No doubt Notes is the greatest thing ever. Certainly EDS thinks so, 'cause they talked Corporate into switching everything over by the end of the year. All of which I might laugh at except....
There's no Unix client, which means that my e-mail access is about to go away. This doesn't seem to be part of EDS' roadmap, and I can see that the problems of insignificant engineers might not concern them. I, on the other hand, do at least half of my work via e-mail in one form or another and for some reason don't share their opinion wrt its importance.
Which raises the question: does anyone know of a way to work around a Notes server and get mail, short of external Webmail services?
So far it's only in the discussion stages, but the FEC's rules on independent TV spots is ripe for abuse. The scheme goes like this: Assume that you are trying to defeat the Hon. Cross Palmer. You register as a PAC and place ads showing archive footage of Rep. Palmer at his stupidest, along with double-meaning captioning written by a really good satirist. The upshot is that the Palmer not only becomes a public laughingstock, but YOUR budget gets charged against HIS spending limits.
It's the TV equivalent of a Joe-job. The same tactic works even better on the Net. You register as a PAC and put up a page on Geocities with the Best of Palmer done in slack-stick and then report to the FEC listing the Geocities server farm as a contribution. Now multiply by the thousands of people who think that the public has been serviced enough by Palmer.
MUCH more practical and cost-efficient than television. But then we knew that, didn't we?
Let's face it, folks -- there are two fundamental problems with software patents. One is that doing a search to avoid infringement is a horrible resource hog, and the other is that (a la Ted Sturgeon) 95% are junk.
Both of these problems are solvable, though, in the same sense that "with enough eyes all bugs are shallow." A distributed project to review software patents would go a short way to index and cross-reference the pests, but it could go a long way toward pointing out how silly most of them are. Most of these toads got past the Patent Office because the PO doesn't review against unpatented prior art, and until recently none of the prior art was patented.
So! The trick is to scan through the patent database for software patents. Sort them into categories. Apply an open review process which rates patents by impact (e.g., a patent on linked lists) and if possible identifies prior art.
With enough ludicrous examples we might even be able to enlist some press to shame Congress into dealing with the problem.
GuidoDKP wisely observed: People have been predicting the death of visible-light lithography for years now, but it's so useful that people keep coming up with ways to extend its life -- I think the latest enhancement was the use of phase-shifted interference patterns to draw features with visible light that were smaller than the wavelength of the light used...
Yup, phase-shifting is the latest trick but it's been pushed about as far as it can go. (My employer uses it and little bitty polygons are what I get paid for, at least in part.) Although "visible light" is a stretch when you're talking about ultraviolet.
Spiffy ultrahigh resolution line-drawing is monumentally valuable, though, if only as a means of producing those next-generation mask sets.
So -- wouldn't want to misrepresent your position -- you'd be OK with the local 'phone company only providing modem service to their own ISP subsidiary?
I know you're thrilled at the idea of getting cable, but will you still be thrilled when they tell you that to get & stay connected you have to run Win98SE? Terms of service, contracts, stuff like that.
Their cable, their rules (even if they put it in with a City charter). We're still doing 56K.
rde groused: One thing that grated about the article; speaking about four-colour nanolithography. When you're dealing with molecules, colour loses its meaning, as the wavelength of visible light is up to twenty-five times the width of the line being drawn. Of course, you could draw lots of lines, but that'd take a hell of a long time.
I think they're talking about the method, not the result. Four-colour lithography depends on having chemically distinct dots Real Close Together.
Another reason to think in terms of four 'colours' is the famous topological theorem.
These guys could end up buying MICROS~1 out of petty cash. From the sounds of it they've found the semiconductor fabrication Philosopher's Stone: a way to do sub-X-ray physical patterns. Right now semiconductor feature sizes are in the 120+ nanometer range and severly limited by quantum optical effects. If they can lay down a precise grid at 15 nm and use it to sharpen up the lithography, they are going to own the semiconductor industry.
Gartner Group: IDC figures show Linux comprised just 3.5 per cent of PC server shipments last year.
Translation: Linux accounted for only 3.5% of the $$$ billed for OS orders last year. (IOW, for each 1000 copies of NT at ~$700 each there were 490 copies of Red Hat shipped at $50 each, and who knows how many times each of them was installed.)
Alternate translation: Linux preinstalls accounted for only 3.5% of the $$$ billed last year for shipped servers. (Which is a lot more preinstalls than I would have expected, and why I doubt this translation.)
The thing to remember is that the Gartner Group is a marketing research firm, and they track MONEY, not UNITS. That little detail was one of the main reasons that IBM and Compaq kept their top spots for as long as they did against higher-volume but lower-priced competition.
Kintanon asked: Ok, I might get flamed to hell for this, but... ummm What does powerpoint DO?? I know a lot of the marketing people around here use it, but I've never seen it DO anything? Help me out here?
Powerpoint absorbs huge amounts of time that management, marketers, and other suits might otherwised spend doing real harm.
The brain is one seriously ad-hoc architecture. At a basic level, there is impulse: smell coffee, raise cup. To avoid having this cause problems in awkward circumstances (like when using hands for more interesting things [more interesting than coffee? Hmmmm]) there is a review process (think gatekeeper) that checks impulses before execution.
This gatekeeper, being a relative latecomer in evolutionary terms, isn't particularly robust. Lots of things can take it offline, such as alcohol, which explains why a depressant (alcohol) can act as an apparent stimulant.
Likewise, people with ADHD actually have soporific gatekeepers and thus for them the very thought is the deed ("No, Honey, I really didn't mean to spill coffee all over that white shirt while we were kissing!") Stimulants "wake up" the 'gatekeeper' and allow the subject to stop foot before oral insertion.
Lots more of this over on alt.support.attn-deficit
Actually, this is specsmanship. Where IEEE1394 does its arbitration, framing, etc. at a minimum of 100 Mb/s and usually at 400 Mb/s, USB of any flavor does its polling (notice, it is *not* a peer interconnect) and framing at no more than 12 Mb/s (assuming that there are no low-speed 1.5 Mb/s devices in the path).
Each packet in USB2 starts out with the same preamble (12Mb/s) and ends with the same postamble (12Mb/s); the new twist is that there may be a tone burst in the middle of 480 Mb/s or whatever. Unless you ship really huge packets (which aren't allowed by the protocol anyway) the average transfer rate is a lot lower than 408 Mb/s.
You do not want to be doing escape velocity at sealevel. The thermal equivalent is deep ultraviolet and the area loading is in the crumbling-diamonds range. The 600 MPH limit is probably aerodynamic. Max Q.
An absurdly huge amount of fuel goes into getting up to Mach One; something like 50%. 600 MPH isn't bad at all.
Nine Gs is the limit for fighter pilots sitting more or less upright (actually about 30 degrees from upright.) Lying down on your back in NASA fashion makes for much higher limits.
The biggest payoff would be if they could get up to ramjet speeds using a local nuclear plant. Or the Grid at off-peak times, keeping in mind that when a shuttle launches it briefly matches the power output of the entire US electrical generation industry. Still, dragging a launch vehicle up to ramjet speeds would allow ditching most of the LOX, and that's two-thirds of the fuel load.
Don't uncritically accept the story that Rambus is a superior technology. Dell (a devout Rambus house) did some very interesting benchmarking, and although Intel and Rambus had the results censored from the IDF procedings you can see a copy at InQuest Market Research
Bottom line: Rambus appears to be substantially (like 15-40%) slower than PC100 SDRAM for typical applications. Oops.
laymil wrote: Rambus, the company that developed the technology was not the source of the problem. It was Intel, and their 820 chipset, which is why it was the duty of Intel to fix the problem. So there.
Actually, there doesn't seem to be all that much wrong with the 820, aside from some speed limitations in the RAC. The big gotcha is a PWB-level signal integrity problem with the reference implementation of Rambus which wasn't anticipated by the relatively superficial signal-integrity analysis that the Rambus gang did.
From what I've gathered of the signaling problems with Rambus there are only two main ways they could 'fix' the present problem, and neither is exactly a fix.
Increase the trace separation to reduce the crosstalk and pattern-dependent impedance mismatches. Along with this an adjustment in termination resistance might be some help.
Change the trace lengths to move the resonant point away from the line fundamental frequency. Apparently a resonant mode near the signaling rate was one of the main gotchas.
In both cases new motherboard layouts will be needed, and since both will take up more space the whole floorplan may change. At best, this will take a few months to get the MBs designed, through validation and regulatory approval (not a trivial issue with this kind of bandwidth!) and into the production pipe. Kiss Q4 goodbye and probably Q1; the memory shops won't be seeing any demand until Q2 at soonest, if at all.
On top of that there will always be the charming issue (which Rambus seems to have in other areas as well) that the operating area for the memory subsystem will have a Swiss-cheese character. Instead of a 'schmoo' plot, with a maximum frequency of operation and constraints on voltage and temperature, there will be areas of operation and failure, alternating. Maybe 300 MHz and 800 MHz will be OK, but 700 will be out. In fact, that seems to be the situation right now.
The best guide for UI design I ever learned was from an educational psychologist who explained it this way:
For any of us at any given time, a task is either associative or cognitive. Associative tasks are basically stimulus/response patterns while cognitive ones require some symbol manipulation (eg internal dialog.) At any time you can have an almost unlimited number of associative tasks going but only one cognitive one. Thus when you're driving along a familiar route in good weather and light traffic you can have an interesting discussion going with your passenger(s) but bad weather, an unfamiliar neighborhood, or the kid running out into the street suddenly leaves you no attention for talk beyond mindless "unh hunh, sure-- shut up!"
That's why pilots (for instance) drill relentlessly on procedures. In an air emergency you've already got way too many things going on requiring thought for you to waste any on trivia like turn mechanics. (The movie Apollo 13 alluded to this rather nicely.)
How this relates to UI design is simple: the UI is there to let you get other things done, not to be an end in itself. Someone writing should be thinking about the words, not about "now where is the control for..."
Of course, as NASA has proven, any task can be made automatic with enough training. It's a learning-curve sort of thing, with one of the big rules being that you don't surprise the user. If dragging a file from one folder to another moves the file in one case it had danged well better not copy it in another. Far better if, having learned to move things by dragging, dragging ALWAYS moved them.
An interesting question is whether there are any parties out there with the cojones to act on Bernstein and "Publish and be damned!"
The decision matrix on this is interesting. Will the USgovt wait until years have passed and the USSC has ruled, and then bring charges? How many juries will convict given the Defense pointing out that the Defendent was acting in accord with the law as decided both in Court and on appeal?
On the other hand, the USgovt could move for an injunction. That would take a lot of confidence to go before a judge and try to explain that irreperable harm would be done by exporting a copy of source code that originated on a non-US server and will continue to be on that server no matter the Court's decision. The whole proceding would be a Heaven-sent opportunity to lampoon all of the nonsense arguments in front of someone whose very job description requires filtering through BS.
[earthworm jim] Better than pro wrestling! [/earthworm jim]
Integrated circuits are little rectangles of silicon and have to be connected to the rest of the world. On the chip itself there are exposed metal spots ("bond pads") connected to the device's I/O circuitry. Various means are used to connect the bond pads to the package, but the most common are veeeery thin gold wires attached to the chip at one end and a package conductor at the other.
Since it's inconvenient and expensive to maintain separate IC designs, it's often easier to have one design which can be used in multiple ways depending on external signals, jumpers, or whatever. Rather than bring these configuration lines out of the package, though, they can be connected internally at the time the bond wires are attached. Likewise, signals that aren't needed in a particular configuration can be left out of the package I/O set.
Hate to break the news, but the problem isn't really with the Camino north bridge chip. It's a signal integrity problem with the Rambus system architecture. Nasty combination of crosstalk and a resonant mode in the data lines that takes received data out of signaling spec (the line fails to cross above the logic threshold when the RAM is sending a HIGH.) Intel is taking the fall on this one as though it's a silicon problem because it doesn't really matter where the problem is, the 820-based product isn't going to ship. Inside of Intel there's a big bloodletting going on between the engineers and the suits, because the suits are having a hard time dealing with the concept that there are some things that can't be changed by management fiat and the engineers aren't real amused by egos under the delusion that they can order back the tide. There's been some good discussion on this over on SI-LIST
Via has a DDR-200/266 chipsetfor the K7 on the ramp right now. The memories are shipping but not on the shelves due to lack of sockets. (All new SDRAM parts are actually DDR with a bond-out option)
Not really surprising. Engineering by fiat fails if trumped by a Higher Authority (e.g., physics)
The problem isn't the Camino chip, it's the physics. Turns out that Rambus has a major signal-integrity failure mode that sort of got swept under the rug until the systems houses got bit by it.
The DRAM companies never liked Rambus, but had their arms twisted by Intel. Now they have a chance to bail and are taking it.
The comment about DDR (double-data-rate SDRAM) having no standard will come as quite a surprise to the people at the memory companies and in particular JEDEC's JC-42 memory committee, which thinks that they have issued one, and AMII, which is sponsored by the memory industry (including NEC and Samsung) to promote its use.
For doubters, there's confirmation of sorts at ZDNN
To me, part of the question is whether this is the price of getting AOL to back off the pressure for open access or whether Steve Case will be at least follow through on all the rhetoric of the last few months.
As it stands, @Home is the only high-bandwidth connection to my neighborhood and their AUP doesn't allow me to hook up my home Linux network -- I'm stuck at 56k for the forseeable. (Yeah, I know they don't enforce their AUP but I'm funny that way.)
Slashdot Mirror
No doubt Notes is the greatest thing ever. Certainly EDS thinks so, 'cause they talked Corporate into switching everything over by the end of the year. All of which I might laugh at except ....
There's no Unix client, which means that my e-mail access is about to go away. This doesn't seem to be part of EDS' roadmap, and I can see that the problems of insignificant engineers might not concern them. I, on the other hand, do at least half of my work via e-mail in one form or another and for some reason don't share their opinion wrt its importance.
Which raises the question: does anyone know of a way to work around a Notes server and get mail, short of external Webmail services?
Rules can always be manipulated.
So far it's only in the discussion stages, but the FEC's rules on independent TV spots is ripe for abuse. The scheme goes like this: Assume that you are trying to defeat the Hon. Cross Palmer. You register as a PAC and place ads showing archive footage of Rep. Palmer at his stupidest, along with double-meaning captioning written by a really good satirist. The upshot is that the Palmer not only becomes a public laughingstock, but YOUR budget gets charged against HIS spending limits.
It's the TV equivalent of a Joe-job. The same tactic works even better on the Net. You register as a PAC and put up a page on Geocities with the Best of Palmer done in slack-stick and then report to the FEC listing the Geocities server farm as a contribution. Now multiply by the thousands of people who think that the public has been serviced enough by Palmer.
MUCH more practical and cost-efficient than television. But then we knew that, didn't we?
Let's face it, folks -- there are two fundamental problems with software patents. One is that doing a search to avoid infringement is a horrible resource hog, and the other is that (a la Ted Sturgeon) 95% are junk.
Both of these problems are solvable, though, in the same sense that "with enough eyes all bugs are shallow." A distributed project to review software patents would go a short way to index and cross-reference the pests, but it could go a long way toward pointing out how silly most of them are. Most of these toads got past the Patent Office because the PO doesn't review against unpatented prior art, and until recently none of the prior art was patented.
So! The trick is to scan through the patent database for software patents. Sort them into categories. Apply an open review process which rates patents by impact (e.g., a patent on linked lists) and if possible identifies prior art.
With enough ludicrous examples we might even be able to enlist some press to shame Congress into dealing with the problem.
GuidoDKP wisely observed:
People have been predicting the death of visible-light lithography for years now, but it's so useful that people keep coming up with ways to extend its life -- I think the latest enhancement was the use of phase-shifted interference patterns to draw features with visible light that were smaller than the wavelength of the light used...
Yup, phase-shifting is the latest trick but it's been pushed about as far as it can go. (My employer uses it and little bitty polygons are what I get paid for, at least in part.) Although "visible light" is a stretch when you're talking about ultraviolet.
Spiffy ultrahigh resolution line-drawing is monumentally valuable, though, if only as a means of producing those next-generation mask sets.
So -- wouldn't want to misrepresent your position -- you'd be OK with the local 'phone company only providing modem service to their own ISP subsidiary?
I know you're thrilled at the idea of getting cable, but will you still be thrilled when they tell you that to get & stay connected you have to run Win98SE? Terms of service, contracts, stuff like that.
Their cable, their rules (even if they put it in with a City charter). We're still doing 56K.
rde groused:
One thing that grated about the article; speaking about four-colour nanolithography. When you're dealing with molecules, colour loses its meaning, as the wavelength of visible light is up to twenty-five times the width of the line being drawn. Of course, you could draw lots of lines, but that'd take a hell of a long time.
I think they're talking about the method, not the result. Four-colour lithography depends on having chemically distinct dots Real Close Together.
Another reason to think in terms of four 'colours' is the famous topological theorem.
These guys could end up buying MICROS~1 out of petty cash. From the sounds of it they've found the semiconductor fabrication Philosopher's Stone: a way to do sub-X-ray physical patterns. Right now semiconductor feature sizes are in the 120+ nanometer range and severly limited by quantum optical effects. If they can lay down a precise grid at 15 nm and use it to sharpen up the lithography, they are going to own the semiconductor industry.
Gartner Group:
IDC figures show Linux comprised just 3.5 per cent of PC server shipments last year.
Translation:
Linux accounted for only 3.5% of the $$$ billed for OS orders last year. (IOW, for each 1000 copies of NT at ~$700 each there were 490 copies of Red Hat shipped at $50 each, and who knows how many times each of them was installed.)
Alternate translation:
Linux preinstalls accounted for only 3.5% of the $$$ billed last year for shipped servers. (Which is a lot more preinstalls than I would have expected, and why I doubt this translation.)
The thing to remember is that the Gartner Group is a marketing research firm, and they track MONEY, not UNITS. That little detail was one of the main reasons that IBM and Compaq kept their top spots for as long as they did against higher-volume but lower-priced competition.
Kintanon asked:
Ok, I might get flamed to hell for this, but... ummm What does powerpoint DO?? I know a lot of the marketing people around here use it, but I've never seen it DO anything? Help me out here?
Powerpoint absorbs huge amounts of time that management, marketers, and other suits might otherwised spend doing real harm.
The brain is one seriously ad-hoc architecture. At a basic level, there is impulse: smell coffee, raise cup. To avoid having this cause problems in awkward circumstances (like when using hands for more interesting things [more interesting than coffee? Hmmmm]) there is a review process (think gatekeeper) that checks impulses before execution.
This gatekeeper, being a relative latecomer in evolutionary terms, isn't particularly robust. Lots of things can take it offline, such as alcohol, which explains why a depressant (alcohol) can act as an apparent stimulant.
Likewise, people with ADHD actually have soporific gatekeepers and thus for them the very thought is the deed ("No, Honey, I really didn't mean to spill coffee all over that white shirt while we were kissing!") Stimulants "wake up" the 'gatekeeper' and allow the subject to stop foot before oral insertion.
Lots more of this over on alt.support.attn-deficit
HTH. HAND.
Actually, this is specsmanship. Where IEEE1394 does its arbitration, framing, etc. at a minimum of 100 Mb/s and usually at 400 Mb/s, USB of any flavor does its polling (notice, it is *not* a peer interconnect) and framing at no more than 12 Mb/s (assuming that there are no low-speed 1.5 Mb/s devices in the path).
Each packet in USB2 starts out with the same preamble (12Mb/s) and ends with the same postamble (12Mb/s); the new twist is that there may be a tone burst in the middle of 480 Mb/s or whatever. Unless you ship really huge packets (which aren't allowed by the protocol anyway) the average transfer rate is a lot lower than 408 Mb/s.
Don't uncritically accept the story that Rambus is a superior technology. Dell (a devout Rambus house) did some very interesting benchmarking, and although Intel and Rambus had the results censored from the IDF procedings you can see a copy at InQuest Market Research
Bottom line: Rambus appears to be substantially (like 15-40%) slower than PC100 SDRAM for typical applications. Oops.
laymil wrote:
Rambus, the company that developed the technology was not the source of the problem. It was Intel, and their 820 chipset, which is why it was the duty of Intel to fix the problem. So there.
Actually, there doesn't seem to be all that much wrong with the 820, aside from some speed limitations in the RAC. The big gotcha is a PWB-level signal integrity problem with the reference implementation of Rambus which wasn't anticipated by the relatively superficial signal-integrity analysis that the Rambus gang did.
In both cases new motherboard layouts will be needed, and since both will take up more space the whole floorplan may change. At best, this will take a few months to get the MBs designed, through validation and regulatory approval (not a trivial issue with this kind of bandwidth!) and into the production pipe. Kiss Q4 goodbye and probably Q1; the memory shops won't be seeing any demand until Q2 at soonest, if at all.
On top of that there will always be the charming issue (which Rambus seems to have in other areas as well) that the operating area for the memory subsystem will have a Swiss-cheese character. Instead of a 'schmoo' plot, with a maximum frequency of operation and constraints on voltage and temperature, there will be areas of operation and failure, alternating. Maybe 300 MHz and 800 MHz will be OK, but 700 will be out. In fact, that seems to be the situation right now.
The best guide for UI design I ever learned was from an educational psychologist who explained it this way:
..."
For any of us at any given time, a task is either associative or cognitive. Associative tasks are basically stimulus/response patterns while cognitive ones require some symbol manipulation (eg internal dialog.) At any time you can have an almost unlimited number of associative tasks going but only one cognitive one. Thus when you're driving along a familiar route in good weather and light traffic you can have an interesting discussion going with your passenger(s) but bad weather, an unfamiliar neighborhood, or the kid running out into the street suddenly leaves you no attention for talk beyond mindless "unh hunh, sure-- shut up!"
That's why pilots (for instance) drill relentlessly on procedures. In an air emergency you've already got way too many things going on requiring thought for you to waste any on trivia like turn mechanics. (The movie Apollo 13 alluded to this rather nicely.)
How this relates to UI design is simple: the UI is there to let you get other things done, not to be an end in itself. Someone writing should be thinking about the words, not about "now where is the control for
Of course, as NASA has proven, any task can be made automatic with enough training. It's a learning-curve sort of thing, with one of the big rules being that you don't surprise the user. If dragging a file from one folder to another moves the file in one case it had danged well better not copy it in another. Far better if, having learned to move things by dragging, dragging ALWAYS moved them.
An interesting question is whether there are any parties out there with the cojones to act on Bernstein and "Publish and be damned!"
The decision matrix on this is interesting. Will the USgovt wait until years have passed and the USSC has ruled, and then bring charges? How many juries will convict given the Defense pointing out that the Defendent was acting in accord with the law as decided both in Court and on appeal?
On the other hand, the USgovt could move for an injunction. That would take a lot of confidence to go before a judge and try to explain that irreperable harm would be done by exporting a copy of source code that originated on a non-US server and will continue to be on that server no matter the Court's decision. The whole proceding would be a Heaven-sent opportunity to lampoon all of the nonsense arguments in front of someone whose very job description requires filtering through BS.
[earthworm jim]
Better than pro wrestling!
[/earthworm jim]
An AC asked,
What is a bond-out option?
Integrated circuits are little rectangles of silicon and have to be connected to the rest of the world. On the chip itself there are exposed metal spots ("bond pads") connected to the device's I/O circuitry. Various means are used to connect the bond pads to the package, but the most common are veeeery thin gold wires attached to the chip at one end and a package conductor at the other.
Since it's inconvenient and expensive to maintain separate IC designs, it's often easier to have one design which can be used in multiple ways depending on external signals, jumpers, or whatever. Rather than bring these configuration lines out of the package, though, they can be connected internally at the time the bond wires are attached. Likewise, signals that aren't needed in a particular configuration can be left out of the package I/O set.
HTH.
Hate to break the news, but the problem isn't really with the Camino north bridge chip. It's a signal integrity problem with the Rambus system architecture. Nasty combination of crosstalk and a resonant mode in the data lines that takes received data out of signaling spec (the line fails to cross above the logic threshold when the RAM is sending a HIGH.)
Intel is taking the fall on this one as though it's a silicon problem because it doesn't really matter where the problem is, the 820-based product isn't going to ship. Inside of Intel there's a big bloodletting going on between the engineers and the suits, because the suits are having a hard time dealing with the concept that there are some things that can't be changed by management fiat and the engineers aren't real amused by egos under the delusion that they can order back the tide.
There's been some good discussion on this over on SI-LIST
Via has a DDR-200/266 chipsetfor the K7 on the ramp right now. The memories are shipping but not on the shelves due to lack of sockets. (All new SDRAM parts are actually DDR with a bond-out option)
Check out the BAT keyboard. One-handed, never take the other off of the mouse.
For doubters, there's confirmation of sorts at ZDNN
To me, part of the question is whether this is the price of getting AOL to back off the pressure for open access or whether Steve Case will be at least follow through on all the rhetoric of the last few months.
As it stands, @Home is the only high-bandwidth connection to my neighborhood and their AUP doesn't allow me to hook up my home Linux network -- I'm stuck at 56k for the forseeable. (Yeah, I know they don't enforce their AUP but I'm funny that way.)