This joystick-shaped mouse is the most comfortable pointing device I've ever used—better than the Marble FX Trackball or the Wacom pen tablet. It was formerly sold as the Anir Mouse until 3M bought the rights. I hope they sell millions of them. They are about $45.
RF is not that bad. I myself am much more concerned about the relatively high energy visible-spectrum radiation in my office. Has it been conclusively shown that I will not get rabies from this constant dose of radiation? I'm concerned.
once you get a CD burner, you now need a faster harddrive so you don't burn as many coasters
I have relatively shitty CPUs and disks in my system, feeding my shiny new 12x burner. But it is a Plextor burner with BURN-PROOF. When it gets a buffer underrun, it transparently pauses until the data is streaming again. End result, it burns no coasters at all. If my system is too overloaded it might take 10 minutes to burn, instead of 7.
Dammit, it's plain you know that isn't actually the case, but let's not go around posting statements that aren't true. It is bad for the discourse. Mmmkay?
What processing, exactly, would be quicker to do on a server then on my own machine?
Here are some things that are done better (and in some cases quicker) on a large, professionally managed server:
Storage of any important data. Backup is just too much of a pain for most people. It should be taken off the table.
Long-running processes or periodic processes. Desktop machines, for various reasons like noise/power/instability, tend to get turned off or restarted. Multi-month or -year uptime is not a reality for most desktops, even if the O/S is reliable.
Processes that use a large amount of system resources. Better to rent a few gigs of RAM or a few hundred gigs of disk when you need them, rather than plowing all that money into hardware that you will only occasionally use.
Services that should be accessible to other users. A desktop is not normally going to have the reliable high speed connectivity and scalable load capacity of a larger, more managed system.
I would agree that the functions of MS Office, as we are accustomed to using them, don't seem like a good fit for this model, but there are some good possibilities in there.
What you say!! Not half. As I recall there was about 38K free in the bootup message, after graphics and BASIC and sprites. The pixel-addressable screen was something like 320x200 2-bit color (wild ass guess), which would be 16K for the graphics bitmap. Add in 8 sprites at 8 bytes each, I believe. God I miss sprites.
Maybe the Atari Jaguar was the first 64-bit OS. It said 64-bit right on the box.
Thanks for the link. Raskin headed up the original Macintosh project and it's always interesting to see what he's thinking. He is critical of OS X, but he would direct just the same criticism to Windows or MacOS 9 or Gnome or CDE.
hee hee... in mindstrm's defense, Bill Gates made just the same malapropism in The Road Ahead, substituting "factor large primes" for "find prime factors of large numbers".
Yes! The larger issue is that "official strategy guides" are nothing more than a way to squeeze extra money out of suckers. Just like expansion packs, gold editions, the belated "speech pack" phenomenon, and hell, even most sequels.
George Broussard's position is disgusting to me, but not so much because he is stepping on the free speech of third party "strategy guide" publishers. It's offensive that he wants to milk every last dollar out of his users. Obviously the game and documentation will be designed in such a way that you need the official guide to get the most out of it. That's why he wants 100% of the guide market.
Well, you're quite correct to say that "There is no real reason your computer should stop playing music, printing, downloading or whatever because the OS is busy with something else." But your proposed remedy won't float.
The supremacy of low cost solutions in the volume PC marketplace is evident. It dates back to about the 486, when PCs became fast enough to handle a WIMP GUI plus every application that mainstream users need. People stopped buying something fast "so it won't become obsolete"; hell, my old 486 really is not obsolete yet, since it deals fine with Win9x, IE, Office 98. As time marched on, and the silicon improved twentyfold, the need for any intelligent devices in new systems "evaporated in the warm light of fast CPUs and cheap memory." (Nemeth et al., 3ed., p.489)
There is just no way to compete any more with a high priced solution. So forget about intelligent peripherals; they are dead forever at the low end, which these days is most of the picture.
As another poster remarked, a solution is better O/S and driver software engineering. It is by no means an open problem to deliver an O/S that can handle the kind of soft real time requirements you are laying out.
(BTW, you were thinking of integrated video subsystems, not AGP video cards.)
ACID properties are what define a transaction. They do not define a database. There are plenty of databases that do not have transaction support. You are taking an abnormally narrow view of the definition of a database if you require transaction support.
Reference: Intro To Database Systems, 8 ed., C.J.Date, Chapter 14.2 "Transactions" and 14.3 "Transaction Recovery".
On the performance issue: my only point is that the performance I see from the Usenet can be excellent, without any extra contribution from the original poster. I don't understand the distinction you're drawing between "user interaction and speed" in characterizing download performance. Client-server performance can also be excellent if the resources are there, paid for by the service provider. If the resources are not there, what you get is an FTP site providing 2KB/sec/user!
...even if you base Usenet P2P-ness soley on the fact that content comes from the user...
Sorry, I wasn't clear here. When I said "every user contributes network resources", I was thinking of the network which propagates messages, not the content on the network. Every user's ISP contributes a Usenet server to the network, or equivalently subcontracts with a news provider. Either way, each user pays his own way by kicking in disk and network capacity through his ISP relationship.
Of course you are correct that most people don't contribute new content. But this does not constitute freeloading; on the Usenet, the only free rider would be someone illegitimately accessing a private server. A freeloader consumes a scarce resource on a P2P network without kicking in. That's not possible on Usenet, since you must be authorized to connect to well-performing Usenet servers. Mojonation goes to great pains to make freeloading impossible via a cryptographically secured economy. Napster is certainly very freeloader-vulnerable.
I'm hearing you say that Usenet uploaders are freeloaders. I don't think that is the commonly understood meaning of the term. Yes, the uploader gets the right to put a copy of his file on all the Usenet servers in the world. But in general this is considered a contribution, not a theft of others' network resources.
You're quite correct to say that most file transfers wouldn't make sense on the Usenet if they were legal. There are only a few exceptions I can think of, where the cost of distribution is high. I've used i-drive; it's annoying to use and I think it might be one of those VC-based business models like iwon.com that may not be around much longer.
[p2p content] probably can't be indexed effectively, forcing users to search everything. It probably couldn't be effectively rated, promoted, improved, etc.
I too am disappointed with the current primitive tools available for these functions using our existing p2p tools. mojonation has some interesting efforts to address the indexing and rating problems in a distributed way.
My point is not that the corporations must be made rich, rather that revenue is _ultimately_ necessary for the creation and distribution of _most_ desired resources. Hence, even where systems like Napster appear to work, they only work in a context where you ignore the creation of value. Yes, I know that issues like open source software are debatable, but even there I'd assert that open source leaves, and will always leave, a lot to be desired.
Looking around us at the world today, I'm sure you're correct to say that it has taken revenue and profit to create most of the interesting stuff out there. Free software is an exception, but it's a small piece of the creative and engineering pie. And if you have revenue, client-server is the better way to go in distributing your wares. So I do agree with your basic point, if I've understood it correctly.
Maybe in some utopian future, productivity and resources will become so vast that most creative output will be released for free, sort of the gift economy scenario. But this isn't the case today.
Usenet is really not Peer to Peer... it's less Peer to Peer than... IRC...
I consider the Usenet to be p2p because every user contributes network resources. With DCC, or anon FTP, or a BBS, that's not the case. The result is superb performance compared to any DCC/anonFTP/BBS. (BTW, Napster works pretty darn well too, compared to a lot of FTPs out there.)
I got an RCN cable modem nine months ago and the performance from my Usenet server dwarfs my Internet performance, which is nothing to sneeze at. I get 0.5 - 1 megabits/sec of throughput from slashdot on a big comment page, and that's about as good as it gets for me Internet-wise. Most servers, of course, deliver far less throughput. But my usenet server gives me 2.5 - 7 megabits/sec of throughput.
Why would anyone want to put up with... either Napster or Usenet when they can... [download] a verifiable copy of software (or music/videos) from a legitimate source?
You have a good point here; P2P is a security problem. But digital signatures by trusted folks would be a good solution to this problem. Here's hoping they catch on one of these years. In the meanwhile, this is only a problem for software; audiovisual material is fine.
...if Usenet and the like are so efficient, why aren't there any substantial and legitimate downloads on them?
Actually there are such downloads. They are large media objects that are being given away for free by individuals. I'm thinking of amateur pornography in particular, but it's just the most extreme case because of its size. If I have five hundred megabytes of video that I want to distribute for free, I can't use an HTTP or FTP server to do it. The cost will be much too high. P2P, and Usenet in particular, is the only reasonable solution at this time.
The free software community is so tech savvy that we keep ourselves nicely set up with ftp and web sites, even with no revenue to spend. Other free data communities, like fan artwork or cult video communities, find p2p more attractive.
I agree that client/server is best for a revenue generating corporation. It delivers ease of use, ease of administration, and control. But in the realm of free data, it's necessary for end users to kick in. Distributed p2p (including Usenet) is a viable solution that is working today. "Monied and legitimate downloads" is an inadequate measure; the non-monied stuff is of interest to me also.
It's hard to respond directly to your hyperbolic straw man "I'm a slashdot user! It's my god-given right! Me me me!" remarks. I guess I can at least say that I block ads without adopting the particular ridiculous posture you accuse us of.
Then you say, "...I don't... see what's wrong with putting some advertisements... on commercial-provided free content." Hey, I don't see what's wrong with it, either. Again it's hard to respond directly to your almost content-free claim.
But then you deliver a conclusion, which has no discernable relationship to the rest of your post: "...reading sites with advertising and purposefully blocking out that advertising is extremely immoral."
I don't really know what your rationale is here. You describe this practice as "stealing bandwidth". Buddy, these web sites are a PUBLIC FREE SERVICE. There is no obligation whatsoever incurred by visiting. Banner blocking is no more stealing than using a free newspaper as birdcage lining is stealing. The site operator is putting the site out there for everyone, with zero claim on the user's time or resources.
Other users have pointed out that any avoidance of ads in any ad supported medium would be stealing by your definition. Fast forwarding through commercials, going to the movie theater late to avoid commercials at the start of the show, tearing ads out of a magazine like Newsweek: it's all stealing according to your argument. In every case the user is denying the advertiser of his impressions, which I take it you believe are the rightful property of the advertiser (!). I doubt that even you fully believe the logical consequences of your claim.
You wrote: if we were to do away with copyright tomorrow, music "sharing" (piracy) services would surely revert back to client-server architecture, as it's a more efficient, more reliable, easier to use, and easier to maintain design.
That sounds pretty reasonable, until you consider the gigantic and growing p2p piracy system known as the usenet. The advantage of the usenet over the web is the same as napster over client-server systems: by distributing the load among all the participating sites, the aggregate storage and bandwidth available is tremendously larger than any individual pirate could ever contribute. Sure, a big company could buy a lot of bandwidth, but pirates don't have access to those resources.
The Usenet is very hairy, but if copyright were abolished, I don't think it would go away. Same for p2p systems like Napster.
What was the last turn-based game you played? What about the one before that?
Well that's easy: Alpha Centauri yesterday, Panzer General a couple of days ago. Last realtime strategy game? Now that's harder to remember. Realtime strategy just sucks and sucks compared to turn based strategy.
BTW, playing against the computer gets tiresome after not too much time in any game; I play Panzer head to head, generally hotseat. Now that's good gaming.
If the kernel is the engine, GCC is the auto factory. Gnu utilities are the smaller parts of the car; it just so happens that the Gnu versions of those parts do a lot more than the BSD (or BTL) versions, while retaining excellent robustness & correctness.
Kids love features, so good luck with your ressentiment-fueled vendetta against Gnu utilities. People just like 'em.
Well of course, any idiot could tell you that the open source letter should be placed under CVS source control with anonymous read access. Duh. It's unclear how these other letter writing campaigns ever get off the ground without sound change management procedures and full visibility.
It is a given that you should always maintain the ability to roll the system back to any earlier revision. Doesn't matter if it is split up over a network, etc. If there is no version control in place, that would be a good project before one embarks on any kind of rewriting. A project that is not under version control is playing russian roulette. And it is not just a matter of implementing the version control technology; you need to get everyone who edits the code to use the system properly and consistently. It's worth it!
Slashdot Mirror
This joystick-shaped mouse is the most comfortable pointing device I've ever used—better than the Marble FX Trackball or the Wacom pen tablet. It was formerly sold as the Anir Mouse until 3M bought the rights. I hope they sell millions of them. They are about $45.
This book has useful things to say about workspace design, posture, stretching and breaks.
RF is not that bad. I myself am much more concerned about the relatively high energy visible-spectrum radiation in my office. Has it been conclusively shown that I will not get rabies from this constant dose of radiation? I'm concerned.
I have relatively shitty CPUs and disks in my system, feeding my shiny new 12x burner. But it is a Plextor burner with BURN-PROOF. When it gets a buffer underrun, it transparently pauses until the data is streaming again. End result, it burns no coasters at all. If my system is too overloaded it might take 10 minutes to burn, instead of 7.
Dammit, it's plain you know that isn't actually the case, but let's not go around posting statements that aren't true. It is bad for the discourse. Mmmkay?
Here are some things that are done better (and in some cases quicker) on a large, professionally managed server:
- Storage of any important data. Backup is just too much of a pain for most people. It should be taken off the table.
- Long-running processes or periodic processes. Desktop machines, for various reasons like noise/power/instability, tend to get turned off or restarted. Multi-month or -year uptime is not a reality for most desktops, even if the O/S is reliable.
- Processes that use a large amount of system resources. Better to rent a few gigs of RAM or a few hundred gigs of disk when you need them, rather than plowing all that money into hardware that you will only occasionally use.
- Services that should be accessible to other users. A desktop is not normally going to have the reliable high speed connectivity and scalable load capacity of a larger, more managed system.
I would agree that the functions of MS Office, as we are accustomed to using them, don't seem like a good fit for this model, but there are some good possibilities in there.What you say!! Not half. As I recall there was about 38K free in the bootup message, after graphics and BASIC and sprites. The pixel-addressable screen was something like 320x200 2-bit color (wild ass guess), which would be 16K for the graphics bitmap. Add in 8 sprites at 8 bytes each, I believe. God I miss sprites.
Maybe the Atari Jaguar was the first 64-bit OS. It said 64-bit right on the box.
Thanks for the link. Raskin headed up the original Macintosh project and it's always interesting to see what he's thinking. He is critical of OS X, but he would direct just the same criticism to Windows or MacOS 9 or Gnome or CDE.
hee hee... in mindstrm's defense, Bill Gates made just the same malapropism in The Road Ahead, substituting "factor large primes" for "find prime factors of large numbers".
Yes! The larger issue is that "official strategy guides" are nothing more than a way to squeeze extra money out of suckers. Just like expansion packs, gold editions, the belated "speech pack" phenomenon, and hell, even most sequels.
George Broussard's position is disgusting to me, but not so much because he is stepping on the free speech of third party "strategy guide" publishers. It's offensive that he wants to milk every last dollar out of his users. Obviously the game and documentation will be designed in such a way that you need the official guide to get the most out of it. That's why he wants 100% of the guide market.
I'm sorry, is that a rhetorical question?
ALGOL 60
CPL
BCPL
B
C
C++
try to keep it straight
Well, you're quite correct to say that "There is no real reason your computer should stop playing music, printing, downloading or whatever because the OS is busy with something else." But your proposed remedy won't float.
The supremacy of low cost solutions in the volume PC marketplace is evident. It dates back to about the 486, when PCs became fast enough to handle a WIMP GUI plus every application that mainstream users need. People stopped buying something fast "so it won't become obsolete"; hell, my old 486 really is not obsolete yet, since it deals fine with Win9x, IE, Office 98. As time marched on, and the silicon improved twentyfold, the need for any intelligent devices in new systems "evaporated in the warm light of fast CPUs and cheap memory." (Nemeth et al., 3ed., p.489)
There is just no way to compete any more with a high priced solution. So forget about intelligent peripherals; they are dead forever at the low end, which these days is most of the picture.
As another poster remarked, a solution is better O/S and driver software engineering. It is by no means an open problem to deliver an O/S that can handle the kind of soft real time requirements you are laying out.
(BTW, you were thinking of integrated video subsystems, not AGP video cards.)
Plextor drives just don't make mistakes in ripping CDDA. It's the same every time, bit-for-bit, and at high speed.
Reference: Intro To Database Systems, 8 ed., C.J.Date, Chapter 14.2 "Transactions" and 14.3 "Transaction Recovery".
Thanks for the pointer; it looks like the paper you referenced is the one at http://www.cs.virginia.edu/~evans/cs551/saltzer/ "The Protection of Information in Computer Systems". Interesting paper.
Of course you are correct that most people don't contribute new content. But this does not constitute freeloading; on the Usenet, the only free rider would be someone illegitimately accessing a private server. A freeloader consumes a scarce resource on a P2P network without kicking in. That's not possible on Usenet, since you must be authorized to connect to well-performing Usenet servers. Mojonation goes to great pains to make freeloading impossible via a cryptographically secured economy. Napster is certainly very freeloader-vulnerable.
I'm hearing you say that Usenet uploaders are freeloaders. I don't think that is the commonly understood meaning of the term. Yes, the uploader gets the right to put a copy of his file on all the Usenet servers in the world. But in general this is considered a contribution, not a theft of others' network resources.
You're quite correct to say that most file transfers wouldn't make sense on the Usenet if they were legal. There are only a few exceptions I can think of, where the cost of distribution is high. I've used i-drive; it's annoying to use and I think it might be one of those VC-based business models like iwon.com that may not be around much longer.
I too am disappointed with the current primitive tools available for these functions using our existing p2p tools. mojonation has some interesting efforts to address the indexing and rating problems in a distributed way. Looking around us at the world today, I'm sure you're correct to say that it has taken revenue and profit to create most of the interesting stuff out there. Free software is an exception, but it's a small piece of the creative and engineering pie. And if you have revenue, client-server is the better way to go in distributing your wares. So I do agree with your basic point, if I've understood it correctly.Maybe in some utopian future, productivity and resources will become so vast that most creative output will be released for free, sort of the gift economy scenario. But this isn't the case today.
I got an RCN cable modem nine months ago and the performance from my Usenet server dwarfs my Internet performance, which is nothing to sneeze at. I get 0.5 - 1 megabits/sec of throughput from slashdot on a big comment page, and that's about as good as it gets for me Internet-wise. Most servers, of course, deliver far less throughput. But my usenet server gives me 2.5 - 7 megabits/sec of throughput.
You have a good point here; P2P is a security problem. But digital signatures by trusted folks would be a good solution to this problem. Here's hoping they catch on one of these years. In the meanwhile, this is only a problem for software; audiovisual material is fine. Actually there are such downloads. They are large media objects that are being given away for free by individuals. I'm thinking of amateur pornography in particular, but it's just the most extreme case because of its size. If I have five hundred megabytes of video that I want to distribute for free, I can't use an HTTP or FTP server to do it. The cost will be much too high. P2P, and Usenet in particular, is the only reasonable solution at this time.The free software community is so tech savvy that we keep ourselves nicely set up with ftp and web sites, even with no revenue to spend. Other free data communities, like fan artwork or cult video communities, find p2p more attractive.
I agree that client/server is best for a revenue generating corporation. It delivers ease of use, ease of administration, and control. But in the realm of free data, it's necessary for end users to kick in. Distributed p2p (including Usenet) is a viable solution that is working today. "Monied and legitimate downloads" is an inadequate measure; the non-monied stuff is of interest to me also.
Insightful, my ass.
It's hard to respond directly to your hyperbolic straw man "I'm a slashdot user! It's my god-given right! Me me me!" remarks. I guess I can at least say that I block ads without adopting the particular ridiculous posture you accuse us of.
Then you say, "...I don't... see what's wrong with putting some advertisements... on commercial-provided free content." Hey, I don't see what's wrong with it, either. Again it's hard to respond directly to your almost content-free claim.
But then you deliver a conclusion, which has no discernable relationship to the rest of your post: "...reading sites with advertising and purposefully blocking out that advertising is extremely immoral."
I don't really know what your rationale is here. You describe this practice as "stealing bandwidth". Buddy, these web sites are a PUBLIC FREE SERVICE. There is no obligation whatsoever incurred by visiting. Banner blocking is no more stealing than using a free newspaper as birdcage lining is stealing. The site operator is putting the site out there for everyone, with zero claim on the user's time or resources.
Other users have pointed out that any avoidance of ads in any ad supported medium would be stealing by your definition. Fast forwarding through commercials, going to the movie theater late to avoid commercials at the start of the show, tearing ads out of a magazine like Newsweek: it's all stealing according to your argument. In every case the user is denying the advertiser of his impressions, which I take it you believe are the rightful property of the advertiser (!). I doubt that even you fully believe the logical consequences of your claim.
That sounds pretty reasonable, until you consider the gigantic and growing p2p piracy system known as the usenet. The advantage of the usenet over the web is the same as napster over client-server systems: by distributing the load among all the participating sites, the aggregate storage and bandwidth available is tremendously larger than any individual pirate could ever contribute. Sure, a big company could buy a lot of bandwidth, but pirates don't have access to those resources.
The Usenet is very hairy, but if copyright were abolished, I don't think it would go away. Same for p2p systems like Napster.
BTW, playing against the computer gets tiresome after not too much time in any game; I play Panzer head to head, generally hotseat. Now that's good gaming.
If the kernel is the engine, GCC is the auto factory. Gnu utilities are the smaller parts of the car; it just so happens that the Gnu versions of those parts do a lot more than the BSD (or BTL) versions, while retaining excellent robustness & correctness.
Kids love features, so good luck with your ressentiment-fueled vendetta against Gnu utilities. People just like 'em.
Well of course, any idiot could tell you that the open source letter should be placed under CVS source control with anonymous read access. Duh. It's unclear how these other letter writing campaigns ever get off the ground without sound change management procedures and full visibility.
mod this up -- it's exactly right
It is a given that you should always maintain the ability to roll the system back to any earlier revision. Doesn't matter if it is split up over a network, etc. If there is no version control in place, that would be a good project before one embarks on any kind of rewriting. A project that is not under version control is playing russian roulette. And it is not just a matter of implementing the version control technology; you need to get everyone who edits the code to use the system properly and consistently. It's worth it!
s/better wordsmith than I/better wordsmith than me