few rules (basically, don't claim you wrote the BSD licensed parts). You can create a new project that has the GPL as an overall license, but some files under a BSD license, but you can't replace the BSD license with the GPL unless you are the copyright holder.
Well, he doesn't actually say what isotopes of either participate in the reaction, but the difference in mass between a neutron and a proton is about 0.14%. The relevant isotopes have about 64 nucleons, so we're looking at about 0.000022 grams of mass being lost in a 1g reaction if one neutron becomes a proton. Plugging that into e=mc^2 gives about 2 gigajoules. Unfortunately, one tonne of oil equivalent is defined as about 42GJ, so we're a long way off.
Well, both nickel and copper have a variety of different stable isotopes, and he hasn't actually said which ones are consumed and produced. This alone makes me somewhat suspicious, because that would let us know if the reaction is even theoretically possible...
No, a catalyst is just something that participates in a reaction, but is not consumed by it. For example, muons have been proposed as a catalyst for hydrogen fusion. If you replace the electron in orbit around a proton with a muon, you get an atom with a much smaller radius than a normal hydrogen atom and no charge. Moving two of these close enough together for the strong attraction to overcome the electrostatic repulsion requires a lot less energy than with normal hydrogen atoms. The only slight problem is that muons don't tend to last long enough to go into orbit around nucleons, and if you could persuade them to then you'd probably have some much easier ways of generating a lot of power.
The idea of chemicals that act as nuclear catalysts was a staple of science fiction in the '50s and '60s, but largely went out of fashion once people realised how improbable it was that you'd find one. A hypothetical catalyst might have a reversible decay, for example emitting a neutron in one transition and then absorbing a neutron in the other direction, so the neutron could hit another atom, cause fission, and then an emitted neutron from that decay would return the catalyst to its original state. As far as I know, there are no known isotopes with this property.
A fusion catalyst is a lot more difficult to imagine. That said, he isn't claiming to have achieved true fusion, so much as proton capture (which is potentially equally interesting), so for his reaction to work I think you'd need something that would absorb a neutron and then decay and emit a proton - a neutron capture followed by a beta decay followed by fission. Something that would do that would count as a catalyst, because it would return to its original state after the reaction.
Let's say they are a huge success. Just how long, and for what prices, will I be able to buy a bag of nickel to feed into the machine?
You seem to have no idea of the energy densities of nuclear reactions. A bag of nickel that you can lift would power an entire country for a year. If it works (big 'if') then the cost of nickel is not going to be a problem.
The real problem is that fusing hydrogen and nickel into copper is an energy-negative reaction. I just tried to do the sums to work out how much energy would be released, and came out with a negative number. If I'd checked the periodic table first, I'd have known to expect this - nickel and copper are both after iron in the periodic table, and anything after iron becomes more table by moving towards iron, not away.
That doesn't necessarily mean that this isn't some kind of energy positive nuclear reaction. For example, nickel 63 to copper 63 would be energy positive, but then you have to account for where two neutrons go - three if one of them doesn't become a proton - and I'm pretty sure they would notice if that many high-energy neutrons were being emitted...
Wow, that's probably close to $15K of hardware. I am insanely jealous. Those machines are still pretty expensive - they go for more than PowerPC Macs second hand - although that's probably more due to rarity than utility. I wish Apple would release the i486 version of OPENSTEP as a free download, like they did with MacOS 7.5 for people to run in emulators. NeXTSTEP / OPENSTEP is incredibly impressive when you consider that the machines it ran on were an order of magnitude slower than a first generation iPhone.
I remember downloading movies with a 26.4Kb/s modem (well, the modem was rated higher but that was all the phone line could support). They'd be uploaded in 1MB chunks to some of the free hosting sites. A load of us would each download a few of them, and pass the chunks on ZIP disk to someone who would burn a VideoCD of the final version for each of us. It took about a decade for the movie industry to give a legal way of doing something similar, and even then the client is crap and they only do streaming of a limited selection...
Virgin documents their throttling rules, and I've found that they do follow their policy. I can get 1.1MB/s on my 10Mb/s line until I go over the cap, then it drops to about 250KB/s. I've only managed to hit the cap twice: once when I had to upload about 20GB of video footage to my publisher (screencasts with no interframe compression) and once when I decided to watch a film on iPlayer HD in the early evening (you hit the cap after about 55 minutes at iPlayer HD bitrates - that's what's making me consider switching to the 30Mb/s plan at the moment).
Virgin Media just rolled out 100Mb/s here (only 10Mb/s up though), and in most urban parts of the UK, so it's not like BT has no competition. I'm on their 10Mb/s plan at the moment, but I'm quite tempted by something with a bit more upstream.
Virgin is a bit unusual. Most ADSL providers (who all use BT's backbone) do use caps, because BT imposes caps in their wholesale rates and doesn't. Oh, and you do have a cap - you can only upload 6000MB between 3PM and 8PM.
Steve Jobs didn't read those emails either. I tried emailing sjobs@apple.com and got a reply from a junior manager. Very occasionally, a mail would be forwarded all the way up the hierarchy and would get personal attention, but it was rare.
Have you ever written code in an object-oriented language? Ever used a graphical UI designer? Even used a web app? If so, you've used things that Steve Jobs pushed from research into the mainstream while at NeXT.
It actually has an abridged version. The full text from the icon (/Applications/TextEdit.app/Resources/Edit.icns has a 512x512 pixel version, which is very easy to read):
Dear Kate,
Here's to the crazy ones. The misfits. The rebels.
The troublemakers. The round begs in the square
holes. The ones who see things differently. They're
not fond of rules. And they have no respect for the
status quo. You can praise them, disagree with them,
quote them, disbelieve them, glorify or vilify them.
About the only thing you can't do is ignore them.
Because they change things.
Because other handset makers do? The desktop versions of Flash are free, but the mobile versions are not. Handset and tablet makers all pay Adobe for the player. Apple refused, probably because the Mac version sucks and they didn't trust Adobe not to make a crap version for the iPhone - Flash on my TouchPad performs about as well as on my Core 2 Duo MacBook Pro - much better when playing video.
Re:What he took away is more precious than given
on
Steve Jobs Dead At 56
·
· Score: 1
No, not really. Jailbreaking is possible because of security holes. The first iPhone jailbreak that I remember the details of used a vulnerability in Apple's PDF implementation. Any web site that served you a PDF could execute arbitrary code on the device. Was it exploited in the wild? No idea, but it shows that the platform is far from immune to malware. In the month or so between public disclosure and Apple fixing the flaw, any iDevice could have been exploited just from visiting a web page.
The locked-down App Store does very little to add security. Apple doesn't do security audits on submitted applications - they don't even ask for the source code - they just check a few things like whether you are calling any private APIs. The security of iOS apps comes from the sandbox subsystem (also present on OS X), not from the distribution model.
At NeXT, he was the one who brought object oriented programming and rapid application development into the mainstream. Both were ideas that were on his must-have feature list from the day he founded the company. Saying that he wasn't a technologist is just woefully ignorant.
Wow. Where did you go to school? The cheapest computers NeXT ever sold were around $5,000, well out of the budget of most high schools. It was hard enough for the company to sell them to universities...
Which wasn't really a problem. Sony just called it i.Link. Other people just called it IEEE1394.
USB 1 and FireWire weren't anything like competitors though. USB was very cheap to put in devices, while FireWire devices needed a full host chip. USB only ran at 11Mb/s (less before 1.1), while FireWire ran at 400Mb/s. FireWire supported isochronous transfer, USB didn't. FireWire placed very little load on the CPU, USB was a CPU hog at high transfer rates. The iMac had both: USB for things like mice, keyboards, MIDI devices, and so on, FireWire for external disks and video cameras.
The thing that spurred the PC industry to adopt USB but not FireWire was Intel putting USB controllers in their south bridge chips. If you used an Intel chipset, you got USB for the cost of connecting the pins to a port. If you wanted FireWire, you needed an extra chip. This made motherboards with FireWire significantly more expensive. For years, PCs were crippled if you wanted an external hard disk. You were limited to 11Mb/s via USB as the fastest external connector on most, while the few that supported FireWire found the bottleneck in the drive, not the interconnect.
I wouldn't say that Xerox were idiots. The licensing fees from the laser printer patents alone completely covered the cost of operating PARC. They also got a big chunk of Apple shares (not sure when - or if - they sold them) in exchange for the guided tour of the centre. I don't know how much they made from Ethernet, but I wouldn't be surprised if that added up to a lot as well...
Pixar before Jobs was a support division of LucasFilm that did some special effects. After Jobs, it was a company that made films. Everything from Luxo Jr. onwards was Steve Jobs' Pixar. Before Jobs, Pixar had not even made a short film, let alone a feature film, it had just done effects work.
Slashdot Mirror
There are people who are reluctant to put effort into a project, which then may not be usable by them in some future situation. And, looking at the number of active contributors to LLVM, compared to GCC, I'd say there are more of these people than people who'd rather contribute to a GPL'd project. Sticking with the BSDL might improve chances of the project staying alive.
See? It works both ways. Oh, and for the record I get paid to write BSDL code fairly often, but I've never been paid to write GPL'd code (although I do sometimes get paid to write LGPL'd code).
Which doesn't contradict anything that the grandparent said. MIT and BSD licenses are GPL compatible. Any license that is more permissive than the GPL is GPL compatible, because the GPL only states that code may not have more restrictions placed on it than are present in the GPL (thus GPLv3 is incompatible with GPLv2, because it includes restrictions not present in GPLv2). You can include a BSDL component in a GPL'd project. There are a few BSDL projects hosted on GNU Savannah, although not very many.
The BSD license does not grant you the right to change the license of the code. It does, however, allow you to create new derived works that incorporate the existing work and are released under a new license, as long as you follow a few rules (basically, don't claim you wrote the BSD licensed parts). You can create a new project that has the GPL as an overall license, but some files under a BSD license, but you can't replace the BSD license with the GPL unless you are the copyright holder.
Well, he doesn't actually say what isotopes of either participate in the reaction, but the difference in mass between a neutron and a proton is about 0.14%. The relevant isotopes have about 64 nucleons, so we're looking at about 0.000022 grams of mass being lost in a 1g reaction if one neutron becomes a proton. Plugging that into e=mc^2 gives about 2 gigajoules. Unfortunately, one tonne of oil equivalent is defined as about 42GJ, so we're a long way off.
Well, both nickel and copper have a variety of different stable isotopes, and he hasn't actually said which ones are consumed and produced. This alone makes me somewhat suspicious, because that would let us know if the reaction is even theoretically possible...
Which is moving away from iron, and therefore an energy negative reaction...
No, a catalyst is just something that participates in a reaction, but is not consumed by it. For example, muons have been proposed as a catalyst for hydrogen fusion. If you replace the electron in orbit around a proton with a muon, you get an atom with a much smaller radius than a normal hydrogen atom and no charge. Moving two of these close enough together for the strong attraction to overcome the electrostatic repulsion requires a lot less energy than with normal hydrogen atoms. The only slight problem is that muons don't tend to last long enough to go into orbit around nucleons, and if you could persuade them to then you'd probably have some much easier ways of generating a lot of power.
The idea of chemicals that act as nuclear catalysts was a staple of science fiction in the '50s and '60s, but largely went out of fashion once people realised how improbable it was that you'd find one. A hypothetical catalyst might have a reversible decay, for example emitting a neutron in one transition and then absorbing a neutron in the other direction, so the neutron could hit another atom, cause fission, and then an emitted neutron from that decay would return the catalyst to its original state. As far as I know, there are no known isotopes with this property.
A fusion catalyst is a lot more difficult to imagine. That said, he isn't claiming to have achieved true fusion, so much as proton capture (which is potentially equally interesting), so for his reaction to work I think you'd need something that would absorb a neutron and then decay and emit a proton - a neutron capture followed by a beta decay followed by fission. Something that would do that would count as a catalyst, because it would return to its original state after the reaction.
Let's say they are a huge success. Just how long, and for what prices, will I be able to buy a bag of nickel to feed into the machine?
You seem to have no idea of the energy densities of nuclear reactions. A bag of nickel that you can lift would power an entire country for a year. If it works (big 'if') then the cost of nickel is not going to be a problem.
The real problem is that fusing hydrogen and nickel into copper is an energy-negative reaction. I just tried to do the sums to work out how much energy would be released, and came out with a negative number. If I'd checked the periodic table first, I'd have known to expect this - nickel and copper are both after iron in the periodic table, and anything after iron becomes more table by moving towards iron, not away.
That doesn't necessarily mean that this isn't some kind of energy positive nuclear reaction. For example, nickel 63 to copper 63 would be energy positive, but then you have to account for where two neutrons go - three if one of them doesn't become a proton - and I'm pretty sure they would notice if that many high-energy neutrons were being emitted...
Wow, that's probably close to $15K of hardware. I am insanely jealous. Those machines are still pretty expensive - they go for more than PowerPC Macs second hand - although that's probably more due to rarity than utility. I wish Apple would release the i486 version of OPENSTEP as a free download, like they did with MacOS 7.5 for people to run in emulators. NeXTSTEP / OPENSTEP is incredibly impressive when you consider that the machines it ran on were an order of magnitude slower than a first generation iPhone.
I remember downloading movies with a 26.4Kb/s modem (well, the modem was rated higher but that was all the phone line could support). They'd be uploaded in 1MB chunks to some of the free hosting sites. A load of us would each download a few of them, and pass the chunks on ZIP disk to someone who would burn a VideoCD of the final version for each of us. It took about a decade for the movie industry to give a legal way of doing something similar, and even then the client is crap and they only do streaming of a limited selection...
RFC1149
Virgin documents their throttling rules, and I've found that they do follow their policy. I can get 1.1MB/s on my 10Mb/s line until I go over the cap, then it drops to about 250KB/s. I've only managed to hit the cap twice: once when I had to upload about 20GB of video footage to my publisher (screencasts with no interframe compression) and once when I decided to watch a film on iPlayer HD in the early evening (you hit the cap after about 55 minutes at iPlayer HD bitrates - that's what's making me consider switching to the 30Mb/s plan at the moment).
Virgin Media just rolled out 100Mb/s here (only 10Mb/s up though), and in most urban parts of the UK, so it's not like BT has no competition. I'm on their 10Mb/s plan at the moment, but I'm quite tempted by something with a bit more upstream.
Virgin is a bit unusual. Most ADSL providers (who all use BT's backbone) do use caps, because BT imposes caps in their wholesale rates and doesn't. Oh, and you do have a cap - you can only upload 6000MB between 3PM and 8PM.
Steve Jobs didn't read those emails either. I tried emailing sjobs@apple.com and got a reply from a junior manager. Very occasionally, a mail would be forwarded all the way up the hierarchy and would get personal attention, but it was rare.
Have you ever written code in an object-oriented language? Ever used a graphical UI designer? Even used a web app? If so, you've used things that Steve Jobs pushed from research into the mainstream while at NeXT.
Yes. Unfortunately, with Woz at the helm, it would be a company that failed to sell stuff that's technically cool and utterly unhip...
Dear Kate,
Here's to the crazy ones. The misfits. The rebels.
The troublemakers. The round begs in the square
holes. The ones who see things differently. They're
not fond of rules. And they have no respect for the
status quo. You can praise them, disagree with them,
quote them, disbelieve them, glorify or vilify them.
About the only thing you can't do is ignore them.
Because they change things.
Take Care,
John Appleseed
Because other handset makers do? The desktop versions of Flash are free, but the mobile versions are not. Handset and tablet makers all pay Adobe for the player. Apple refused, probably because the Mac version sucks and they didn't trust Adobe not to make a crap version for the iPhone - Flash on my TouchPad performs about as well as on my Core 2 Duo MacBook Pro - much better when playing video.
No, not really. Jailbreaking is possible because of security holes. The first iPhone jailbreak that I remember the details of used a vulnerability in Apple's PDF implementation. Any web site that served you a PDF could execute arbitrary code on the device. Was it exploited in the wild? No idea, but it shows that the platform is far from immune to malware. In the month or so between public disclosure and Apple fixing the flaw, any iDevice could have been exploited just from visiting a web page.
The locked-down App Store does very little to add security. Apple doesn't do security audits on submitted applications - they don't even ask for the source code - they just check a few things like whether you are calling any private APIs. The security of iOS apps comes from the sandbox subsystem (also present on OS X), not from the distribution model.
At NeXT, he was the one who brought object oriented programming and rapid application development into the mainstream. Both were ideas that were on his must-have feature list from the day he founded the company. Saying that he wasn't a technologist is just woefully ignorant.
Wow. Where did you go to school? The cheapest computers NeXT ever sold were around $5,000, well out of the budget of most high schools. It was hard enough for the company to sell them to universities...
Which wasn't really a problem. Sony just called it i.Link. Other people just called it IEEE1394.
USB 1 and FireWire weren't anything like competitors though. USB was very cheap to put in devices, while FireWire devices needed a full host chip. USB only ran at 11Mb/s (less before 1.1), while FireWire ran at 400Mb/s. FireWire supported isochronous transfer, USB didn't. FireWire placed very little load on the CPU, USB was a CPU hog at high transfer rates. The iMac had both: USB for things like mice, keyboards, MIDI devices, and so on, FireWire for external disks and video cameras.
The thing that spurred the PC industry to adopt USB but not FireWire was Intel putting USB controllers in their south bridge chips. If you used an Intel chipset, you got USB for the cost of connecting the pins to a port. If you wanted FireWire, you needed an extra chip. This made motherboards with FireWire significantly more expensive. For years, PCs were crippled if you wanted an external hard disk. You were limited to 11Mb/s via USB as the fastest external connector on most, while the few that supported FireWire found the bottleneck in the drive, not the interconnect.
I wouldn't say that Xerox were idiots. The licensing fees from the laser printer patents alone completely covered the cost of operating PARC. They also got a big chunk of Apple shares (not sure when - or if - they sold them) in exchange for the guided tour of the centre. I don't know how much they made from Ethernet, but I wouldn't be surprised if that added up to a lot as well...
Pixar before Jobs was a support division of LucasFilm that did some special effects. After Jobs, it was a company that made films. Everything from Luxo Jr. onwards was Steve Jobs' Pixar. Before Jobs, Pixar had not even made a short film, let alone a feature film, it had just done effects work.