We need faster processors not to do more things, or do the same things faster, but to simply maintain the same levels of productivity because our OSes and applications are ever increasingly bloated and inefficient. We need faster processors because the code is less efficient; and we don't write efficient code because we have faster processors. This is the loop of BS we've lived with for years.
I wrote a program in 1996 or so called busybox, which has been adopted by the embedded Linux industry. It, the Linux kernel, and the Almquist shell make a pretty comfortable command-line environment. At the time I wrote it, the whole environment would boot from a 1.44 MB floppy. It's still very small, even with the 2.6 kernel and the fact that 100 commands have been added to busybox since then, and runs in a lot of embedded devices. If you want tiny environments, they are still available, and some of us have not forgotten how to write them.
I wrote 29000 bit-slice microcode in the 80's at Pixar, and could sit all day working on getting the last clock tick possible out of a single tight loop. That CPU ran the predecessor of MMX at 40 MHz. Today, if I did that it might not even matter, because once you get over a certain speed, you can't read and write RAM as fast as the CPU can demand it.
It's not really that we need more powerful machines to do the same thing. It is that we are using every bit of power that is offered to us. For instance, 3D animated film frames used to take 5 minutes to render on a 250 MHz CPU. Today, with a 4000 MHz CPU, they take 5 minutes to render! But the animated characters now have realistic looking hair, which would never have been possible with that VAX.
It's true that I often write slower code these days, because I write more in Ruby than in C. But you can buy a new computer for what it would cost to have me work on something for three hours. So, it's a good trade.
So, I was teaching a class yesterday at a company that makes software. And at one point while answering a question, I blurted out "but then Java's a horrible language". And a few people seemed to be very surprised that I said that. At which point I added "but C and C++ are worse".
Yes, I did some image processing at Pixar. My point was that you can't read or write the memory as fast as you can do an integer compare. So, below the point at which memory bandwidth becomes the governing factor, the speed of your algorithm makes no difference.
People have 4 GHz CPUs these days and somebody is worried about the difference between !p and p==NULL? If any compiler actually assembles code with a speed difference for those examples these days, it might well be that the speed difference is one billionth of a second. And even if that code is in a tight loop, it's entirely irrelevant given that the memory bandwidth is probably the governing factor rather than how fast the CPU can do integer vs. 0 tests.
That tells us how it works, doesn't it? It must set up traps so that things that need to be emulated are caught, and then directly execute the code that would otherwise be emulated. This requires some systems programming smarts and a lot of knowledge of processor internals, but even if the module is never Open Sourced, it's possible for some third party to produce an Open Source equivalent.
As an ex-HP manager, I thought I'd pen some thoughts on Carly's exit.
First, don't feel sorry for Carly. She's made at least $50
Million from HP, probably more. This is a good time for her to
make her exit, whether or not she's had a disagreement with the
board.
Itanium, upon which HP has been a partner with Intel, is a
disaster. HP transferred its Precision Architecture (PA-RISC), the basis of
Itanium, to Intel, transferred its silicon designers to Intel,
shut down chip foundries that it had spent Billions to build. All of this was tied into their
Itanium partnership with Intel, which HP thought would be
producing the dominant microprocessor architecture. Now it is much more
likely that Itanium will return no significant revenue to HP.
Intel, eager to save their own
ship after having bet their company on Itanium, has transferred
Itanium innovations to their Pentium line, which they can do without any
significant return to HP. Indeed, due to Intel's court-compelled
cross-licensing with AMD, we might even see HP technology pop up there.
HP must be starting to see some delayed negative effects
of the merger - which was always a daring bid with many
naysayers. I think you can read IBM's attempted sale of its PC
manufacturing division to a Chinese company as an indictment of
the HP-Compaq merger strategy. Where HP chose
to "fix" a marginally profitable division at great expense, IBM
did not see that its forte was competing at the low end.
Over 6 or 7 years we have seen HP in a transition from a
high-margin to a low-margin company. Computers are becoming
commoditized, and the 70% margins that we used to pay for
workstations are gone forever. But now HP does have to compete at
that low end, a very difficult business requiring an almost
ruthless focus on efficiency that is opposite of the corporate
culture with which they went into this change.
There is also the problem that much of the innovation that
drove HP left when they spun off Agilent. That was a high-margin,
low-volume business that required a lot of innovation. It wasn't
very much like HP's main profit-centers, but it created a lot of
ideas that transferred to other divisions.
In both Nusphere v. MySQL and a German case regarding an embedded device, the judge has made a preliminary finding that the GPL was valid. I think both cases settled after that.
FSF offers past oblivion in exchange for present compliance. They just want you to remedy the infringement.
If you use my GPL work in your proprietary application, you are denying me payment. I put my work out to be used in the work of other people who would share their work on the same terms as mine. That's a sort of payment too. If you want to use my GPL work in your proprietary software, offer me money for a proprietary license. As copyright holder, I can give out any number of different licenses for my work. But not for the work of others, so those who pursue such dual-licensing plans must consider the licensing strategy that they will apply to contributed modifications.
That's a good point. If I had to crunch to render a film in time for Christmas release, I might go for 10,000 nodes on a moment's notice. Indeed, Sun once did lend Pixar a whole bunch of computers on very short notice, to do just that.
So, this could be a good proposition for bursty use. If you can forecast your needs, you might at least provide for your baseline capacity in a more cost-effective manner. But you could supplement that baseline with grid services.
I have a Dell dedicated server that I've never seen, in a rack somewhere in Texas, with really great networking and professional management. It costs $250/month. The folks who did my server recently did a 500-node install for a grid customer. It's not a partition in a Sun N1 grid. It's a high-end Intel CPU, a lot of memory, and reasonably good I/O, and I don't share that with anyone. I'm not clear what the bottom line difference would be between my system and the grid.
There is a maxim in film computer graphics: A frame always takes 5 minutes to render. If you add faster networking, better CPU, more memory, etc., the animators use it. They add enough complexity to the scene that you are back where you used to be.
So, instead of anything working faster we get realistic looking hair, which we were formerly unable to do.
So I'm not sure where we stand today regarding local disk.
At $744/month (24 * 31) you can buy at least three of the best Intel or AMD CPUs in a dedicated server along with the rest of the system including disk storage. I used to render movies. They were so I/O intensive that it did not work to have the data - including a lot of photographic textures - all live "on the network". It had to be on the local disk.
Odd you have sat on panels constituted to hand out the same magazine awards that you suggest are just media publicity vehicules ! Why ?
Because I'm an insider. I am nominated to those awards panels from time to time, and being involved, of course I see that activity for what it is. We create photo ops, get links from the front page of Slashdot, and the august members of the panel and the organization writing the check share the spotlight with the meritorious winners of the award.
I don't think I've ever had to spend more than 15 minutes on an award selection. I don't see that the other panelists are putting in more time than that.
I understand that OpenOffice is a very large contribution. Indeed, it should be the second most important program we have in Open Source. It should have a community to match that importance, and that would not obsolete anyone's job at Sun.
I despair at how badly Sun has botched strategy at every point in the entire OpenOffice saga. I spent a large part of Tuesday talking with two Sun representatives. They admit the strategic botches, and unfortunately they have so much trouble getting Sun's front office to listen that they are not able to prevent similar botches going forward.
The worst thing about Apollo was that its goal, though of course ambitious for the time, was too shallow. Land a man on the moon and safely return him to earth we did, and then ran out of goal and the motivation to go any farther. If the goal is not to establish a viable self-sustaining human presence in space, a permanent colony away from the perils of Earth, there is no point in sending more people out there. If the goal is just scientific exploration, robots are 1000 times more cost-effective.
They spent a lot of resources over many years and eventually got a settlement offer that they were comfortable with and could move on.
I was naive enough to believe that the plaintiffs in the anti-trust cases actually wanted to see justice done and were not simply in it for their own settlement. But this was not the case, even Ed Black took the money.
The blather was meant to convey that I was not taking you seriously. To understand why, let's revisit what you suggested I say in a PR statement
We admire Sun's ingenuity, but we're a little concerned about their lack of transparency. If they were to make their business a little more transparent, everybody would benefit from the collaboration that would result. By choosing not to adopt a philosophy of transparency, Sun is passing up an opportunity to create new, valuable partnerships.
Having read that, I have two choices. One is to believe that you are being sarcastic and you really do know better. But you write things like that consistently enough that I don't believe that's the case. What I am left with is the belief that you know nothing of my own success in doing PR for this movement over some years, and the fact that my success is due to the fact that I don't communicate in the manner you espouse.
It's just clueless PR-speak, not grand conspiracy by an evil genius.
Simon,
The outside PR person for that release is an Open Source expert that you may have met, and is quite clueful. She was doing what Sun asked of her and her company. She and Sun know that some simple facts will be taken out of the release by the press, and some more esoteric ones will not, and that because of the construction of the release the press would in almost every case overstate the importance of the patent release. I have done my best this week to counter that overstatement.
Regarding copyright aggregation, in every case I know of except for Sun, the copyright aggregator is an IRS-certified 501(c)3 charitable non-profit organization. Thus, there is a higher standard regarding the aggregator's benificient behavior than can be applied to Sun. Sun, too, announced their intention to create an outside foundation to hold the OpenOffice software, and then they reneged upon that promise.
Aggregation may have something to do with "governance", I'm not sure what, but Sun promoted its reason as having to do with defending the entire program in a court of law. That rationale still survives in garbled form in the licensing FAQ where it says The JCA ensures that Sun can defend license violations if necessary. Note that it doesn't say prosecute license violations. The original language was defend the entire program.
You should be aware that I have been critical of IBM and its patents. When Bob Macmillan of IDG brought a question of mine into an IBM executive presentation at LinuxWorld, it helped goad IBM into declaring that it wouldn't enforce its own patents upon the Linux kernel. I also did not refrain from mentioning IBM in connection with the Ravicher report, which was presented at my LinuxWorld press conference, and probably was a big factor in inspiring the IBM patent grant.
This was done because the requirements of the MPL seemed overly specific and burdensome
"Burdensome" means "we didn't want to do it". Had they wanted to leave the decision up to the individual Open Source project (presumably OpenSolaris), they would have left that appendix in place in skeletal form. I've never heard of anyone on the Mozilla project who considers that appendix to be burdensome.
Even the Mozilla community doesn't seem to consistently adhere to this practice.
It's critical that they do adhere to this practice, or otherwise the Gecko renderer could not be included in GNOME. That's the first one I thought of, no doubt there are other relationships like that which must be maintained through dual-licensing.
Oh, so you think I'm an anti-capitalist. When I was too young to take care of my own money. Actually, I used to be an ante-capitalist. Then I was a venture capitalist. See my resume. Now I'm not a venture capitaist any longer, so perhaps I'm an ad-venture-capitalist. I also own pieces of software companies here and there. Hm, that's property, isn't it? I must be a very virtuous person. I publish books for money, in fact I publish books with Open Source texts, and still make money on them. I must be some sort of gimme-gimme-gimme, cause there's certainly no other way to do that. And I am on the board of an insurance company. Definitely a pinko commie fag, nobody else would sit on the board of directors of an insurance company!
Saying you are stunned is a rhetorical device. And you're doing it wrong, anyway. Try this way:
I am stunned,
stunned that you...
It's important to italicize the second one. It gives it that breathless sort of flavor.
Slavery was the first one I thought of, and certainly it's the most obvious sort of property ownership that is held to be offensive these days but in an earlier day was legal, and socially acceptable - indeed the ownership of many slaves was held to be a sign of high social status at one time. As a society, we've grown up since then. But we still have much growing to do.
No doubt there are hundreds of forms of property ownership that our society has chosen to prohibit, limit, or tax.
So, your claim that software patents can wear some sort of mantle of virtue that is accorded to property ownership is bogus.
In fact, I am stunned, stunned, that you even hold such an opinion.:-)
Well, that's kind of implicit in any culture that accepts the right to own property, isn't it?
No. Our society has decided to throw off certain forms of property ownership. For example, we had a legal institution of slave ownership, and a civil war when Southern states felt that the Northern-dominated government acted illegally to restrict their property ownership. [Civil war buffs, I know there were other reasons too.]
It's moot that the legislature is only allowed to prohibit, not to allow. Indeed so moot I don't understand why you would bring it up except to obfuscate. In the absence of prohibition, all is legal. In the presence of prohibition, the legislature or courts act to make something legal by removing a prohibition. They do not in general go back and reverse all prior verdicts and administrative actions when they do so, and so we have the situation that an action generally held to be illegal becomes legal as a result of a court finding.
There was a combination of law and policy of the institution that grants the legal monopoly of the patent that made software patenting illegal until some 20 years ago. This is not an ancient property law embedded in tradition.
Over at Groklaw they have copies of some contracts between MS and Sun, including the "Stand-Still Agreement" which binds Sun to not interfere while MS sues OpenOffice developers. These would also be at Edgar, as they were part of a 10-Q report or some such. The stand-still agreement is especially irksome because those same developers assigned their copyrights to Sun. Why? Because Sun said they needed to own the entire copyright so that they could defend the entire program. I feel there was some bad faith in even making the MS agreement after accepting those copyrights on that premise. But I suppose that a choice between a Billion dollars and repudiating their word was pretty easy for them to make.
So, when Scott McNealy says that they're not buddies, I suggest you look at the contracts and at the money.
If Sun didn't plan to do anything in the face of Linux, they would not have promoted their grant as one to all Open Source developers.
It would help defuse the "Gimme, gimme, gimme" public image that you people have cultivated ever so diligently over the past few years.
You sound as if you are deliberately trying to be offensive. If you haven't noticed, we are the folks who have been creating software and giving it away with licenses designed to keep the software free for everyone and without the patent strings that Sun is imposing on the process. We just want the right to continue to give away our own work and have everyone use it as they please. The gimmie-gimmie-gimmie is coming from folks who think they have the right to own ideas and keep others from using those ideas.
You send the message that you're about the destruction of property rights that we've held as axiomatic for centuries.
You mean since approximately 1984. There was no software patenting before the court case that made it legal in the U.S.. Even the U.S. patent office thought it was a bad idea. You can't possibly be that ignorant of history. You must just be trolling.
Slashdot Mirror
I wrote a program in 1996 or so called busybox, which has been adopted by the embedded Linux industry. It, the Linux kernel, and the Almquist shell make a pretty comfortable command-line environment. At the time I wrote it, the whole environment would boot from a 1.44 MB floppy. It's still very small, even with the 2.6 kernel and the fact that 100 commands have been added to busybox since then, and runs in a lot of embedded devices. If you want tiny environments, they are still available, and some of us have not forgotten how to write them.
I wrote 29000 bit-slice microcode in the 80's at Pixar, and could sit all day working on getting the last clock tick possible out of a single tight loop. That CPU ran the predecessor of MMX at 40 MHz. Today, if I did that it might not even matter, because once you get over a certain speed, you can't read and write RAM as fast as the CPU can demand it.
It's not really that we need more powerful machines to do the same thing. It is that we are using every bit of power that is offered to us. For instance, 3D animated film frames used to take 5 minutes to render on a 250 MHz CPU. Today, with a 4000 MHz CPU, they take 5 minutes to render! But the animated characters now have realistic looking hair, which would never have been possible with that VAX.
It's true that I often write slower code these days, because I write more in Ruby than in C. But you can buy a new computer for what it would cost to have me work on something for three hours. So, it's a good trade.
Bruce
Bruce
Bruce
Hey, my phone has a 400 MHz processor.
Bruce
Bruce
First, don't feel sorry for Carly. She's made at least $50 Million from HP, probably more. This is a good time for her to make her exit, whether or not she's had a disagreement with the board.
Itanium, upon which HP has been a partner with Intel, is a disaster. HP transferred its Precision Architecture (PA-RISC), the basis of Itanium, to Intel, transferred its silicon designers to Intel, shut down chip foundries that it had spent Billions to build. All of this was tied into their Itanium partnership with Intel, which HP thought would be producing the dominant microprocessor architecture. Now it is much more likely that Itanium will return no significant revenue to HP.
Intel, eager to save their own ship after having bet their company on Itanium, has transferred Itanium innovations to their Pentium line, which they can do without any significant return to HP. Indeed, due to Intel's court-compelled cross-licensing with AMD, we might even see HP technology pop up there.
HP must be starting to see some delayed negative effects of the merger - which was always a daring bid with many naysayers. I think you can read IBM's attempted sale of its PC manufacturing division to a Chinese company as an indictment of the HP-Compaq merger strategy. Where HP chose to "fix" a marginally profitable division at great expense, IBM did not see that its forte was competing at the low end.
Over 6 or 7 years we have seen HP in a transition from a high-margin to a low-margin company. Computers are becoming commoditized, and the 70% margins that we used to pay for workstations are gone forever. But now HP does have to compete at that low end, a very difficult business requiring an almost ruthless focus on efficiency that is opposite of the corporate culture with which they went into this change.
There is also the problem that much of the innovation that drove HP left when they spun off Agilent. That was a high-margin, low-volume business that required a lot of innovation. It wasn't very much like HP's main profit-centers, but it created a lot of ideas that transferred to other divisions.
Thanks
Bruce Perens
FSF offers past oblivion in exchange for present compliance. They just want you to remedy the infringement.
Bruce
Bruce
So, this could be a good proposition for bursty use. If you can forecast your needs, you might at least provide for your baseline capacity in a more cost-effective manner. But you could supplement that baseline with grid services.
Thanks
Bruce
Bruce
So, instead of anything working faster we get realistic looking hair, which we were formerly unable to do.
So I'm not sure where we stand today regarding local disk.
Bruce
Bruce
No, I've explained this in pieces on Slashdot and elsewhere. I should put them all together and publish that.
Bruce
Because I'm an insider. I am nominated to those awards panels from time to time, and being involved, of course I see that activity for what it is. We create photo ops, get links from the front page of Slashdot, and the august members of the panel and the organization writing the check share the spotlight with the meritorious winners of the award.
I don't think I've ever had to spend more than 15 minutes on an award selection. I don't see that the other panelists are putting in more time than that.
I understand that OpenOffice is a very large contribution. Indeed, it should be the second most important program we have in Open Source. It should have a community to match that importance, and that would not obsolete anyone's job at Sun.
I despair at how badly Sun has botched strategy at every point in the entire OpenOffice saga. I spent a large part of Tuesday talking with two Sun representatives. They admit the strategic botches, and unfortunately they have so much trouble getting Sun's front office to listen that they are not able to prevent similar botches going forward.
Bruce
Bruce
I was naive enough to believe that the plaintiffs in the anti-trust cases actually wanted to see justice done and were not simply in it for their own settlement. But this was not the case, even Ed Black took the money.
Bruce
Having read that, I have two choices. One is to believe that you are being sarcastic and you really do know better. But you write things like that consistently enough that I don't believe that's the case. What I am left with is the belief that you know nothing of my own success in doing PR for this movement over some years, and the fact that my success is due to the fact that I don't communicate in the manner you espouse.
Bruce
Simon,
The outside PR person for that release is an Open Source expert that you may have met, and is quite clueful. She was doing what Sun asked of her and her company. She and Sun know that some simple facts will be taken out of the release by the press, and some more esoteric ones will not, and that because of the construction of the release the press would in almost every case overstate the importance of the patent release. I have done my best this week to counter that overstatement.
Regarding copyright aggregation, in every case I know of except for Sun, the copyright aggregator is an IRS-certified 501(c)3 charitable non-profit organization. Thus, there is a higher standard regarding the aggregator's benificient behavior than can be applied to Sun. Sun, too, announced their intention to create an outside foundation to hold the OpenOffice software, and then they reneged upon that promise.
Aggregation may have something to do with "governance", I'm not sure what, but Sun promoted its reason as having to do with defending the entire program in a court of law. That rationale still survives in garbled form in the licensing FAQ where it says The JCA ensures that Sun can defend license violations if necessary. Note that it doesn't say prosecute license violations. The original language was defend the entire program.
You should be aware that I have been critical of IBM and its patents. When Bob Macmillan of IDG brought a question of mine into an IBM executive presentation at LinuxWorld, it helped goad IBM into declaring that it wouldn't enforce its own patents upon the Linux kernel. I also did not refrain from mentioning IBM in connection with the Ravicher report, which was presented at my LinuxWorld press conference, and probably was a big factor in inspiring the IBM patent grant.
Thanks
Bruce
This was done because the requirements of the MPL seemed overly specific and burdensome
"Burdensome" means "we didn't want to do it". Had they wanted to leave the decision up to the individual Open Source project (presumably OpenSolaris), they would have left that appendix in place in skeletal form. I've never heard of anyone on the Mozilla project who considers that appendix to be burdensome.
Even the Mozilla community doesn't seem to consistently adhere to this practice.
It's critical that they do adhere to this practice, or otherwise the Gecko renderer could not be included in GNOME. That's the first one I thought of, no doubt there are other relationships like that which must be maintained through dual-licensing.
Thanks
Bruce
We're not arguing. It's just contradiction :-)
Bruce
Saying you are stunned is a rhetorical device. And you're doing it wrong, anyway. Try this way:
It's important to italicize the second one. It gives it that breathless sort of flavor.
Slavery was the first one I thought of, and certainly it's the most obvious sort of property ownership that is held to be offensive these days but in an earlier day was legal, and socially acceptable - indeed the ownership of many slaves was held to be a sign of high social status at one time. As a society, we've grown up since then. But we still have much growing to do.
No doubt there are hundreds of forms of property ownership that our society has chosen to prohibit, limit, or tax.
So, your claim that software patents can wear some sort of mantle of virtue that is accorded to property ownership is bogus.
In fact, I am stunned, stunned, that you even hold such an opinion. :-)
Bruce
No. Our society has decided to throw off certain forms of property ownership. For example, we had a legal institution of slave ownership, and a civil war when Southern states felt that the Northern-dominated government acted illegally to restrict their property ownership. [Civil war buffs, I know there were other reasons too.]
It's moot that the legislature is only allowed to prohibit, not to allow. Indeed so moot I don't understand why you would bring it up except to obfuscate. In the absence of prohibition, all is legal. In the presence of prohibition, the legislature or courts act to make something legal by removing a prohibition. They do not in general go back and reverse all prior verdicts and administrative actions when they do so, and so we have the situation that an action generally held to be illegal becomes legal as a result of a court finding.
There was a combination of law and policy of the institution that grants the legal monopoly of the patent that made software patenting illegal until some 20 years ago. This is not an ancient property law embedded in tradition.
Bruce
So, when Scott McNealy says that they're not buddies, I suggest you look at the contracts and at the money.
If Sun didn't plan to do anything in the face of Linux, they would not have promoted their grant as one to all Open Source developers.
Bruce
You sound as if you are deliberately trying to be offensive. If you haven't noticed, we are the folks who have been creating software and giving it away with licenses designed to keep the software free for everyone and without the patent strings that Sun is imposing on the process. We just want the right to continue to give away our own work and have everyone use it as they please. The gimmie-gimmie-gimmie is coming from folks who think they have the right to own ideas and keep others from using those ideas.
You send the message that you're about the destruction of property rights that we've held as axiomatic for centuries.
You mean since approximately 1984. There was no software patenting before the court case that made it legal in the U.S.. Even the U.S. patent office thought it was a bad idea. You can't possibly be that ignorant of history. You must just be trolling.
Bruce