"I rather have XP on an Eee and be open to all the possibilities of interaction with other peripherals"
And what peripherals would those be?
Video.., no Sound.., no Wireless.., no Network.., no Storage.., no Mouse.. no, Keyboard.. no (illustrating the way the thought process SHOULD be going).
Ooo, ooo, let me! Microsoft Fingerprint Scanner! HDTV USB input! Frankly, both of those (the first being external, the second requiring far too much storage) shouldn't be in consideration.
To give an example. I bought a media extender, and a high-def player. Models: D-Link DSM-320 and Toshiba A3. Both of these products use GPL licensed software (Linux) as a base.
My expectation is that I can change the firmware in these devices to add features or to use them for a different purpose. I paid money for these products; I feel that I received good value for the money. I would actually pay more for a product that I can upgrade (for example, I paid a bit more for a WRT54GL router).
I don't care about the "freedom of the software", or the "freedom of the developer". I don't care about "communism" vs "capitalism". What I care about is the freedom to load my OWN firmware onto the hardware. What is bizarre is that manufactures (even those using GPL code as a base) rarely get it --
GPL should be used to preserve hacking rights. Yes, it is "viral". I like that. This simply (and hopefully, effectively) encourages hacking and hackability.
The original Pentium (which went to 166Mhz, at the end, not just 75Mhz), used U and V execution pipes. No translation to micro-ops, and no "out of order". Indeed, there shouldn't be a need for that in Larrabee, anyway, given the number of cores. It would almost be better to get rid of the V pipe, and add SIMD, instead.
Your comments on CISC are bit off-base; the idea is to execute shaders in x86 machine code. They can be simple (limited flow control), or complex (general CPU/GPU).
"out-of-order" (ei. Pentium Pro and better) is not so good with that many cores doing that kind of work. It would get the hardware into a lot of trouble. Better to keep it simple, and add more cores.
A better start point would probably have been ARM, but that would lose the compatibility edge. If Larrabee works, it will take the GP-GPU market by storm. It needs:
1 - to publish itself as an NUMA access CPU (add a bit to tell the OS what it is for) 2 - compiler optimizations for the particular CPU architecture, preferably broken into two pieces: 2a - "straight line" shader code 2b - branching code 3 - a guide to the new NUMA characteristics.
With that in place, a standard (BSD/LINUX) OS will be able to use it for regular jobs. Or, for those special "I need the SIMD unit" jobs. The biggest hassle is trying to split control of those new CPU units between OpenGL and the regular scheduler (this is a kernel hack that Intel will have to make). It would be easier to jam this into OpenSolaris, but that isn't anywhere near popular enough.
Don't you want your video card to assist compiling large source when not gaming/modeling? Why not?
And, a few "extra" points
- Intel already has an optimizing compiler for the P54C architecture, and we have gcc. - The architecture, including U/V pipelines only used 3.1 million transistors. - A GeForce 7800 GTX has 302 million transistors -- 100x the number of the original Pentium processor.
So, I would think that using 32 "Pentium Classic" cores reduced would be quite feasible -- you need some (lots) of logic to ensure that they can all access their respective memories. The general SIMD implementation will take quite a bit of real estate as well. There is probably a budget of 600M transistors (wild ass guess) to Larrabee, estimate derived from power consumption estimates.
The gate size shrink should result in higher speeds. There may be a danger in the complex instruction interpretation routines, but these can be corrected. The single cycle instructions are already a (more than less) synchronous design, and should scale trivially.
Anything I am missing?
I, for one, am looking forward to buying a desktop super-computer with Larrabee.
You stated (I won't quote, just look at the parent post):
RMS is a criminal: WRONG - Microsoft is the criminal. RMS doesn't hold a position of importance: WONG - otherwise explain your reaction RMS doesn't live in the real world: What kind of crap is that? I presume that there is ANOTHER world that you are aware of?
All this to wrap an idea: that somehow the vast profits made by illegal activity are good, because people will be fed, and diseases cured.
I take it that you are an ethical relativist. After all, it's all relative...
Let's take an example from the "seamier side": an AIDs organization drops Artie Langes contributions because he indulges in gay-bashing. Right or wrong?
RMS views Microsoft as evil because Microsoft restricts software freedom. Simple fact.
Microsoft has done illegal things, that have brought them a LOT of money. Convicted for it. Simple fact.
Microsoft has made more money with illegal activity than Al Capone. Convicted for it. Simple fact.
You are happy to receive money from Microsoft. Simple fact.
In conclusion, you are happy to receive money obtained from illegal activities (of some sort). You are a moral relativist. QED.
Personally? I couldn't care less WHAT Microsoft sells. Especially now that I can buy x86 computers without Windows. I don't "hate" or even dislike Windows -- it just isn't my preference. I just much prefer Unix. (I did like OS/2, back in the day, though).
And, if I can have a platform that is (1) Unix, and (2) maintains software freedom, I choose that. Simple, really. But that's not the point! Here's the point:
I wouldn't accept a single dime from the B&MG foundation. I personally think it would be immoral. Different reason from the reason I don't (generally) use Windows, though.
Two different issues. My stance on both of them? Just say no. On the other hand, I will code for Windows. Doesn't break my moral stance, although it does break my software preference. Go figure.
No, I haven't ever figured it out -- I'll consider forking xmove and x2x, and doing a bit of maintaining. The largest issue is access to "off-brand" X servers for testing. As a general question: is this something that others want?
My experience: I worked on the Hummingbird X Server for Windows, including LBX proxy; but haven't been in the space for a decade. I would like generic xmove/x2x myself, I just want to know if its worth something to anyone else. I would need "off-brand" testing with other PC X Servers, Mac X Server, HP, IBM, old DEC. I work with SUN kit all day, and Xorg, but don't have easy access to XFree86. Send me email at fred_weigel@hotmail.com if you are interested, and have access to those X Servers, and are capable of using tunneling with ssh.
We are talking about terminology. And you just did it again - what "network mesh of computers"?
Please do not refer to a network as a mesh, unless it actually is. Treating a non-mesh as a mesh may well end up breaking things horribly. I give bittorrent as an example (which works, but is causing telcos to scream, simply because it is trying to use the internet AS a mesh).
I understand that network applications are changing, and the issue on the table is whether to refer to an interactive content that happens to layer on http as an extension of the web or as the web.
You are right in that regard. Simply put, it is what the users of it make it, or want it to be.
Interconnected computers is a "network". I have a network at home. The "topology" is the picture that these make when connected. That is, 5 computers are connected in a "star topology". 3 more in a "star topology", and there is a "bus topology" connecting the two "stars".
In turn, this network is connected to a cable modem, which connects this ENTIRE network to other networks. How? We are not sure of the topology, so we draw it as a "cloud".
This "network of networks" is the "internet".
Nothing to do with the data. That would be defined by public or private protocols. The "web" is defined as clients using servers with "http" protocol over "tcp/ip". It also defines a "URL" to allow linking from one server to itself or another, thus implementing hypertext documents.
Very useful stuff. But it only passes encoded data. Not food, water, or hygiene products. So I don't get where you get "need". It also doesn't solve problems. By facilitating information flow, the internet may provide you with data to solve a problem, but it's only data.
You are right -- the internet is not lynx and gopher. It is simply the idea that routing can be pushed to the edge, allowing networks to be trivially connected, and that the result would be useful.
In a sense, it is a fractal.
At the beginning I mentioned that I have two networks, each in a star topology, connected by a bus. That bus is the internet. Remember, the internet backbone started as a 56k link.
For convenience, we "users" of the internet allow certain functions to become part of the bedrock. As yet the "web" isn't there. What is there? ip, tcp, udp, dns, and routing protocols.
After all, we need a lingua franca; and dns is just too convenient to give us. Maybe the "web" will join in, but not until it loses bloat (as a hypertext publication method, not an application carrier).
As an example, I give you telnet. Once a noble and (considered) indispensable part of the "internet", it is now deceased. RIP.
Each choice point requires a core. You WILL run out of cores; it really doesn't matter how many you have... Unless, of course, you solve THAT problem, which scores you a Nobel Prize.
The Science of Algorithms. Expression, and analysis.
As far as I know, it doesn't have much to do with coding. And it sure doesn't have much to do with computer hardware. In fact, about the only "real" computer sciences job out there is teaching. Or, coming up with new algorithms.
Sometimes, you get to make a breakthrough; a single such breakthrough can make your career -- p-v semaphores and quicksort would be exemplars.
Which makes a "Computer Science" degree about as practically useful as a fine arts degree. Which, in turn, is why schools are tending to concentrate on teaching practical programming skills.
Now, you have indicated that you don't want to go into the programming end; which leaves the purer expression of Computer Science. Go for an advanced degree.
An interesting point -- but who created the internet and home computers for you?
Yep -- we are all now in our 50's and up.
But we didn't grow up on TV either -- the first TV in our family was used to watch the moon landing in '69. But there was no "cable"; we could only receive three stations. Wasn't worth watching, most of the time (except for exceptional events, like the moon landing).
The previous generation (take my mother-in-law - she's in her '70s) didn't see a TV until their late twenties/early thirties -- it certainly isn't a formative part.
Still, census disagrees with me a bit -- TV penetration in households in the USA was nearly complete by 1960 (I guess our family was a hold-out):
It may be that viewers born 1960 (and before) to 1970 (ei. those who did NOT start with cable) view TV programs as an "event" rather than as disposable entertainment, which may drive that demographic to watch first airings.
IBM did produce a split version of the Model M for a while -- had a big rotater knob to control the angle of the wedges. If you want to try it, it was known as the Model M15, and originally brought out in 1994, under the "Options" line:
So -- the problem is caused by non-symmetrical connections. This causes people who wish to exchange information to use "p2p". The efficient methods for controlling a swarm tend to be decentralized, by permitting the clients to ask for random sections, and discuss among themselves.
And -- we know the network grows by pushing routing decisions to the edge. Or, at least the "networking engineers" should know this. If the issue is that the ISP core routers are being swamped (rather than a bandwidth issue, as you imply), the solution seems evident. Push the routing back to the edge.
In other words, we have to solve the "last mile" problem to cure this. Obviously, allowing a monopoly for wired connections won't work, unless the organization granted that monopoly can guarantee service levels.
Another (hold on tight) solution is simply to equalize upload and download rates. This would make the swarm less effective (after all, much larger pieces could be downloaded from a single site).
But...
WAIT A MINUTE -- bittorrent doesn't ITSELF break pieces into MTU chunks! bittorrent loads at 128k+ (4mb isn't uncommon) chunk sizes.
This transacts 128 to over 4000 packets over the same path! Indeed, it's probably more efficient in terms of router use than automated web browsing.
Which means that the ISP simply cannot bear the load. And it doesn't matter whether that is caused by raw bandwidth or switching. Which brings us back to requiring the SLA.
SCREEN for persistent terminal connections, XMOVE for X programs (although xmove is a bit of a bear to use, and is suffering a slight case of bitrot) ftp://ftp.cs.columbia.edu/pub/xmove/ VNC is probably a better choice for "generic" use. VNC can be tunneled through SSH, which is better than opening a port for it (by itself, VNC isn't secure).
1 - "Geeks" do work around "latency" issues -- I have been doing so since the dialup era. I run my own DNS server for this reason, and use nscd on local machines. I use SquidProxy.
2 - I don't work around traffic shaping. But a packet is just a packet. What the fuck does the ISP think it is doing when (re)prioritizing my packets? I guess I simply don't deserve ANY priority, right?
3 - You didn't address your argument for keeping shaping private. Explain WHY it should be private. I will gladly add quality of service requests to my packets -- IF THE SHAPING POLICY IS NOT PRIVATE.
4 - "A few geeks demanding maximum bandwidth". Sure, why not? What does Bell DSL advertise?
"Technical Specifications: 720x480 maximum resolution, 30 fps, 1,500-2,200 kbps average bit rate. These videos will play on all 4th and higher generation Archos portable media players. THESE VIDEOS WILL NOT PLAY ON MOST PORTABLE VIDEO PLAYERS AND ARE NOT IPOD COMPATIBLE."
So this ISP does bandwidth shaping -- why? Aren't they selling Total Internet Performance, the Bell Better Home (tm), Audio downloads and Video (even streaming) at near HD quality?
A FEW GEEKS DEMANDING MAXIMUM BANDWIDTH??? Hell, I would like to see the beavers deliver a fraction of the promises. And let's not get into "never shared, never slow" crap they advertised.
It's worse. I don't CARE about "not paying". When I buy such a box (I have a DSM 320), I want the freedom to REPLACE the firmware. I want the manufacturer to make it EASY to replace the firmware; after all the firmware is based on GPL licensed material. This has nothing to do with "free as in beer".
I am perfectly willing to spend money buying the device. But I choose devices that use GPL software to allow me the freedom to modify those devices to suit my needs (for example, I want to add flash video to the DSM 320 -- it already has an internet connection,). This shouldn't bother the vendor (it did with Tivo, because Tivo sold a subscription, D-Link didn't), because the box has ALREADY BEEN PURCHASED.
So, make it easy, make it useful, and add a clause stating that the guarantee only applies if the original firmware has been restored.
I understand that the GPL may be incompatible with the "give away the razor, make money on the blades" model; but there are plenty of non-GPL alternatives (that means YOU, Tivo).
But if I buy the whole thing, I want to use as I see fit. And, for a general purpose device, that may go well beyond, or sideways to the intent of the manufacturer (eg. not a GPL example, but XBOX as media center).
I am a "coder" (well, an architect, but I occasionally do still code).
I do NOT get paid royalties or residuals. Any money I make "up front" on an hourly rate. I am willing to work for LESS, if the final product is GPL licensed. But, the works are "works for hire". I write hundreds of pages a year, and I am happy if my NAME remains on the material as author.
One of my co-workers is also an O'Reilly author (Using Samba). The authors insisted that the book material is available for free over the internet -- O'Reilly still makes money on the printed book (note that enough was made to go to at least a second edition).
This world is not comparable to the "Music industry". At least I make a living, even though I don't have royalties. (Perhaps it is: music acts go on tour.)
I don't work for free. If I couldn't make money from it, I would pick another job, and probably code and teach as a hobby.
The ISPs will lose revenue. The Government will step in to save them from bankruptcy. The "network effect" will be lessened. The "pirates" will find another method. Big advances in encrypted onion ring routing, and multi-site partial anonymous storage. Drop in "owned" systems, as less experienced users are simply cut off
Losers: ISPs and less experienced users. Government and other "snoops"
Winners: Other users due to decrease in SPAM. Peer to Peer forced evolution. Forced evolution in privacy preserving networks.
Neutral: "pirates"
To quote Nike: "Just Do It!". Stop threatening, and let's get this ball into play.
I would LOVE to be the CEO of a billion-dollar corporation. Is it ever going to happen? No, it won't. There, I'm disillusioned, and will become depressed.
Give me my happy pills! And get the hell off of my lawn!
So, a product produced in the US costs 1/2 in a foreign country. Am I allowed to now purchase Chinese product and sell in the US? To find out if this is worth it, let's get a retail (not msrp) price on Windows Vista Home Premium. Amazon lists some prices (no, we don't want the "upgrade" edition):
My understanding is that IF a case has been withdrawn that was based on the same matter as the case in question; the case then brought cannot be withdrawn without adjudication. So the judge can award costs?
Of course I don't live in the USA, so I may be totally off-base here.
Since the RIAA has ALREADY backed down from the related John Doe filing, they are not allowed to simply throw in the towel on this case. Procedure should force adjudication on the merits. Of course the Judge can decide to simply let them go home, or can judge the case anyway. So, the RIAA could stand to lose, EVEN IF THEY DON'T PURSUE. To quote Boston Legal: "I am the decider!". And, suitable compensation for the defense can be awarded (at the Judges discretion).
Bad form to back away from a case TWICE. Once, for the John Doe, and now, again.
Slashdot Mirror
"I rather have XP on an Eee and be open to all the possibilities of interaction with other peripherals"
And what peripherals would those be?
Video.., no Sound.., no Wireless.., no Network.., no Storage.., no Mouse.. no, Keyboard.. no (illustrating the way the thought process SHOULD be going).
Ooo, ooo, let me! Microsoft Fingerprint Scanner! HDTV USB input! Frankly, both of those (the first being external, the second requiring far too much storage) shouldn't be in consideration.
But, it's the POSSIBILITY that irks you, right?
The GPL is not communistic. It protects users.
To give an example. I bought a media extender, and a high-def player. Models: D-Link DSM-320 and Toshiba A3. Both of these products use GPL licensed software (Linux) as a base.
My expectation is that I can change the firmware in these devices to add features or to use them for a different purpose. I paid money for these products; I feel that I received good value for the money. I would actually pay more for a product that I can upgrade (for example, I paid a bit more for a WRT54GL router).
I don't care about the "freedom of the software", or the "freedom of the developer". I don't care about "communism" vs "capitalism". What I care about is the freedom to load my OWN firmware onto the hardware. What is bizarre is that manufactures (even those using GPL code as a base) rarely get it --
GPL should be used to preserve hacking rights. Yes, it is "viral". I like that. This simply (and hopefully, effectively) encourages hacking and hackability.
The original Pentium (which went to 166Mhz, at the end, not just 75Mhz), used U and V execution pipes. No translation to micro-ops, and no "out of order". Indeed, there shouldn't be a need for that in Larrabee, anyway, given the number of cores. It would almost be better to get rid of the V pipe, and add SIMD, instead.
Your comments on CISC are bit off-base; the idea is to execute shaders in x86 machine code. They can be simple (limited flow control), or complex (general CPU/GPU).
"out-of-order" (ei. Pentium Pro and better) is not so good with that many cores doing that kind of work. It would get the hardware into a lot of trouble. Better to keep it simple, and add more cores.
A better start point would probably have been ARM, but that would lose the compatibility edge. If Larrabee works, it will take the GP-GPU market by storm. It needs:
1 - to publish itself as an NUMA access CPU (add a bit to tell the OS what it is for)
2 - compiler optimizations for the particular CPU architecture, preferably broken into two pieces:
2a - "straight line" shader code
2b - branching code
3 - a guide to the new NUMA characteristics.
With that in place, a standard (BSD/LINUX) OS will be able to use it for regular jobs. Or, for those special "I need the SIMD unit" jobs. The biggest hassle is trying to split control of those new CPU units between OpenGL and the regular scheduler (this is a kernel hack that Intel will have to make). It would be easier to jam this into OpenSolaris, but that isn't anywhere near popular enough.
Don't you want your video card to assist compiling large source when not gaming/modeling? Why not?
And, a few "extra" points
- Intel already has an optimizing compiler for the P54C architecture, and we have gcc.
- The architecture, including U/V pipelines only used 3.1 million transistors.
- A GeForce 7800 GTX has 302 million transistors -- 100x the number of the original Pentium processor.
So, I would think that using 32 "Pentium Classic" cores reduced would be quite feasible -- you need some (lots) of logic to ensure that they can all access their respective memories. The general SIMD implementation will take quite a bit of real estate as well. There is probably a budget of 600M transistors (wild ass guess) to Larrabee, estimate derived from power consumption estimates.
The gate size shrink should result in higher speeds. There may be a danger in the complex instruction interpretation routines, but these can be corrected. The single cycle instructions are already a (more than less) synchronous design, and should scale trivially.
Anything I am missing?
I, for one, am looking forward to buying a desktop super-computer with Larrabee.
You stated (I won't quote, just look at the parent post):
RMS is a criminal: WRONG - Microsoft is the criminal.
RMS doesn't hold a position of importance: WONG - otherwise explain your reaction
RMS doesn't live in the real world: What kind of crap is that? I presume that there is ANOTHER world that you are aware of?
All this to wrap an idea: that somehow the vast profits made by illegal activity are good, because people will be fed, and diseases cured.
I take it that you are an ethical relativist. After all, it's all relative...
Let's take an example from the "seamier side": an AIDs organization drops Artie Langes contributions because he indulges in gay-bashing. Right or wrong?
RMS views Microsoft as evil because Microsoft restricts software freedom. Simple fact.
Microsoft has done illegal things, that have brought them a LOT of money. Convicted for it. Simple fact.
Microsoft has made more money with illegal activity than Al Capone. Convicted for it. Simple fact.
You are happy to receive money from Microsoft. Simple fact.
In conclusion, you are happy to receive money obtained from illegal activities (of some sort). You are a moral relativist. QED.
Personally? I couldn't care less WHAT Microsoft sells. Especially now that I can buy x86 computers without Windows. I don't "hate" or even dislike Windows -- it just isn't my preference. I just much prefer Unix. (I did like OS/2, back in the day, though).
And, if I can have a platform that is (1) Unix, and (2) maintains software freedom, I choose that. Simple, really. But that's not the point! Here's the point:
I wouldn't accept a single dime from the B&MG foundation. I personally think it would be immoral. Different reason from the reason I don't (generally) use Windows, though.
Two different issues. My stance on both of them? Just say no. On the other hand, I will code for Windows. Doesn't break my moral stance, although it does break my software preference. Go figure.
No, I haven't ever figured it out -- I'll consider forking xmove and x2x, and doing a bit of maintaining. The largest issue is access to "off-brand" X servers for testing. As a general question: is this something that others want?
My experience: I worked on the Hummingbird X Server for Windows, including LBX proxy; but haven't been in the space for a decade. I would like generic xmove/x2x myself, I just want to know if its worth something to anyone else. I would need "off-brand" testing with other PC X Servers, Mac X Server, HP, IBM, old DEC. I work with SUN kit all day, and Xorg, but don't have easy access to XFree86. Send me email at fred_weigel@hotmail.com if you are interested, and have access to those X Servers, and are capable of using tunneling with ssh.
We are talking about terminology. And you just did it again - what "network mesh of computers"?
Please do not refer to a network as a mesh, unless it actually is. Treating a non-mesh as a mesh may well end up breaking things horribly. I give bittorrent as an example (which works, but is causing telcos to scream, simply because it is trying to use the internet AS a mesh).
I understand that network applications are changing, and the issue on the table is whether to refer to an interactive content that happens to layer on http as an extension of the web or as the web.
You are right in that regard. Simply put, it is what the users of it make it, or want it to be.
Interconnected computers is a "network". I have a network at home. The "topology" is the picture that these make when connected. That is, 5 computers are connected in a "star topology". 3 more in a "star topology", and there is a "bus topology" connecting the two "stars".
In turn, this network is connected to a cable modem, which connects this ENTIRE network to other networks. How? We are not sure of the topology, so we draw it as a "cloud".
This "network of networks" is the "internet".
Nothing to do with the data. That would be defined by public or private protocols. The "web" is defined as clients using servers with "http" protocol over "tcp/ip". It also defines a "URL" to allow linking from one server to itself or another, thus implementing hypertext documents.
Very useful stuff. But it only passes encoded data. Not food, water, or hygiene products. So I don't get where you get "need". It also doesn't solve problems. By facilitating information flow, the internet may provide you with data to solve a problem, but it's only data.
You are right -- the internet is not lynx and gopher. It is simply the idea that routing can be pushed to the edge, allowing networks to be trivially connected, and that the result would be useful.
In a sense, it is a fractal.
At the beginning I mentioned that I have two networks, each in a star topology, connected by a bus. That bus is the internet. Remember, the internet backbone started as a 56k link.
For convenience, we "users" of the internet allow certain functions to become part of the bedrock. As yet the "web" isn't there. What is there? ip, tcp, udp, dns, and routing protocols.
After all, we need a lingua franca; and dns is just too convenient to give us. Maybe the "web" will join in, but not until it loses bloat (as a hypertext publication method, not an application carrier).
As an example, I give you telnet. Once a noble and (considered) indispensable part of the "internet", it is now deceased. RIP.
http://www.urbandictionary.com/define.php?term=get+off+my+lawn
American Cultural Reference -- not sure of origin. Take the first meaning above; I am just a cranky old man...
tongue in cheek.
Yes, but...
Each choice point requires a core. You WILL run out of cores; it really doesn't matter how many you have... Unless, of course, you solve THAT problem, which scores you a Nobel Prize.
The Science of Algorithms. Expression, and analysis.
As far as I know, it doesn't have much to do with coding. And it sure doesn't have much to do with computer hardware. In fact, about the only "real" computer sciences job out there is teaching. Or, coming up with new algorithms.
Sometimes, you get to make a breakthrough; a single such breakthrough can make your career -- p-v semaphores and quicksort would be exemplars.
Which makes a "Computer Science" degree about as practically useful as a fine arts degree. Which, in turn, is why schools are tending to concentrate on teaching practical programming skills.
Now, you have indicated that you don't want to go into the programming end; which leaves the purer expression of Computer Science. Go for an advanced degree.
An interesting point -- but who created the internet and home computers for you?
Yep -- we are all now in our 50's and up.
But we didn't grow up on TV either -- the first TV in our family was used to watch the moon landing in '69. But there was no "cable"; we could only receive three stations. Wasn't worth watching, most of the time (except for exceptional events, like the moon landing).
The previous generation (take my mother-in-law - she's in her '70s) didn't see a TV until their late twenties/early thirties -- it certainly isn't a formative part.
Still, census disagrees with me a bit -- TV penetration in households in the USA was nearly complete by 1960 (I guess our family was a hold-out):
http://www.tvb.org/rcentral/mediatrendstrack/tvbasics/02_TVHouseholds.asp
It may be that viewers born 1960 (and before) to 1970 (ei. those who did NOT start with cable) view TV programs as an "event" rather than as disposable entertainment, which may drive that demographic to watch first airings.
(Ob: Now get off my lawn, you damn kids!)
IBM did produce a split version of the Model M for a while -- had a big rotater knob to control the angle of the wedges. If you want to try it, it was known as the Model M15, and originally brought out in 1994, under the "Options" line:
http://www.clickykeyboards.com/index.cfm/fa/categories.main/parentcat/12675
I use a "plain" Model M, simply because I learned to type on typewriters, and I find the split angling to be almost too awkward to use.
But do yourself a favor and try one out. If you type for a living, I think you'll thank me. (however, getting an M15 may be difficult).
So -- the problem is caused by non-symmetrical connections. This causes people who wish to exchange information to use "p2p". The efficient methods for controlling a swarm tend to be decentralized, by permitting the clients to ask for random sections, and discuss among themselves.
And -- we know the network grows by pushing routing decisions to the edge. Or, at least the "networking engineers" should know this. If the issue is that the ISP core routers are being swamped (rather than a bandwidth issue, as you imply), the solution seems evident. Push the routing back to the edge.
In other words, we have to solve the "last mile" problem to cure this. Obviously, allowing a monopoly for wired connections won't work, unless the organization granted that monopoly can guarantee service levels.
Another (hold on tight) solution is simply to equalize upload and download rates. This would make the swarm less effective (after all, much larger pieces could be downloaded from a single site).
But...
WAIT A MINUTE -- bittorrent doesn't ITSELF break pieces into MTU chunks! bittorrent loads at 128k+ (4mb isn't uncommon) chunk sizes.
This transacts 128 to over 4000 packets over the same path! Indeed, it's probably more efficient in terms of router use than automated web browsing.
Which means that the ISP simply cannot bear the load. And it doesn't matter whether that is caused by raw bandwidth or switching. Which brings us back to requiring the SLA.
SCREEN for persistent terminal connections,
XMOVE for X programs (although xmove is a bit of a bear to use, and is suffering a slight case of bitrot) ftp://ftp.cs.columbia.edu/pub/xmove/
VNC is probably a better choice for "generic" use.
VNC can be tunneled through SSH, which is better than opening a port for it (by itself, VNC isn't secure).
Hope this helps.
Wha, wha, wha, WHAT?
1 - "Geeks" do work around "latency" issues -- I have been doing so since the dialup era. I run my own DNS server for this reason, and use nscd on local machines. I use SquidProxy.
2 - I don't work around traffic shaping. But a packet is just a packet. What the fuck does the ISP think it is doing when (re)prioritizing my packets? I guess I simply don't deserve ANY priority, right?
3 - You didn't address your argument for keeping shaping private. Explain WHY it should be private. I will gladly add quality of service requests to my packets -- IF THE SHAPING POLICY IS NOT PRIVATE.
4 - "A few geeks demanding maximum bandwidth". Sure, why not? What does Bell DSL advertise?
Let me give you a hint:
http://www.bellvideostore.ca/help/
This is Bells "download video" service.
"Technical Specifications:
720x480 maximum resolution, 30 fps, 1,500-2,200 kbps average bit rate. These videos will play on all 4th and higher generation Archos portable media players. THESE VIDEOS WILL NOT PLAY ON MOST PORTABLE VIDEO PLAYERS AND ARE NOT IPOD COMPATIBLE."
http://www.bell.ca/support/PrsCSrvGnl_mover_offers.page?region=ON&language=en&EXT=PDL2_SA_MOVE_2008Q2_AR_EN_ON_Google_bell%20dsl&s_kwcid=bell%20dsl|1920763120
"If you're moving, now's the perfect time to move up and into a Bell Better HomeTM: ...
* The most powerful Internet"
Of course Bell (the ISP) offers a music store in addition to the video store:
http://musicstore.sympatico.msn.ca/help_en/index.html
Conflict of interest? No, they really need to shape traffic to keep up with that audio and video goodness.
Let's see how Bell advertises the service for new buyers (note the lack of fine print):
http://www.bell.ca/shopping/PrsShpInt_NewAccess.page?userType=NEW
"Total Internet Performance
Download high quality music files, stream video,or play games.
Best price
in a bundle$42.95/mo.
Bonus: Get 2 months free
Regular price: $47.95/mo. *"
(editors comment - the * footnote is NOT on this page, I have no idea what it means)
And my personal favorite:
http://youtube.com/watch?v=ArpmbnxIQIQ
This ad aired until fairly recently.
So this ISP does bandwidth shaping -- why? Aren't they selling Total Internet Performance, the Bell Better Home (tm), Audio downloads and Video (even streaming) at near HD quality?
A FEW GEEKS DEMANDING MAXIMUM BANDWIDTH??? Hell, I would like to see the beavers deliver a fraction of the promises. And let's not get into "never shared, never slow" crap they advertised.
It's worse. I don't CARE about "not paying". When I buy such a box (I have a DSM 320), I want the freedom to REPLACE the firmware. I want the manufacturer to make it EASY to replace the firmware; after all the firmware is based on GPL licensed material. This has nothing to do with "free as in beer".
I am perfectly willing to spend money buying the device. But I choose devices that use GPL software to allow me the freedom to modify those devices to suit my needs (for example, I want to add flash video to the DSM 320 -- it already has an internet connection,). This shouldn't bother the vendor (it did with Tivo, because Tivo sold a subscription, D-Link didn't), because the box has ALREADY BEEN PURCHASED.
So, make it easy, make it useful, and add a clause stating that the guarantee only applies if the original firmware has been restored.
I understand that the GPL may be incompatible with the "give away the razor, make money on the blades" model; but there are plenty of non-GPL alternatives (that means YOU, Tivo).
But if I buy the whole thing, I want to use as I see fit. And, for a general purpose device, that may go well beyond, or sideways to the intent of the manufacturer (eg. not a GPL example, but XBOX as media center).
That's freedom.
I am a "coder" (well, an architect, but I occasionally do still code).
I do NOT get paid royalties or residuals. Any money I make "up front" on an hourly rate. I am willing to work for LESS, if the final product is GPL licensed. But, the works are "works for hire". I write hundreds of pages a year, and I am happy if my NAME remains on the material as author.
One of my co-workers is also an O'Reilly author (Using Samba). The authors insisted that the book material is available for free over the internet -- O'Reilly still makes money on the printed book (note that enough was made to go to at least a second edition).
This world is not comparable to the "Music industry". At least I make a living, even though I don't have royalties. (Perhaps it is: music acts go on tour.)
I don't work for free. If I couldn't make money from it, I would pick another job, and probably code and teach as a hobby.
Cut 'em off.
The ISPs will lose revenue. The Government will step in to save them from bankruptcy. The "network effect" will be lessened. The "pirates" will find another method. Big advances in encrypted onion ring routing, and multi-site partial anonymous storage. Drop in "owned" systems, as less experienced users are simply cut off
Losers: ISPs and less experienced users. Government and other "snoops"
Winners: Other users due to decrease in SPAM. Peer to Peer forced evolution. Forced evolution in privacy preserving networks.
Neutral: "pirates"
To quote Nike: "Just Do It!". Stop threatening, and let's get this ball into play.
I would LOVE to be the CEO of a billion-dollar corporation. Is it ever going to happen? No, it won't. There, I'm disillusioned, and will become depressed.
Give me my happy pills! And get the hell off of my lawn!
No it doesn't...
It is hard to compare directly, but let's try.
Vista Home Premium (it has media center functionality): In China (according to Microsoft), 899 yuan ( http://blogs.techrepublic.com.com/tech-news/?p=953 )
In USA: 239.95 usd ( http://www.microsoft.com/windows/products/windowsvista/editions/homepremium/default.mspx ).
Asking google for the conversion:
899 Chinese yuan = 130.628296 U.S. dollars
So, a product produced in the US costs 1/2 in a foreign country. Am I allowed to now purchase Chinese product and sell in the US? To find out if this is worth it, let's get a retail (not msrp) price on Windows Vista Home Premium. Amazon lists some prices (no, we don't want the "upgrade" edition):
http://www.amazon.com/gp/product/B0013O54OE/ref=dp_cp_ob_title_1
This prices at 209.99 usd (although it claims msrp is 269.99).
All told, the Chinese version is considerably less expensive.
Actually, the "success" of the iPhone is attributed to it being an Apple product. It certainly does not replace the Blackberry.
And, as to replacing Nokia? Good on Apple if they can. But... where is the $40 iPhone?
Which leads me to a question: just how are you defining success?
Ray
My understanding is that IF a case has been withdrawn that was based on the same matter as the case in question; the case then brought cannot be withdrawn without adjudication. So the judge can award costs?
Of course I don't live in the USA, so I may be totally off-base here.
Since the RIAA has ALREADY backed down from the related John Doe filing, they are not allowed to simply throw in the towel on this case. Procedure should force adjudication on the merits. Of course the Judge can decide to simply let them go home, or can judge the case anyway. So, the RIAA could stand to lose, EVEN IF THEY DON'T PURSUE. To quote Boston Legal: "I am the decider!". And, suitable compensation for the defense can be awarded (at the Judges discretion).
Bad form to back away from a case TWICE. Once, for the John Doe, and now, again.
Um... it's flash.
So, don't run ext3. Use ext2 instead. The use of ext3 is expected to cut the usable life of the media in half.