But if they're G4s, at a gigaflop each (for moderately SIMDable computations)...
MAN what a box that could be! (Then think about a cluster of them...)
Can you imagine the consternation among the export controllers, when a high-end home-computer can finally be used to predict the weather several days out or model nuclear devices?
The Millenium is upon us, and the Singularity is not far beyond.
It DOES allow the holder to install additional security services into MS's security framework. But the holder has to have access to your machine to do that.
So that means they also need a trapdoor that lets them run on the machine, but once they have it they can use this trapdoor to hack your security modules.
This trapdoor is where it was easily findable because that's where the code prety much had to be in order to augment the security install checking. But how do you know the other one isn't in there somewhere?
If you're considering BSD variants you should remember that OpenBSD is the open-source BSD variant whose focus is security. (And it's maintained in and distributed from Canada, so no export crimps on encryption.)
I hereby propose the term "Pink Fedora" to refer to a Red-Hat derived item. For instance, a machine loaded with a stock Red Had 6.0 but without the boxed set would be said to be loaded with Pink Fedora 6.0, and so on.
It's close enough you know what you're talking about and far enough away to convey that there may be divergence, the buyer should beware, and that it's the sellers, not Red Hat's, fault if there's something wrong.
... if it's uncachable it gets more hits on the adds. B-)
Seriously: making it more cachable would have cut the load DRASTICALLY - and still will for the new config. When a user tries to follow a thread, he hits the server again when he pops back up to the level above, doubling the number of hits on the servers.
It also is a pain in the tail when the servers are slow - which even the new config will eventually become.
Mormons make good neighbors.
on
911 Calls Linux
·
· Score: 1
Everybody's religion looks screwey from the outside. That of the Mormons has the advantage that it convinces essentially all of them to really BE good to their neighbors - especially non-Mormons - and to keep their religious squabbles within the various factions of their church.
The system Ted Nelson invented is a global naming system for very large data sets; not unlike the Dewey decimal system some libraries use for indexing books.
Sadly the naming system can't handle changes to the referenced text, so in order not to break the references, every version of the document that was ever created has to be kept.
Not really. (You're thinking of his early descriptions of the concept.)
In "gold" the issues of location and labeling were separated, as was version tracking. You can inquire about intermediate versions of a document, but if no server you can reach happens to have saved a copy of them them you can't read them. Servers can in principle "burn all the copies" of a published version, archive them and lose the archive, and so on.
For some kinds of content the server can compute it on the fly, so nobody ever has to save a copy (though other servers might cache it). This corresponds to documents with dynamic content, such as query retrievals and other CGI-generated stuff.
The web also has a hierarchical global naming system, and while it can't index down to a specific paragraph in a document, it can index as far as a well-known entry point within a document.
While the web does have problems with global names, those problems would only be worse with a system like Xanadu.
Again you're confusing indexing with identity. "Gold" handles identity as a separate item. The identity of a document is a history track across versions. "The current issue of Wired magazine" would be one such identity, as would "the latest rough of the December 1999 issue of Wired magazine". You "Publish" the December issue by "hopping the bert" of the current issue onto the same version as the "bert" of the rough of the December issue, creating a new entry in the issue history with exactly the same content as the rough.
The history of the "current issue" bert is thus the set of published issues, while the history of the "December 1999 rough" is the history of the assembly of the issue. The current state of the "current issue" bert is the copy on the newsstand, you can see the publishing history by viewing the bert history, and you can read the back issues (if somebody saved the bits and will serve them to you) by viewing through the previous states of the bert.
As for importing the whole web, "gold" lets you create a placeholder for an external document. Initially you can view the external document through the placeholder - and your server will compose a query to the external system if possible, or tell you that you can't view it yet. Later, if the external document can be and is imported into the system, the owner of the placeholder can "unify" it to the on-system version, declaring the on-system verson to be canonical and giving up his ownership rights to the owner of the canonical version.
Ted wanted to have the RIAA model in place, but the actual implementors were working mainly on other issues and hadn't gotten around to putting in royalty hooks.
They hadn't put them into "gold" as of the point where Autodesk bailed out, and I don't think they were in "green" at the point where it was sheleved in favor of work on "gold".
I have a couple class-C (/24, i.e. 256) IP addresses ranges that I registered in the good old days, before NSI's monopoly and the spinoff of ARIN (American Registry of Internet Numbers).
I moved a while back, and just noticed that ARIN didn't have my current address - and had constructed a new NIC handle from my old one when they spun off.
While I was there I took a look at their fee structure - and was floored. $2,500 for a class-C - IF you qualify. And then $30 a year to maintain it, too. This is on top of my domain name, of course. Don't like it? Rent some numbers from your ISP, or THEIR ISP, or the one above them. And pay their markup. Of course if you move, or change ISPs, you'll have to change your numbers. (A private block is now called "portable".) And the ISP will want you to pay for all those numbers (that THEY have to pay for) whether they're routing packets to them or not.
And if I want an ASN (Autonomous System Number) so I can participate in the more advanced routing protocols, that's ANOTHER 500 bucks, and ANOTHER $30 annual fee.
They've started allocating IPv6 addresses, too. You know, the new version of internet protocol with about 10**39 addresses - enough addresses to assign a big block to every atom in the lithospere. (Or is that enough to give several to every subatomic particle in the known universe?) But they want to conserve those, too (maybe encoding routing information into them - and thus casting it in concrete and recreating the pre-dotist mail routing problems). So pony up a few kilobux if you want an allocation. And keep yet another $30 annual fee coming, forever.
Was it McCracken or Belluzzo who kept bringing Clinton and Gore to their site?
For a while there was a bi-weekly round of high-tech schmoozing, photo-ops, announcements of new programs to pave information highways with tax money and to pick the winners. All based at SGI's facility. With two days of more-snarled-than-usual traffic every time.
I am reminded of the time a few years back when the host of a NPR station's classical music show proclaimed the "death of Rock'n'Roll" - because only 5% of the songs on the playlist stations had been composed in the current year. He was oblivious to the fact that current RnR songs now had several decades of good stuff with which they were compeeting for "earballs" on the commercial stations.
I heroically refrained from calling him up and declaring his classical music to be rotting in its grave, since 0% had been composed in the current year. B-)
Regardless of how much of the phenomenon they report is a sampling artifact and how much is the result of a few good sites being so useful that they get a lot of repeat attention, at least one of their inferences is faulty:
... it suggests that the idea of the Web as a guarantor of infinite choice and endless serendipity is hardly a divine right. Does this... actually threaten the vibrant democratizing force the Web symbolizes and often embodies?
What this misses is that the web remains un-gate-kept, and continues to grow. Commercial convenience sites may cause those pages they link to be found more quickly. But they can't stop ideas they dislike from reaching an audience by ignoring them - just as the editor-controlled mass media no longer can.
Even if ALL the current convenience sites started systematically ignoring pages propagating a particular set of ideas or a particular side of an argument, they'd just discredit themselves and start another backlash, like the one against the establishment media.
Meanwhile, ANYONE can start a NEW convenience site, just as anyone can post a page. Anyone can put up a page linking their favorite non-establishment postings and link to it from any page they control, mail the URL to their friends, pass out handbills, or scrawl it on bathroom walls. Any group of people with some shared non-establishment views can create a "literature" by mutually linking, making a bigger target to be found or linked into.
Just as the Sturgeon's-law chaff didn't stop people from finding the good stuff before, so the flood of distractions doesn't keep people from finding the stuff they want now. There's more they don't want, but there's more of what they're looking for, too. And there's no concentration of the printing presses and broadcast outlets in the hands of a power-structure to bar entry.
They used synthetic tire rubber and liquid oxygen. Gave out paperweights made of it minus the oxygen B-) ). They were originally made up for bouncing off the desk of bureaucrats who wanted to deny them a license because of the "danger" of working with "highly explosive rocket fuel" in their particular patch of desert or whatever.
Their first (and last) sounding rocket test firing had the worst possible failure - the oxygen valve stuck open at 10% - too little thrust to lift off, so it slowly burned up on the pad.
Glad to hear something is left of 'em. (Anyone know if Jim Bennett is still there?)
After the SPARC debacle, where Sun first encouraged others to use the SPARC then pulled the rug out from those who did, I'll wait until somebody else does such a processor before getting excited.
Just write it with an actor-based OOP style. Then it's automatically split into tiny, simple, parallizable chunks.
Instead of having to explicitly declare what's parallizable, you explicitly declare what's interdependent. Typically that's a much smaller set - especially after the message-send/receive dependencies (which are automatically handled for you) are excluded.
If there's any OS out there that is more comprehensively thread oriented (which leads to more application threading) it must be proprietary.
Out there currently, perhaps that's true. But looking back in computing history there's the T.H.E. multiprocessing system (by Djikstra and Riddle), plus an arbitrary number of clones of it, typically living in embedded systems.
I used one done by Mark Weiser, on a Nova, about 1975, and cloned my own onto an 8080 a few years later. Mine was a preemptive multitasking kernel (excluding drivers) a little over 500 bytes long. Add a console driver, a debugger, a network stack (not IP), real-time-clock processing, scheduled event interpreter, instrumentation drivers, a relay logic ladder-diagram interpreter, drivers to receive and send relay/contact signals from/to optoisolators, and a network daemon that downloaded schedules, read meters, examined relay states and stuck virtual screwdrivers in to force them, and it still come in under 2K bytes. This left the other 2K of ROM available for a description of a hysterically-large emulated-relay network.
That sucker flew, too. With the one tweak I added it became exactly an implementation of "actors", perhaps a bit before they were formalized. If you're not familiar with them: Imagine a machine where every program is in C++, but where every instance of every class is a separate thread of execution, every complicated class has been split into a set of simpler classes with one thread-related member function each, every call to a thread-related member functin is an intertask message - at about the cost of a subroutine call (with free queueing of multiple messages), and every thread-related member function (with all the non-thread-related subroutines it calls) can in principle run simultaneously (because they explicitly mutex when they must share a resource, and the free queueing makes such occasions are extremely rare). Now pour all these tiny tasks into the machine, with a half-K kernel to orchestrate them.
On a single processor machine the fact that the individual objects could run in parallel was an unused side-effect of a programming style that simplified writing programs to take maximum advantage of the tiny kernel. But with a more modern hardware platform, with a slightly more complicated kernel and perhaps a little hardware assist, the same style automatically produces a great pile of tiny, simple objects that can all be run in parallel on as many CPUs as you've got.
One question: If these things run on methanol (alcohol), will you have to be over 21 to purchase fuel for your laptop?;-P
No. Booze alcohol is ethanol. Methonol is deadly poison - and what they contaminate ethanol with to "denature" it.
Your enzyme systems, which turn ethanol into acetaldehyde and acetic acid (vinegar), which are only mildly annoying, also turn methanol into formaldehyde and formic acid (ant toxin), which are deadly - especially to nerves. Ingest a little methanol and go blind. Even the fumes will give you a headache. I'd be a lot happier if they fueled them with ethanol - or even isopropanol (rubbing alcohol).
On the other hand, they might run about as well on ethanol as on methanol, unmodified. We'll have to see the details of the devices once they're out.
One thing that impressed me about the referenced story was the lack of any information on how the computer actually worked or what kind of computation it did.
Modern mainstream news organizations have come up with a content-free grammar.
I think the possibility of global, untracable (FBI Suit, "Yeah, the call is coming from the eastern hemisphere") communications scared the crap out of our own (US) gov'
Hard to trace, perhaps, but VERY easy to tap. Just point an antenna at every satellite and you've got 'em all. What signal goes up must come down - at least while all the handsets are on the planet's surface. Tapping regular cellular, meanwhle, needs a receiver in each cell of interest unless you can tap the network feeding the base stations.
And you might be able to get some location info from the satellites, too.
The PETA people anthropomorphize incessantly. If the animal's face is shaped like the human's fright expression, they'll claim it's scared and stressed.
Given that penguins in the wild are gregarious (or at least the Arctic types are) and these were raised by humans and calm enough to be rubbing beaks and eating treats, I suspect the only stress they might be experiencing is insufficent crowding.
Compare that to the life of a wild penguin - with predators behind every visual barrier - and you'll see these little birdies were likely in penguin heaven.
Slashdot Mirror
MAN what a box that could be! (Then think about a cluster of them...)
Can you imagine the consternation among the export controllers, when a high-end home-computer can finally be used to predict the weather several days out or model nuclear devices?
The Millenium is upon us, and the Singularity is not far beyond.
security services into MS's security framework. But the holder has to have access to your
machine to do that.
So that means they also need a trapdoor that lets them run on the machine, but once they have it they can use this trapdoor to hack your security modules.
This trapdoor is where it was easily findable because that's where the code prety much had to be in order to augment the security install checking. But how do you know the other one isn't in there somewhere?
If you're considering BSD variants you should remember that OpenBSD is the open-source BSD variant whose focus is security. (And it's maintained in and distributed from Canada, so no export crimps on encryption.)
(This comment intentionally left blank.)
It's close enough you know what you're talking about and far enough away to convey that there may be divergence, the buyer should beware, and that it's the sellers, not Red Hat's, fault if there's something wrong.
Seriously: making it more cachable would have cut the load DRASTICALLY - and still will for the new config. When a user tries to follow a thread, he hits the server again when he pops back up to the level above, doubling the number of hits on the servers.
It also is a pain in the tail when the servers are slow - which even the new config will eventually become.
I used to be "Ungrounded Lightning Rod."
Everybody's religion looks screwey from the outside. That of the Mormons has the advantage that it convinces essentially all of them to really BE good to their neighbors - especially non-Mormons - and to keep their religious squabbles within the various factions of their church.
(And he could give Mike@ABC his free number.)
The system Ted Nelson invented is a global naming system for very large data sets; not unlike the Dewey decimal system some libraries use for indexing books.
Sadly the naming system can't handle changes to the referenced text, so in order not to break the references, every version of the document that was ever created has to be kept.
Not really. (You're thinking of his early descriptions of the concept.)
In "gold" the issues of location and labeling were separated, as was version tracking. You can inquire about intermediate versions of a document, but if no server you can reach happens to have saved a copy of them them you can't read them. Servers can in principle "burn all the copies" of a published version, archive them and lose the archive, and so on.
For some kinds of content the server can compute it on the fly, so nobody ever has to save a copy (though other servers might cache it). This corresponds to documents with dynamic content, such as query retrievals and other CGI-generated stuff.
The web also has a hierarchical global naming system, and while it can't index down to a specific paragraph in a document, it can index as far as a well-known entry point within a document.
While the web does have problems with global names, those problems would only be worse with a system like Xanadu.
Again you're confusing indexing with identity. "Gold" handles identity as a separate item. The identity of a document is a history track across versions. "The current issue of Wired magazine" would be one such identity, as would "the latest rough of the December 1999 issue of Wired magazine". You "Publish" the December issue by "hopping the bert" of the current issue onto the same version as the "bert" of the rough of the December issue, creating a new entry in the issue history with exactly the same content as the rough.
The history of the "current issue" bert is thus the set of published issues, while the history of the "December 1999 rough" is the history of the assembly of the issue. The current state of the "current issue" bert is the copy on the newsstand, you can see the publishing history by viewing the bert history, and you can read the back issues (if somebody saved the bits and will serve them to you) by viewing through the previous states of the bert.
As for importing the whole web, "gold" lets you create a placeholder for an external document. Initially you can view the external document through the placeholder - and your server will compose a query to the external system if possible, or tell you that you can't view it yet. Later, if the external document can be and is imported into the system, the owner of the placeholder can "unify" it to the on-system version, declaring the on-system verson to be canonical and giving up his ownership rights to the owner of the canonical version.
They hadn't put them into "gold" as of the point where Autodesk bailed out, and I don't think they were in "green" at the point where it was sheleved in favor of work on "gold".
I wouldn't mind a banner ad on every screen if they'd stop charging for domain names. B-)
I moved a while back, and just noticed that ARIN didn't have my current address - and had constructed a new NIC handle from my old one when they spun off.
While I was there I took a look at their fee structure - and was floored. $2,500 for a class-C - IF you qualify. And then $30 a year to maintain it, too. This is on top of my domain name, of course. Don't like it? Rent some numbers from your ISP, or THEIR ISP, or the one above them. And pay their markup. Of course if you move, or change ISPs, you'll have to change your numbers. (A private block is now called "portable".) And the ISP will want you to pay for all those numbers (that THEY have to pay for) whether they're routing packets to them or not.
And if I want an ASN (Autonomous System Number) so I can participate in the more advanced routing protocols, that's ANOTHER 500 bucks, and ANOTHER $30 annual fee.
They've started allocating IPv6 addresses, too. You know, the new version of internet protocol with about 10**39 addresses - enough addresses to assign a big block to every atom in the lithospere. (Or is that enough to give several to every subatomic particle in the known universe?) But they want to conserve those, too (maybe encoding routing information into them - and thus casting it in concrete and recreating the pre-dotist mail routing problems). So pony up a few kilobux if you want an allocation. And keep yet another $30 annual fee coming, forever.
For a while there was a bi-weekly round of high-tech schmoozing, photo-ops, announcements of new programs to pave information highways with tax money and to pick the winners. All based at SGI's facility. With two days of more-snarled-than-usual traffic every time.
I heroically refrained from calling him up and declaring his classical music to be rotting in its grave, since 0% had been composed in the current year. B-)
Regardless of how much of the phenomenon they report is a sampling artifact and how much is the result of a few good sites being so useful that they get a lot of repeat attention, at least one of their inferences is faulty:
What this misses is that the web remains un-gate-kept, and continues to grow. Commercial convenience sites may cause those pages they link to be found more quickly. But they can't stop ideas they dislike from reaching an audience by ignoring them - just as the editor-controlled mass media no longer can.
Even if ALL the current convenience sites started systematically ignoring pages propagating a particular set of ideas or a particular side of an argument, they'd just discredit themselves and start another backlash, like the one against the establishment media.
Meanwhile, ANYONE can start a NEW convenience site, just as anyone can post a page. Anyone can put up a page linking their favorite non-establishment postings and link to it from any page they control, mail the URL to their friends, pass out handbills, or scrawl it on bathroom walls. Any group of people with some shared non-establishment views can create a "literature" by mutually linking, making a bigger target to be found or linked into.
Just as the Sturgeon's-law chaff didn't stop people from finding the good stuff before, so the flood of distractions doesn't keep people from finding the stuff they want now. There's more they don't want, but there's more of what they're looking for, too. And there's no concentration of the printing presses and broadcast outlets in the hands of a power-structure to bar entry.
Their first (and last) sounding rocket test firing had the worst possible failure - the oxygen valve stuck open at 10% - too little thrust to lift off, so it slowly burned up on the pad.
Glad to hear something is left of 'em. (Anyone know if Jim Bennett is still there?)
Now if we can just get everybody else to do it... B-)
After the SPARC debacle, where Sun first encouraged others to use the SPARC then pulled the rug out from those who did, I'll wait until somebody else does such a processor before getting excited.
Instead of having to explicitly declare what's parallizable, you explicitly declare what's interdependent. Typically that's a much smaller set - especially after the message-send/receive dependencies (which are automatically handled for you) are excluded.
Out there currently, perhaps that's true. But looking back in computing history there's the T.H.E. multiprocessing system (by Djikstra and Riddle), plus an arbitrary number of clones of it, typically living in embedded systems.
I used one done by Mark Weiser, on a Nova, about 1975, and cloned my own onto an 8080 a few years later. Mine was a preemptive multitasking kernel (excluding drivers) a little over 500 bytes long. Add a console driver, a debugger, a network stack (not IP), real-time-clock processing, scheduled event interpreter, instrumentation drivers, a relay logic ladder-diagram interpreter, drivers to receive and send relay/contact signals from/to optoisolators, and a network daemon that downloaded schedules, read meters, examined relay states and stuck virtual screwdrivers in to force them, and it still come in under 2K bytes. This left the other 2K of ROM available for a description of a hysterically-large emulated-relay network.
That sucker flew, too. With the one tweak I added it became exactly an implementation of "actors", perhaps a bit before they were formalized. If you're not familiar with them: Imagine a machine where every program is in C++, but where every instance of every class is a separate thread of execution, every complicated class has been split into a set of simpler classes with one thread-related member function each, every call to a thread-related member functin is an intertask message - at about the cost of a subroutine call (with free queueing of multiple messages), and every thread-related member function (with all the non-thread-related subroutines it calls) can in principle run simultaneously (because they explicitly mutex when they must share a resource, and the free queueing makes such occasions are extremely rare). Now pour all these tiny tasks into the machine, with a half-K kernel to orchestrate them.
On a single processor machine the fact that the individual objects could run in parallel was an unused side-effect of a programming style that simplified writing programs to take maximum advantage of the tiny kernel. But with a more modern hardware platform, with a slightly more complicated kernel and perhaps a little hardware assist, the same style automatically produces a great pile of tiny, simple objects that can all be run in parallel on as many CPUs as you've got.
No. Booze alcohol is ethanol. Methonol is deadly poison - and what they contaminate ethanol with to "denature" it.
Your enzyme systems, which turn ethanol into acetaldehyde and acetic acid (vinegar), which are only mildly annoying, also turn methanol into formaldehyde and formic acid (ant toxin), which are deadly - especially to nerves. Ingest a little methanol and go blind. Even the fumes will give you a headache. I'd be a lot happier if they fueled them with ethanol - or even isopropanol (rubbing alcohol).
On the other hand, they might run about as well on ethanol as on methanol, unmodified. We'll have to see the details of the devices once they're out.
Barry was quoting Thomas Jefferson.
Modern mainstream news organizations have come up with a content-free grammar.
hemisphere") communications scared the crap out of our own (US) gov'
Hard to trace, perhaps, but VERY easy to tap. Just point an antenna at every satellite and you've got 'em all. What signal goes up must come down - at least while all the handsets are on the planet's surface. Tapping regular cellular, meanwhle, needs a receiver in each cell of interest unless you can tap the network feeding the base stations.
And you might be able to get some location info from the satellites, too.
Given that penguins in the wild are gregarious (or at least the Arctic types are) and these were raised by humans and calm enough to be rubbing beaks and eating treats, I suspect the only stress they might be experiencing is insufficent crowding.
Compare that to the life of a wild penguin - with predators behind every visual barrier - and you'll see these little birdies were likely in penguin heaven.