The only problem with Apple including the name and email address in purchased music is that the iTunes UI won't let me filter on it for smart playlists! It's really useful information, and I want to be able to _use_ it to automatically separate the music that my wife buys from the music that I buy.
This is why I expect Apple will do everything they can to fight against people running a flexible system that can run whatever content they want on their artificially cheap hardware. I would be surprised if Apple's lawyers didn't start coming out of the woodworks soon. Yeah, that's why an Apple spokesman was quoted as saying "it's your box, do with it what you please -- but be mindful of voiding that warranty" a couple days ago. See http://www.engadget.com/2007/04/05/apple-not-fight ing-back-against-apple-tv-hacks/.
2 FPUs/ CPU * 1 floating point operation per cycle per FPU = 2 flop per CPU per cycle
2 flops per CPU per cycle * 2 Gcycles per second = 4 Gflops per second per CPU 4 Gflops/s per CPU * 2 CPU per machine = 8 Gflops/s per machine where does the extra 2x come from?
A fused multiply-add is f0 = f1 * f2 + f3, which is two floating point operations in a single instruction. Each FPU on a G5 can execute an FMADD each cycle. So:
That would depend on whether you would call Apple originally charging $1 a pop for using its FireWire trademark "gouging".
No, the $1 a pop (now down to $.25 per device), is the license fee for the IEEE 1394 patent pool. It's use of the actual technological guts of the port, and manufacturers pay it whether they call it FireWire or not.
_Most_ of that quarter per device goes to Apple, because they did most of the inventing involved and hold most of the patents in the pool.
This announcement doesn't change that fee structure at all. Allowing use of the name is completely unrelated.
but, lets just say treating things **very** optimistically, maybe os x blows everyone away- like (ahem) win95 and everyone buys into it. so then apple realize theyve made the big time and start porting immediately. os x 1.0 according to apple's website is supposed to come out january 01. lalor suggests that os x on x86 debuts in "early 2001". even with apple allegedly keeping os x easy to port, can the entirety of os x, gui and all, be ported in less than six months?
hahahahaha.
thats funny.
Dude, we're talking about the former NeXTSTEP OS codebase here. It originally ran on Intel, and the PowerPC version is the port. Apple has OS X running on Intel internally for every single build as they do development. There is no extra porting involved here.
I agree with you about Apple's control of hardware being the main reason for their lack of compatibility problems, and that Linux will obviously not be standing still over the next year, but if Apple wanted to ship an x86 version of OS X in Spring 2001, they could certainly do it. (It would be a mistake, IMHO, but it is totally technically feasible.)
Back in the OpenStep days (essentially the same technology as Cocoa, the technology we're talking about here), the OS ran on NeXT's 68k hardware, Intel x86, Sun Sparc, and HP workstation hardware. When building your app you could just check some checkboxes, gcc would do all the cross-compilation, and you'd end up with a single "fat binary" application package that ran on all 4 architectures.
Of course, if you were an idiot you could break things by introducing endian-dependencies in your code, etc, but the frameworks provided everything you needed to be sure that you were platform independent. You needed zero #ifdefs or other hacks. There were lots of shareware developers that couldn't afford hardware to test on all 4 archs but would release software anyway, and it generally just worked.
There is no doubt in my mind that were Apple to release an Intel version, our applications would be running correctly on it in less than a week (being a little conservative here.):-)
However, most programmers coming from the Mac world instead of the NeXTSTEP world are using Carbon (a cleaned up version of the venerable Mac toolbox) instead of Cocoa. Parts of Carbon are cross-platform already (i.e. Quicktime) but it wouldn't surprise me if developers had a much harder time porting their Carbon code than we would our Cocoa code.
I doubt that Apple is going to make a move to Intel hardware unless and until they either get most of their major developers writing their apps in Cocoa (and thus are portable), or in desperation if the AIM processor alliance suffers a total meltdown (Motorola continues to be unable to reach higher G4 clockspeeds or something like that).
I actually agree with a lot of what Katz is saying here. Where I think he is wrong, though, is in stressing interactivity. These companies need to concentrate on doing what they already do but in better ways. Sure the Internet is a different medium, but newspapers are still newspapers and book publishers are still book publishers. The New York Times web site wouldn't be improved if it ran the Slashdot code, IMHO. That's not what the NYT is about.
As far as book publishing goes: how many people really want to be reading those Choose-Your-Own-Adventure books? What else would an interactive novel look like? This isn't the way for publishers to go (perhaps game companies, but book publishers certainly don't have any of the expertise or infrastructure...)
So what should they be doing?
eBooks are never going to sell well until the hardware exists to make an eBook that has all the advantages of a paperback. Small size, take it anywhere, very good readability. Not least among the advantages: the publishers can make money with them. There are some projects out there working on "electronic paper" that'll look and feel like paper but be changable electronically. I think this is a ways off though, and the hardest problem will be figuring out how authors can get paid for books distributed electronically. It's just too easy to reproduce the bits.
What book publishers need to be aiming at in the short-term is books printed on demand. You go to the bookstore, or a web site, make your selection, and the book is printed, bound, and at the register or ready to ship in 15 minutes. The publishers and bookstores get to keep their business model, and no longer have to carry large stocks of inventory, a lot of which ends up wasted. Consumers can always find whatever title they want at the bookstore, and no book ever has to go out of print. Advantages for everybody.
(The parallel to how music can be sold or movies rented in this way should be obvious, and have even been mentioned in previous stories on/.)
Sort of on this topic: It is sad that Barnes & Noble chose to split off an entire different company for their web business in an unimaginative attempt to copy Amazon. B&N's big advantage was their potential integration with their brick & mortar stores, and they threw all of the possibilities away with this one decision.
Can we please be more careful about what words are thrown around here? There is a vast difference between trademarks, copyrights and patents. The slashdot item title says "copyright", which is totally wrong. The description and a bunch of the posts here say "patent", which is also totally wrong. Everyone here is aware that these are different things, aren't they? </rant>
This is a trademark dispute. You trademark a logo, a name, a distinctive look (like coke bottles, or UPS puke brown), et cetera. These are things that the general public notices to recognize your product. The purpose of trademark law is to keep you from purchasing some other product by mistake from a disreputable company that disguises its product as the one you actually want.
You're in a hurry, so you rush into the grocery store, grab a case of those distinctively shaped Coca-cola bottles, pay, and leave, and only when you get back to your car do you notice that you've actually bought some of my Greg's Sparkly Brown Water. (Costs the same, tastes like crap!) Without trademark law, there's nothing that Coca-cola can do about it, and I can profit off of unobservant consumers that are foolish enough to make this mistake. (Muahahahaha.)
Obviously there is a big gray area around the question: how similar do products have to be before ignorant consumers start to mistake one for the other? Beige boxes are so common that no one would think they're looking at a particular brand just because the box is square and beige. The judge who granted the preliminary injunction in this case, though, decided that Apple had done enough marketing, and the iMac look was distinctive enough, that consumers could be confused by the look of the eMachines and Daewoo boxes into thinking that they were iMacs and buy them by mistake.
This case was only about one product trying to impersonate another to take advantage of ignorant consumers, it has nothing at all to do with copyright or patent law, and nothing keeps anyone from building another all-in-one computer, or building a blue transparent computer - but trying to cash in on someone else's distinctive look is, IMHO, clearly wrong.
I don't know if I would say The Transparent Society is more upbeat, per se. David Brin essentially agrees with Scott McNealy, or will soon: you have no privacy, so get over it.
The $64,000 question for Brin, though, is whether it is only going to be the government and major corporations that have access to all sorts of personal data about you, or whether that information is going to be available to the general public as well.
His thesis (which is a good one, IMHO) is that losing privacy is inevitable (due to the march of technology), but that if it is a symmetric loss of privacy, if large corporations and governments can't get away with doing anything because they have no privacy either.... then a loss of privacy may not be a bad thing.
This ties in to some other posts made in this thread about passing laws enabling the public to know where telemarketers get your information, to be informed everytime your personal data is used, and to be aware of what databases exist on you, where, and for what purpose.
A world without privacy, but also without corruption, where you are aware of who is gathering or using information about you, even if you can't stop it -- it's not nirvana, but IMHO, it doesn't sound that bad. (Note that I, and Brin, are not saying we should trade privacy for security or anything like that - which would be stupid, IMHO. But if privacy is going to be technically impossible to achieve, let's try to make the best of it...)
Nah, the Chinese had that figured out long before Gutenberg, along with gunpowder and other nifty stuff.
What Gutenburg invented was movable type. The chinese had printing presses for hundreds of years before Gutenburg, but they'd carve a single block for a whole page at a time and print, instead of having a seperate block for each letter which could be reused in different combinations on each page.
Gutenburg had the major advantage of an alphabetic writing system instead of having to deal with pictographs. It makes a lot of sense that the Chinese never came up with movable type on their own, because they'd have needed thousands of different blocks.
So China had a few printed books and printed paper money long long before Europe, but once Gutenburg came around there were many more different books printed in Europe, the possibility of newspapers, et cetera.
This is the guy who provided the source code to ddos_scan right here. Even if he is behind the FBI's differently named tool as Roblimo is saying he obviously isn't asking you to run a tool without source code.
You need to ask whoever it is that is administrating the web site at the FBI why there isn't source code available.
Don't underestimate how nice it is to be fanless even on a normal desktop in an office somewhere. The machine is _quiet_.
The new iMacs are convection cooled without a CPU fan, and when you get used to them, they make normal desktop machines just plain unpleasant to be near.
I agree with the posters who are saying that open source software isn't communism, because communism is a mechanism for deciding how to distribute scarce resources, and that just doesn't apply to software which can be duplicated indefinitely.
I think the open source movement is really a parallel of Pacific Northwest Indian culture before the Europeans arrived. Food and raw materials were so plentiful and easy to get that you didn't have to work that hard to survive. Instead they held great polatches, feasted, and tried to outdo each other giving incredible presents. The culture was impressed by what you gave away, not by what you had - because just having was too easy.
Anyone in a position to really benefit from and appreciate open source software is usually in the same position: there is huge demand for the kind of work that we do, we generally don't need to worry about making enough money to eat or to live a decent life. We also all know how random and capricious the big money is - from the rise of Microsoft to the latest IPO, we all know that people aren't neccesarily getting rich through merit alone.
Giving away our skill and our time, and in such a way that other people within our little hacker sub-culture can be impressed with what we've done - that is what drives open source software. Because there is little else that does impress our peers.
This may be just my cynical take on human nature, but I think this pursuit of status is much closer to the true spirit of open source than any sort of uber-communism or desire to sacrifice ourselves for the good of others.(Not that this is a bad thing - using an innate human drive in a productive way for society is great, the same way capitalism mostly manages to use greed in a good way.)
Very interesting ideas about all sorts of things: auction markets in processor time, continually improving AI spam and spam-filter contests, and some of the philosophical and technological consequences of transfering consciousness into software.
Not to mention a good story.
I don't need ruggedized, just waterproof...
on
Rugged Laptops
·
· Score: 1
I carry a PowerBook around with me all the time, am not too careful with it, and I've never had any problems. Granted this is not tornado country.
The feature that appeals to me about this would be getting a waterproof laptop that I could use in the bath. I do my best thinking underwater.:-)
Dynamic Optimization rulz, should be used compiled
on
HotSpot arrives
·
· Score: 2
I agree that dynamic optimization is a very exciting concept and a lot of compiler research is going to go in that direction. However, if it gives Java an advantage over C/C++, it will only be because more resources are being spent improving the Java JVMs and compilers than are being spent improving the C/C++ compilers.
The initial demonstration of a lot of the concepts in HostSpot are from Self, a O-O prototype-based language, whose implementation does a lot of dynamic optimization. In the Self implementation the optimization actually takes place on machine instructions and not via intermediary VM instructions (which is why the Self implementation is closely tied to Sparc procesors). So having a VM-based language is not a prerequisite for doing these sorts of optimizations.
The future of compilation, IMHO, is compiling your code with profiling information included, running your app for awhile, and then feeding the profile back into the compiler for more information to generate your final optimized executable.
Granted that C/C++ aren't dynamic O-O languages, so many of the specific optimizations that Self/HotSpot are doing won't apply, but I still think there is a lot that could be done. And compiled Java, Objective-C, Eiffel, et cetera could all benefit from the HotSpot optimizations done with compiling and profiling as above.
The study is supposed to be directed towards IT professionals and assorted PHBs that are involved in setting future OS direction, coming up with scaling plans, hardware and software budgets for the next budget period, et cetera.
These people are asking (or at least should be): what should I be running six months to a year from now in order to handle the predicted traffic and server load?
Considering how quickly the semi-conductor industry is still moving, I think 1 GB quad-Xeon boxes are going to be more or less the standard new server box in a years time.
Of course, Win NT scales to 4 processors so badly that Microsoft is artificially retarding the development of SMP hardware, so I could easily be wrong and we'll be stuck with 2-way SMP for longer.
For example, there is almost no available Intel hardware bigger than 4-way SMP, even though handling more processors is well understood: Sun's entry-level(!) enterprise server E450 handles up to 8. There is nothing keeping the Intel hardware vendors from building bigger boxes besides the limitations of MS operating systems.
And the keyboard commands might seem weird, but you can certainly do everything from the keyboard - because the keyboard commands are exactly the same as Lynx.:-)
(Or at least a very early version of Lynx from several years ago - I don't think we've checked to make sure it still corresponds any time lately.)
MkLinux is a single server based off of a Mach 3.x kernel, MacOS X Server is a heavily bastardized Mach 2.2 kernel with a BSD 4.4 personality compiled in.
So MOSXS is architecturally a micro-kernel, but with the BSD personality built into the kernel so you don't have any context-switch overhead going between the microkernel and the BSD OS layer. To some degree this overcomes the drawbacks of having most of your functionality in servers instead of the kernel, while retaining the advantage of fast message passing that Mach is built for.
Supposedly Apple is taking their ex-NeXT folk with all their custom mods to Mach 2.2, putting them together with some of the MkLinux folk inside Apple, and will end up with a "best-of-both-worlds" optimized Mach 3.x microkernel for a future MacOS X release.
As far as a speed comparison goes, it wouldn't surprise me at all if LinuxPPC or even MkLinux are faster servers at this point. This is the first real release of the OS on PPC hardware, after all. I have access to a bunch of MOSXS machines (mostly Blue&White G3s) though, so if anyone wants to propose a particular test, I could probably set it up. . .
I work for a small company that does consulting and shrinkwrap software, and we write quite a bit of free software. Not only is it a lot of fun, there are three major business benefits:
Support: Not only do we sell support contracts, large corporations often just feel better paying for something, even if they could get it for free, because they'll have a vendor standing behind a purchased product.
Admiration: There is no better advertisement for a consulting organization than working code that your potential clients can look at and try out for themselves. We are known in our little niche primarily because of our free software offerings, and it gets us a lot of paying business.
Productivity: We get more work done for our clients in a shorter amount of time because we can reuse our free source code on every project, so we're only building the unique parts of the project instead of reinventing the wheel all the time. More productive means higher rates.
So we don't usually make money directly from free software, per se, it is more of an investment - it pays off over the long run for the aspects of our business that do make money.
People interested in working on this might want to look at OmniNetworking and OWF available from our website. It's free software but not Open Source - we had that argument here on Slashdot a couple months ago.:-)
The former framework is a BSD socket abstraction, and the latter is a set of generic stream / pipeline abstractions that does a lot of what this project sounds like it is aiming at. The basic idea is that you start with a source content, which can be stream-oriented or something simpler like a URL, and the framework constructs a series of translation steps to get from your source-stream to a target-stream in whatever format you ask for. Each translation step can be run in a separate thread - all the stream objects and processors are thread-safe. It provides functionality for treating files, http, ftp, etc as generic sources of streaming content.
It's all Objective-C code for OpenStep / MacOSX Server, so I doubt you'll be directly taking any of the source, but I think it has some good ideas about useful abstractions and ways to provide this functionality in a generic and powerful way. It might help for design ideas.
What I thought was most interesting about the interview is Moore saying that now his law is partly self-fulfilling because the industry drives R&D and plant investment to meet the curve.
Assuming that the law eventually collapses (I don't know enough of the hardware to guess why exactly...), what happens to long-range planning? What happens to all those Asian memory manufacturers that use Moore's law to make long-term plans? Seems like a lot of chaos is in the offing in the semi-conductor industry if enough people come to believe that they can't depend on Moore's law any more...
And what is Microsoft, and all the other makers of bloat-ware going to do when they don't have Moore's law to bail them out and make their products (almost) usable as they go through upgrade after bloated upgrade?
Slashdot Mirror
The only problem with Apple including the name and email address in purchased music is that the iTunes UI won't let me filter on it for smart playlists! It's really useful information, and I want to be able to _use_ it to automatically separate the music that my wife buys from the music that I buy.
A fused multiply-add is f0 = f1 * f2 + f3, which is two floating point operations in a single instruction. Each FPU on a G5 can execute an FMADD each cycle. So:
1 FMADD per cycle = 2 flop/cycle * 2 FPUs = 4 flop/cycle * 2 CPUs = 8 flop/cycle * 2 GHz = 16 Gflop/s
That would depend on whether you would call Apple originally charging $1 a pop for using its FireWire trademark "gouging".
No, the $1 a pop (now down to $.25 per device), is the license fee for the IEEE 1394 patent pool. It's use of the actual technological guts of the port, and manufacturers pay it whether they call it FireWire or not.
_Most_ of that quarter per device goes to Apple, because they did most of the inventing involved and hold most of the patents in the pool.
This announcement doesn't change that fee structure at all. Allowing use of the name is completely unrelated.
but, lets just say treating things **very** optimistically, maybe os x blows everyone away- like (ahem) win95 and everyone buys into it. so then apple realize theyve made the big time and start porting immediately. os x 1.0 according to apple's website is supposed to come out january 01. lalor suggests that os x on x86 debuts in "early 2001". even with apple allegedly keeping os x easy to port, can the entirety of os x, gui and all, be ported in less than six months?
hahahahaha.
thats funny.
Dude, we're talking about the former NeXTSTEP OS codebase here. It originally ran on Intel, and the PowerPC version is the port. Apple has OS X running on Intel internally for every single build as they do development. There is no extra porting involved here.
I agree with you about Apple's control of hardware being the main reason for their lack of compatibility problems, and that Linux will obviously not be standing still over the next year, but if Apple wanted to ship an x86 version of OS X in Spring 2001, they could certainly do it. (It would be a mistake, IMHO, but it is totally technically feasible.)
Back in the OpenStep days (essentially the same technology as Cocoa, the technology we're talking about here), the OS ran on NeXT's 68k hardware, Intel x86, Sun Sparc, and HP workstation hardware. When building your app you could just check some checkboxes, gcc would do all the cross-compilation, and you'd end up with a single "fat binary" application package that ran on all 4 architectures.
:-)
Of course, if you were an idiot you could break things by introducing endian-dependencies in your code, etc, but the frameworks provided everything you needed to be sure that you were platform independent. You needed zero #ifdefs or other hacks. There were lots of shareware developers that couldn't afford hardware to test on all 4 archs but would release software anyway, and it generally just worked.
There is no doubt in my mind that were Apple to release an Intel version, our applications would be running correctly on it in less than a week (being a little conservative here.)
However, most programmers coming from the Mac world instead of the NeXTSTEP world are using Carbon (a cleaned up version of the venerable Mac toolbox) instead of Cocoa. Parts of Carbon are cross-platform already (i.e. Quicktime) but it wouldn't surprise me if developers had a much harder time porting their Carbon code than we would our Cocoa code.
I doubt that Apple is going to make a move to Intel hardware unless and until they either get most of their major developers writing their apps in Cocoa (and thus are portable), or in desperation if the AIM processor alliance suffers a total meltdown (Motorola continues to be unable to reach higher G4 clockspeeds or something like that).
I actually agree with a lot of what Katz is saying here. Where I think he is wrong, though, is in stressing interactivity. These companies need to concentrate on doing what they already do but in better ways. Sure the Internet is a different medium, but newspapers are still newspapers and book publishers are still book publishers. The New York Times web site wouldn't be improved if it ran the Slashdot code, IMHO. That's not what the NYT is about.
/.)
As far as book publishing goes: how many people really want to be reading those Choose-Your-Own-Adventure books? What else would an interactive novel look like? This isn't the way for publishers to go (perhaps game companies, but book publishers certainly don't have any of the expertise or infrastructure...)
So what should they be doing?
eBooks are never going to sell well until the hardware exists to make an eBook that has all the advantages of a paperback. Small size, take it anywhere, very good readability. Not least among the advantages: the publishers can make money with them. There are some projects out there working on "electronic paper" that'll look and feel like paper but be changable electronically. I think this is a ways off though, and the hardest problem will be figuring out how authors can get paid for books distributed electronically. It's just too easy to reproduce the bits.
What book publishers need to be aiming at in the short-term is books printed on demand. You go to the bookstore, or a web site, make your selection, and the book is printed, bound, and at the register or ready to ship in 15 minutes. The publishers and bookstores get to keep their business model, and no longer have to carry large stocks of inventory, a lot of which ends up wasted. Consumers can always find whatever title they want at the bookstore, and no book ever has to go out of print. Advantages for everybody.
(The parallel to how music can be sold or movies rented in this way should be obvious, and have even been mentioned in previous stories on
Sort of on this topic: It is sad that Barnes & Noble chose to split off an entire different company for their web business in an unimaginative attempt to copy Amazon. B&N's big advantage was their potential integration with their brick & mortar stores, and they threw all of the possibilities away with this one decision.
Can we please be more careful about what words are thrown around here? There is a vast difference between trademarks, copyrights and patents. The slashdot item title says "copyright", which is totally wrong. The description and a bunch of the posts here say "patent", which is also totally wrong. Everyone here is aware that these are different things, aren't they?
</rant>
This is a trademark dispute. You trademark a logo, a name, a distinctive look (like coke bottles, or UPS puke brown), et cetera. These are things that the general public notices to recognize your product. The purpose of trademark law is to keep you from purchasing some other product by mistake from a disreputable company that disguises its product as the one you actually want.
You're in a hurry, so you rush into the grocery store, grab a case of those distinctively shaped Coca-cola bottles, pay, and leave, and only when you get back to your car do you notice that you've actually bought some of my Greg's Sparkly Brown Water. (Costs the same, tastes like crap!) Without trademark law, there's nothing that Coca-cola can do about it, and I can profit off of unobservant consumers that are foolish enough to make this mistake. (Muahahahaha.)
Obviously there is a big gray area around the question: how similar do products have to be before ignorant consumers start to mistake one for the other? Beige boxes are so common that no one would think they're looking at a particular brand just because the box is square and beige. The judge who granted the preliminary injunction in this case, though, decided that Apple had done enough marketing, and the iMac look was distinctive enough, that consumers could be confused by the look of the eMachines and Daewoo boxes into thinking that they were iMacs and buy them by mistake.
This case was only about one product trying to impersonate another to take advantage of ignorant consumers, it has nothing at all to do with copyright or patent law, and nothing keeps anyone from building another all-in-one computer, or building a blue transparent computer - but trying to cash in on someone else's distinctive look is, IMHO, clearly wrong.
I don't know if I would say The Transparent Society is more upbeat, per se. David Brin essentially agrees with Scott McNealy, or will soon: you have no privacy, so get over it.
The $64,000 question for Brin, though, is whether it is only going to be the government and major corporations that have access to all sorts of personal data about you, or whether that information is going to be available to the general public as well.
His thesis (which is a good one, IMHO) is that losing privacy is inevitable (due to the march of technology), but that if it is a symmetric loss of privacy, if large corporations and governments can't get away with doing anything because they have no privacy either.... then a loss of privacy may not be a bad thing.
This ties in to some other posts made in this thread about passing laws enabling the public to know where telemarketers get your information, to be informed everytime your personal data is used, and to be aware of what databases exist on you, where, and for what purpose.
A world without privacy, but also without corruption, where you are aware of who is gathering or using information about you, even if you can't stop it -- it's not nirvana, but IMHO, it doesn't sound that bad. (Note that I, and Brin, are not saying we should trade privacy for security or anything like that - which would be stupid, IMHO. But if privacy is going to be technically impossible to achieve, let's try to make the best of it...)
I don't think they actually need to repeat the font tag in every paragraph, but they certainly do need to do it in every single table cell.
Silly, but true. Someone was not thinking a whole lot when they came up with the spec for font tags.
What Gutenburg invented was movable type. The chinese had printing presses for hundreds of years before Gutenburg, but they'd carve a single block for a whole page at a time and print, instead of having a seperate block for each letter which could be reused in different combinations on each page.
Gutenburg had the major advantage of an alphabetic writing system instead of having to deal with pictographs. It makes a lot of sense that the Chinese never came up with movable type on their own, because they'd have needed thousands of different blocks.
So China had a few printed books and printed paper money long long before Europe, but once Gutenburg came around there were many more different books printed in Europe, the possibility of newspapers, et cetera.
You need to ask whoever it is that is administrating the web site at the FBI why there isn't source code available.
Don't underestimate how nice it is to be fanless even on a normal desktop in an office somewhere. The machine is _quiet_.
The new iMacs are convection cooled without a CPU fan, and when you get used to them, they make normal desktop machines just plain unpleasant to be near.
I agree with the posters who are saying that open source software isn't communism, because communism is a mechanism for deciding how to distribute scarce resources, and that just doesn't apply to software which can be duplicated indefinitely.
I think the open source movement is really a parallel of Pacific Northwest Indian culture before the Europeans arrived. Food and raw materials were so plentiful and easy to get that you didn't have to work that hard to survive. Instead they held great polatches, feasted, and tried to outdo each other giving incredible presents. The culture was impressed by what you gave away, not by what you had - because just having was too easy.
Anyone in a position to really benefit from and appreciate open source software is usually in the same position: there is huge demand for the kind of work that we do, we generally don't need to worry about making enough money to eat or to live a decent life. We also all know how random and capricious the big money is - from the rise of Microsoft to the latest IPO, we all know that people aren't neccesarily getting rich through merit alone.
Giving away our skill and our time, and in such a way that other people within our little hacker sub-culture can be impressed with what we've done - that is what drives open source software. Because there is little else that does impress our peers.
This may be just my cynical take on human nature, but I think this pursuit of status is much closer to the true spirit of open source than any sort of uber-communism or desire to sacrifice ourselves for the good of others.(Not that this is a bad thing - using an innate human drive in a productive way for society is great, the same way capitalism mostly manages to use greed in a good way.)
Here's a link to it on amazon.
Very interesting ideas about all sorts of things: auction markets in processor time, continually improving AI spam and spam-filter contests, and some of the philosophical and technological consequences of transfering consciousness into software.
Not to mention a good story.
I carry a PowerBook around with me all the time, am not too careful with it, and I've never had any problems. Granted this is not tornado country.
:-)
The feature that appeals to me about this would be getting a waterproof laptop that I could use in the bath. I do my best thinking underwater.
I agree that dynamic optimization is a very exciting concept and a lot of compiler research is going to go in that direction. However, if it gives Java an advantage over C/C++, it will only be because more resources are being spent improving the Java JVMs and compilers than are being spent improving the C/C++ compilers.
The initial demonstration of a lot of the concepts in HostSpot are from Self, a O-O prototype-based language, whose implementation does a lot of dynamic optimization. In the Self implementation the optimization actually takes place on machine instructions and not via intermediary VM instructions (which is why the Self implementation is closely tied to Sparc procesors). So having a VM-based language is not a prerequisite for doing these sorts of optimizations.
The future of compilation, IMHO, is compiling your code with profiling information included, running your app for awhile, and then feeding the profile back into the compiler for more information to generate your final optimized executable.
Granted that C/C++ aren't dynamic O-O languages, so many of the specific optimizations that Self/HotSpot are doing won't apply, but I still think there is a lot that could be done. And compiled Java, Objective-C, Eiffel, et cetera could all benefit from the HotSpot optimizations done with compiling and profiling as above.
The study is supposed to be directed towards IT professionals and assorted PHBs that are involved in setting future OS direction, coming up with scaling plans, hardware and software budgets for the next budget period, et cetera.
These people are asking (or at least should be): what should I be running six months to a year from now in order to handle the predicted traffic and server load?
Considering how quickly the semi-conductor industry is still moving, I think 1 GB quad-Xeon boxes are going to be more or less the standard new server box in a years time.
Of course, Win NT scales to 4 processors so badly that Microsoft is artificially retarding the development of SMP hardware, so I could easily be wrong and we'll be stuck with 2-way SMP for longer.
For example, there is almost no available Intel hardware bigger than 4-way SMP, even though handling more processors is well understood: Sun's entry-level(!) enterprise server E450 handles up to 8. There is nothing keeping the Intel hardware vendors from building bigger boxes besides the limitations of MS operating systems.
You can see an announcement of a secure-HTTP plugin for OmniWeb at http://www.omnigroup. com/MailArchive/OmniWeb-l/1999/0185.html.
:-)
And the keyboard commands might seem weird, but you can certainly do everything from the keyboard - because the keyboard commands are exactly the same as Lynx.
(Or at least a very early version of Lynx from several years ago - I don't think we've checked to make sure it still corresponds any time lately.)
MkLinux is a single server based off of a Mach 3.x kernel, MacOS X Server is a heavily bastardized Mach 2.2 kernel with a BSD 4.4 personality compiled in.
So MOSXS is architecturally a micro-kernel, but with the BSD personality built into the kernel so you don't have any context-switch overhead going between the microkernel and the BSD OS layer. To some degree this overcomes the drawbacks of having most of your functionality in servers instead of the kernel, while retaining the advantage of fast message passing that Mach is built for.
Supposedly Apple is taking their ex-NeXT folk with all their custom mods to Mach 2.2, putting them together with some of the MkLinux folk inside Apple, and will end up with a "best-of-both-worlds" optimized Mach 3.x microkernel for a future MacOS X release.
As far as a speed comparison goes, it wouldn't surprise me at all if LinuxPPC or even MkLinux are faster servers at this point. This is the first real release of the OS on PPC hardware, after all. I have access to a bunch of MOSXS machines (mostly Blue&White G3s) though, so if anyone wants to propose a particular test, I could probably set it up. . .
Is here.
Anyone who is more legaleeze savvy care to comment on whether the previous brokenness has been fixed?
So we don't usually make money directly from free software, per se, it is more of an investment - it pays off over the long run for the aspects of our business that do make money.
(so I hope you'll forgive me.)
:-)
People interested in working on this might want to look at OmniNetworking and OWF available from our website. It's free software but not Open Source - we had that argument here on Slashdot a couple months ago.
The former framework is a BSD socket abstraction, and the latter is a set of generic stream / pipeline abstractions that does a lot of what this project sounds like it is aiming at. The basic idea is that you start with a source content, which can be stream-oriented or something simpler like a URL, and the framework constructs a series of translation steps to get from your source-stream to a target-stream in whatever format you ask for. Each translation step can be run in a separate thread - all the stream objects and processors are thread-safe. It provides functionality for treating files, http, ftp, etc as generic sources of streaming content.
It's all Objective-C code for OpenStep / MacOSX Server, so I doubt you'll be directly taking any of the source, but I think it has some good ideas about useful abstractions and ways to provide this functionality in a generic and powerful way. It might help for design ideas.
What I thought was most interesting about the interview is Moore saying that now his law is partly self-fulfilling because the industry drives R&D and plant investment to meet the curve.
Assuming that the law eventually collapses (I don't know enough of the hardware to guess why exactly...), what happens to long-range planning? What happens to all those Asian memory manufacturers that use Moore's law to make long-term plans? Seems like a lot of chaos is in the offing in the semi-conductor industry if enough people come to believe that they can't depend on Moore's law any more...
And what is Microsoft, and all the other makers of bloat-ware going to do when they don't have Moore's law to bail them out and make their products (almost) usable as they go through upgrade after bloated upgrade?