The StrongARM is basically an ARM with silicon fab improvements to make it go faster than the standard ARMs of the time.
I don't think this is correct. The StrongARM wasn't designed by ARM (the company), but designed by the same DEC CPU team that did the original DEC Alpha CPU. The SrongARM also has a different pipeline than does the more common ARM7 and ARM9 implementations of the ARM CPU architecture; so it's definitely not just a shrink. Intel ended up with the StrongARM design (and an IC fab line) in a big patent settlement with DEC.
And why shouldn't all embedded devices be considered one "work"? The device will usually not function without both the OS and the drivers; and the driver and OS are usually customized for exact hardware environment being shipped. In many embedded devices all software exists in one memory space, since there is no MMU hardware. The concepts of "userland" and "linking at runtime" don't seem to mean as much if the device performs one specific function and is the device is running as soon as it's manufactured.
The Google people (Schmidt, et.al.) gave a keynote talk at the IEEE Hot Chips conference this year. Basically, they depend on replication and software to get reliability from a vast array of commodity parts and rapidly failing disk drives. They have buyers who check the daily pricing on the cheapest motherboards and disk drives and buy in bulk.
One major limiting factor is a 500W per square foot limit in most hosting facilities (that's why a lot of their systems are still PIII based). But if a low-power blade cost only 20% more per "MIP", it might still be cheaper to pay the server facility for the extra cooling and power.
They said the HP/IBM/Dell salesmen just cry because there's no way those vendors can compete with the cheapest daily far-East motherboard import prices. The only salesmen who must like Google are the ones who sell them the diesel locomotives (err. backup power generators) when they exceed the power limit of some hosting facility.
3) They state that you cannot integrate GPL'd software with proprietery software.
This is true as well.
Not true. The GPL does not interfere with proprietary software business models if the GPL'd software is integrated as a separate process, "linked" via documented pipes, sockets, or ip connections. And these are all just virtual machine abstractions (where's the seperate process when I use Word and gcc together inside VirtualPC?)
The freedom in free software does not belong to the people that use the software, but rather to the software itself.
That's the ticket, enslave the people, and give their freedoms to a bunch of magnetic domains. I don't know about you, but I'd rather that the people be free, and that their tools (including magnetic and electostatic bits) be the slaves.
The m105 is *selling* for $99. To get an idea of what this product *costs*, remember that the m105 was introduced at a price of $250. It's selling now at what is probably a closeout price, with much lower profit margins all around. If the Zire follows the same product maturity pricing curve, it should be selling for under $40 somewhere around Christmas 2003.
Digital, can last forever--at least as long as someone gives them enough care to convert them. Just make sure to save them in an open loss-less format.
That's just the problem. If you don't convert them every couple media generation, no one but a museum specialist can read them.
I have a stack of CP/M formatted 16-sector 5.25" floppy disks in the garage... some Mac 400k single-sided diskettes. I'm pretty sure some of the files are in plain text format. But what do I read them with today? I also have a few photos my father took before the invention of the transistor. I can still look at them without any special tools.
This means that the whole cost of the global satellite network will have to be covered by the customers who cannot obtain the ground based two-way communications
Absolutely. So don't forget that most of the planet is covered by water, and a that lot of *very* rich people (as well as giant shipping companies) own boats... and islands and such.
... would still small potatoes compared to old fashioned "Big Iron" computing. A Cray 1S installation required minimum of.15 megawatts... and a cooling tower in the parking lot.
... would still small potatoes compared to real big iron. A Cray 1S installation required minimum of.15 megawatts, and a cooling tower in the parking lot.
From the perspective of free software, losing copyright isn't such a disaster. You couldn't compel people to cough up modified source code anymore (causing the GPL to behave more like BSD), but you'd simultaneously gain the right to freely distribute and/or plagiarize anything you wanted-- including proprietary source code that some disgruntled employee posted to usenet.
Loss of copyright protection doesn't imply that trade secret laws would also be invalidated. There's still a chance that the chain of possesion could be tracked and that you could be charged with receiving stolen property (unpublished secrets).
Well the beauty of RISC is the PII target performance can easily be ramped up to a P4 3G by simple manufacturing upgrades.
Don't tell that to the talented and experienced processor design groups at SiByte (now Broadcom) and SGI which had to spend serious effort to get the MIPs architecture to run at 1+ GHz with an appropriate performance scaling to match the clock frequency.
1a) Users - users are trusted because their mail is sent via SSL, and signed with a private key the user has (with the mail server having the public key).
That will just push the problem to that of fake private key registrations (via identity theft maybe). Maybe only the only private keys issued by the postal authority (after checking ones ID, and taking fingerprint and DNA samples) and only run on secure machines (Palladium?) should be trusted.
Maybe junk email is part of a conspiracy to convince people of the need to deploy Palladium?
Let's say I publish a program containing an XYZ. If I release my code under the GNU GPL, I thereby declare that recipients of the code have all the rights granted by the GPL. But I don't have the authority to grant them those rights, because some of the code is covered by the ZYX patents.
This is especially bad if my program contains code written by other programmers and released under the GPL. These programmers could then sue me for incorporating their code into software that is not freely redistributable as required by the GPL.
So, by this principle, anyone who contributed some unencumbered code to linux could now sue Linus because he incorporated that code into a kernel which includes some (very likely) patented memory management and TLB routines? Linus, and many other contributers, would lose the right to distribute linux. I wonder if Mumble$oft contributed any GPL'd code for just this purpose?
Actually, motion looks fairly fluid above 25 frames per second.
Motion *starts* to look fluid above 24 fps, but some sampling and judder artifacts are still detectable by the eye/brain up to around 130 fps. Action sports (basketball, hocky, etc.) are rarely filmed at just 25 fps because the either the blur or the sampling artifacts are quite visible. Film directors have to be very careful with the shutter speed and editing of actions scenes in 24 fps movies because of this. 60 field per second video is noticeably better for sports in terms of motion artifacts.
Open source hardware (as in schematics and HDL are public) will do absolutely no good. The problem is physical production. For last years 0.18 micron CMOS process, it would cost about $30k each for the 30 or so masks necessary for any fab to produce wafers for your IC. Who's going to cough up on the order of a million bucks for the first chip of that open-source, non-DRM CPU? And you'll not only need a CPU, but probably a system-bridge chip or two, and a non-DRM hard disk controller (even a low end IDE controller usually contains a RISC CPU and a custom servo DSP these days). We haven't even talked about the "cheap" stuff, like chip testing, packaging, multilayer motherboards, etc. So even if all the HDL, layout and verification were done for free by the open source community, it would still lead nowhere until someone spent on the order of 5 to 10 million USD.
Some say you could just program your system into an FPGA, but that will give you all the performance of a 12 year old computer, if that. Boy, that will be impressive to have a system with less performance than in the cheap handheld games kids buy in the toy stores...
The best bet will be to an write emulator for a non-DRM machine to run on the latest legal high performance machines.
The purpose of general purpose hardware is to execute your software. If some government manages to make some forms of this hardware illegal, it will still be possible to execute your software... with software.
To make all the software tools which allow any college student to write a hardware emulator would be impossible without killing off all software development. That would quickly kill the infrastructure of any modern country.
Yeah but how are the students gonna learn things like memory management and hardware control if they are using a managed, abstracted language like C# ?... You should start with something like C that teaches the fundamentals, then when you know how a computer *really* works, you can move on to a higher level language like C# or Java.
Uniprocessors execute code in a global address space using branch and jump op codes. The closest popular language with these features is good ole BASIC, with its global variables, GOTO statements, PEEK & POKE for hardware control, etc. Maybe it should be the first HLL taught to people who want a feel for what the hardware is doing (until multiprocessors become the dominant paradigm).
Of course, they cannot stop anyone from buying the disc and redistributing the contents.
Not everything on the disc is under GPL. They might even be able to get a "compilation" copyright on their particular mix of GPL stuff which is on the CD. A redistributor would have to rebuild a distribution from their own selection of open source components.
In the event that a license is not legally binding (i.e. a GPL or BSD style license), the terms revert to the default, which is "All Rights Reserved", which is more restrictive than what is being granted by the GPL or BSD license.
AFAIK, this hasn't been tested in court. (IANAL)
*flame-bait-on* Since it seem obvious that anyone who gives their software away without compensation has pretty much abandoned their copyright and should no longer have legal standing to restrict someone from using said software; Micro$oft should just petition congre$$ to pass a law clarifying this. Maybe Senator Holling$ would be intere$ted. *flame-bait-off*
Theft of copyrighted IP is only a crime in the U.S. because U.S. law makes it so. But a U.S. Copyright applies in the U.S. If Malaysia does not grant a copyright to the developer of the needed educational software, or modifies its national copyright laws so that they does not apply to local educational institutions located in Malaysia, then any copying is not pirating, but simply another form of fair use in Malaysia.
The U.S. limits access to copyrighted material because it thinks that encourages R&D. Other countries are perfectly free to try different policies. Then we can see which policy works best.
I've had a similar idea when it comes to making a log server. If it is only physically possible to write to the log server, then there would be no way someone could erase their tracks.
That's an easy one. Use a serial null-modem cable to connect to a log server that's not networked and not running a getty; and cut the TxD line from the log server so the it's write only.
Copyright protection is a privledge extended to writers and muscians in order to incite them to create content.
So test the hypothesis.
Did pre-1989 communist countries, where musicians and artist (supposedly) could not gain monopolistic profits from their creations, realise the creation of more or less great art and music than the capitalist countries with restrictive copyright laws during the same time period?
Slashdot Mirror
I don't think this is correct. The StrongARM wasn't designed by ARM (the company), but designed by the same DEC CPU team that did the original DEC Alpha CPU. The SrongARM also has a different pipeline than does the more common ARM7 and ARM9 implementations of the ARM CPU architecture; so it's definitely not just a shrink. Intel ended up with the StrongARM design (and an IC fab line) in a big patent settlement with DEC.
And why shouldn't all embedded devices be considered one "work"? The device will usually not function without both the OS and the drivers; and the driver and OS are usually customized for exact hardware environment being shipped. In many embedded devices all software exists in one memory space, since there is no MMU hardware. The concepts of "userland" and "linking at runtime" don't seem to mean as much if the device performs one specific function and is the device is running as soon as it's manufactured.
The Power4 is very expensive device to manufacture. Just the chip packaging alone probably cost more than a high-end PowerMac.
One major limiting factor is a 500W per square foot limit in most hosting facilities (that's why a lot of their systems are still PIII based). But if a low-power blade cost only 20% more per "MIP", it might still be cheaper to pay the server facility for the extra cooling and power.
They said the HP/IBM/Dell salesmen just cry because there's no way those vendors can compete with the cheapest daily far-East motherboard import prices. The only salesmen who must like Google are the ones who sell them the diesel locomotives (err. backup power generators) when they exceed the power limit of some hosting facility.
This is true as well.
Not true. The GPL does not interfere with proprietary software business models if the GPL'd software is integrated as a separate process, "linked" via documented pipes, sockets, or ip connections. And these are all just virtual machine abstractions (where's the seperate process when I use Word and gcc together inside VirtualPC?)
That's the ticket, enslave the people, and give their freedoms to a bunch of magnetic domains. I don't know about you, but I'd rather that the people be free, and that their tools (including magnetic and electostatic bits) be the slaves.
The m105 is *selling* for $99. To get an idea of what this product *costs*, remember that the m105 was introduced at a price of $250. It's selling now at what is probably a closeout price, with much lower profit margins all around.
If the Zire follows the same product maturity pricing curve, it should be selling for under $40 somewhere around Christmas 2003.
That's just the problem. If you don't convert them every couple media generation, no one but a museum specialist can read them.
I have a stack of CP/M formatted 16-sector 5.25" floppy disks in the garage... some Mac 400k single-sided diskettes. I'm pretty sure some of the files are in plain text format. But what do I read them with today? I also have a few photos my father took before the invention of the transistor. I can still look at them without any special tools.
Absolutely. So don't forget that most of the planet is covered by water, and a that lot of *very* rich people (as well as giant shipping companies) own boats... and islands and such.
Loss of copyright protection doesn't imply that trade secret laws would also be invalidated. There's still a chance that the chain of possesion could be tracked and that you could be charged with receiving stolen property (unpublished secrets).
Don't tell that to the talented and experienced processor design groups at SiByte (now Broadcom) and SGI which had to spend serious effort to get the MIPs architecture to run at 1+ GHz with an appropriate performance scaling to match the clock frequency.
That will just push the problem to that of fake private key registrations (via identity theft maybe). Maybe only the only private keys issued by the postal authority (after checking ones ID, and taking fingerprint and DNA samples) and only run on secure machines (Palladium?) should be trusted.
Maybe junk email is part of a conspiracy to convince people of the need to deploy Palladium?
This is especially bad if my program contains code written by other programmers and released under the GPL. These programmers could then sue me for incorporating their code into software that is not freely redistributable as required by the GPL.
So, by this principle, anyone who contributed some unencumbered code to linux could now sue Linus because he incorporated that code into a kernel which includes some (very likely) patented memory management and TLB routines? Linus, and many other contributers, would lose the right to distribute linux. I wonder if Mumble$oft contributed any GPL'd code for just this purpose?
Motion *starts* to look fluid above 24 fps, but some sampling and judder artifacts are still detectable by the eye/brain up to around 130 fps. Action sports (basketball, hocky, etc.) are rarely filmed at just 25 fps because the either the blur or the sampling artifacts are quite visible. Film directors have to be very careful with the shutter speed and editing of actions scenes in 24 fps movies because of this. 60 field per second video is noticeably better for sports in terms of motion artifacts.
Some say you could just program your system into an FPGA, but that will give you all the performance of a 12 year old computer, if that. Boy, that will be impressive to have a system with less performance than in the cheap handheld games kids buy in the toy stores...
The best bet will be to an write emulator for a non-DRM machine to run on the latest legal high performance machines.
The purpose of general purpose hardware is to execute your software. If some government manages to make some forms of this hardware illegal, it will still be possible to execute your software... with software.
To make all the software tools which allow any college student to write a hardware emulator would be impossible without killing off all software development. That would quickly kill the infrastructure of any modern country.
You should start with something like C that teaches the fundamentals, then when you know how a computer *really* works, you can move on to a higher level language like C# or Java.
Uniprocessors execute code in a global address space using branch and jump op codes. The closest popular language with these features is good ole BASIC, with its global variables, GOTO statements, PEEK & POKE for hardware control, etc. Maybe it should be the first HLL taught to people who want a feel for what the hardware is doing (until multiprocessors become the dominant paradigm).
Not everything on the disc is under GPL. They might even be able to get a "compilation" copyright on their particular mix of GPL stuff which is on the CD. A redistributor would have to rebuild a distribution from their own selection of open source components.
AFAIK, this hasn't been tested in court. (IANAL)
*flame-bait-on*
Since it seem obvious that anyone who gives their software away without compensation has pretty much abandoned their copyright and should no longer have legal standing to restrict someone from using said software; Micro$oft should just petition congre$$ to pass a law clarifying this. Maybe Senator Holling$ would be intere$ted.
*flame-bait-off*
What's to prevent this scenario?
Theft of copyrighted IP is only a crime in the U.S. because U.S. law makes it so. But a U.S. Copyright applies in the U.S. If Malaysia does not grant a copyright to the developer of the needed educational software, or modifies its national copyright laws so that they does not apply to local educational institutions located in Malaysia, then any copying is not pirating, but simply another form of fair use in Malaysia.
The U.S. limits access to copyrighted material because it thinks that encourages R&D. Other countries are perfectly free to try different policies. Then we can see which policy works best.
That's an easy one. Use a serial null-modem cable to connect to a log server that's not networked and not running a getty; and cut the TxD line from the log server so the it's write only.
The Artistic license is considered an Open Source License, but seems to allow the author greater control over forking.
So test the hypothesis.
Did pre-1989 communist countries, where musicians and artist (supposedly) could not gain monopolistic profits from their creations, realise the creation of more or less great art and music than the capitalist countries with restrictive copyright laws during the same time period?