I'm assuming that somewhere in North America, some power plant will have missed something that matters - maybe not on Jan 1 (e.g. the oiler-widget that's scheduled to oil hourly is expecting to oil something at 12:30am 1/1/100 which is a long time after 1/1/00, but the valve it oils doesn't seize up for a few weeks), but eventually the plant will crash or degrade. The question is how much of the power grid it takes down with it, and how good each power company is at isolating its working plants from the dead parts of the grid. Northern California's PG&E promises it understands that, and has some reassuring words about how this isn't happening during the summer air-conditioning crunch so the system should be pretty stable anyway.
Does your power plant have good plans for that? Or only for not crashing?
I also expect some electic company's billing system to get confused, but shutting off people's power is almost always manual, and any power company that sends around trucks to turn off anybody's power on Jan 2 for non-payment will deserve their Darwin Awards:-)
This week's X-Files was just Brain-Eating Monster Of The Week. But the Preview of NeXT Week'S X-Files has them meeting Frank Black of Millenium, Chris Carter's Other Show!
Just sit right back and you'll hear a tale Of the doom that is our fate. That started when programmers used Two digits for a date. Two digits for a date.
> When you restore power to a computer, it is immediately ready to go - no booting required.
A friend of a friend had a PDP-8 in his garage. When he had it all put back together and powered it up, he didn't need to boot it - the operating system was still sitting happily in genuine magnetic-donut core memory:-)
AT&T fired up the OC-192 POS about a week ago. The first segment is Cambridge MA to New York. We're also running OC-48 POS across the country. The OC-192 boards for the Cisco GSR are brand new, - OC-48 support is much more mature in both the Internet and fiber mux markets, and even OC-48 is still a very large pipe. But internet traffic keeps doubling every 15 minutes or so, so it'll fill up soon enough:-)
Sure, it may cost you some small amount of money to receive the spam, but for most people that's really minor - a few minutes of connect time a week, which for most Americans is free. It also costs some money for your ISP to receive the spam, since they're receiving larger quantities of it, but most ISPs have fixed-size data lines and there's near-zero incremental cost for more traffic on them. Disk drives are also nearly free.
The real cost of spam is the amount of time you spend dealing with it. It's those mental cycles you have to spin looking at mail on your mailing lists that's just spam, or building filters to auto-trash things marked M8K3 M0NEY FA$T, and the sheer annoyance that the spammers don't mind wasting your personal time with their junk, because some small percentage of the readers are suckers who will buy what they're selling.
Some spam is worth hunting down - particularly anything selling spamware or pyramid scams that will lead to lots more spam. It helps if there's an obvious internet address that the spammers are using to have suckers contact them - traceroute makes it easy to find the service providers and ask them to shut down the accounts.
> Attach Bricks/etc to Business Reply Mail It's been said since the 70s that the Post Office doesn't actually return bricks and similar junk attached to reply envelopes. Also, the Post Office no longer accepts packages heavier than 16 ounces in mailboxes - you have to physically hand them to a post office person, because otherwise they might be bombs. So no more 69-pound reply mails.
Obviously most people won't carry them as laptops, but as technology drives costs down, we'll eventually get decent resolution at affordable pri ces. For me that means 300dpi, 8.5x11 for a flat pad, or maybe 19" for a desktop.
My favorite two monitors have higher resolution
Tek 4014 (?) Green Plasma Display Storage tube - A late-70s design, but they had 4K x 4K resolution. They were essentially computer-driven Etch-A-Sketches - you drew on them either with the computer or with thumbwheels, and pressed a button to erase the screen instead of turning it over and shaking.
Sony 2048x2048 36" flat-screen CRT - late 80s, designed for the Air-Traffic-Control industry, amazingly gorgeous color. They still make them (about 1000 a year, about $30K each), and they've recently been trying to get the commercial market more interested, using them for applications like digitized art galleries. I saw one recently showing a landscape - the colors reminded me of Saxony (I was slightly wrong - it was actually Bavaria, but it had the quality of color to convey that depth of precision.)
A decade or so ago, Doug Gwyn wrote "ash", the adventure shell, a somewhat calmer version of the same idea.
$ xyzzy You are in your home directory. A directory named foo leads downward. There are files here. $ throw bar at lineprinter The fierce lineprinter daemon digests your file.
The dust designers are at Livermore Labs - of *course* there's a dark side. The Labs has been spreading its focus to topics other than nuclear weapons design and the scientific research behind it (and the environmental research into how to clean up places like Livermore), but they're still fundamentally a research institute for the military. So as the author said, "deal with it".
We knew this sort of technology would be around eventually anyway. It's interesting that it's this close to reality already. For some fictional treatment of things to do with smart dust, see "A Deepness In The Sky" by Vinge, as well as "The Diamond Age" by Stephenson.
My mom's vision is getting worse, and a friend of hers offered her a TV-sized Fresnel magnifier.
We didn't install it on the Macintosh (where the vision problems are more important) - there wasn't enough physical space on the desk to install the Fresnel and still sit in the chair with the keyboard on the desk and get reasonable viewing distance. Also, the Mac's screen wasn't bright enough, and the monitor keeps going semi-blue, so I set her screen fonts bigger, and printer fonts bigger, and eventually a new monitor will help.
So we installed it on the TV. It's just a big curved Fresnel with some metal-rod hangers that you hang 6-12 inches in front of the TV screen. Works pretty well - you have to sit straight in front of the TV to see it, but the picture does get about 20-30% bigger for reasonable adjustments. It's still bright enough, and being a bit fuzzy around the edges is ok for TV, even though it would be really annoying for reading text. Vision's funny about things like that.
Various people have suggested getting a cheap laptop as a remote terminal, and running a terminal emulator on top of DOS/Windows/MacOS. Most current laptops have infrared capability, so depending on your physical configuration you may be able to use this instead of wire.
Another option is radio - WebGear Aviator is a ~2Mbps short-distance radio system that plugs into a parallel port, and they have other models for USB and PCMCIA. I don't know if there are Linux drivers yet or not. Their older models tend to be on sale at Fry's for about $100-150/pair. This means you can have the real PC in your computer room, but sit somewhere comfortable around the house with your laptop.
There are also phone-line and power-line systems for short-distance communications. Most of them are ~19200 instead of megabit speeds, but that's fine for terminal emulation and many programming activities.
Last time I saw a dumb terminal at a computer fair or the physical stores for Halted Specialties or Weird Stuff, it was about $10-20 for a wyse or equivalent.
Old Macintoshes are about $25. They're not as portable as a laptop, but they're a bit smaller than most dumb terminals, and they're cute.
There are all sorts of small portable computers, laptop-like things, industrial portables, penpads, etc. which can be cheap because they're obsolete but are still good enough to use as a remote terminal while sitting on your back porch; some of the industrial models can even be used in your hot tub:-)
Unfortunately, most older portables used specialized batteries, typically NiCad, and the batteries are often dead or near-dead and new replacements can't be found, so you either have to do battery hacking or run them off A/C, and eithe r way you tend to void the watertightness that lets you use them in the tub, as well as being less convenient.
My Psion 3A pocket organizer has a real keyboard and a screen that's legible at 25x80 (though it's pretty small print), and there's free vt100 terminal software available. It's particularly nice for those times that you need to connect to a router console or server computer that you'd normally access with telnet. The Psion is a wonderful beast, with an interesting operating system, good physical design, and a small but fanatical bunch of supporters putting out free and commercial software. Psion Web Ring
The 3A has a weird I/O requiring a somewhat expensive adapter cable, but the 3C uses a standard RS-232 cable (and also infrared!) and has a backlit screen. The 5 is a major redesign and can do lots of Internet things on its own. You can pick up a used 3-series model for about $150-200.
Of course, you could get a Palm Pilot and one of the keyboards that goes with it; just because I'm a Psion fan doesn't mean I don't know where the market leader is:-) and the Pilot does seem to be well-designed.
This kind of topic has been extensively discussed on the cypherpunks, cryptography, cyberia-l, and other mailing lists.
Unfortunately, Anonymous Coward is correct that, if an American citizen does this, it counts as technical assistance and is therefore as illegal as any other means of providing unlicensed crypto to foreigners (assuming the work is made available for access by foreigners, as opposed to fixing the crypto software at a foreign office of a US company, or fixing a product that was legally exported to a foreign bank or Friend Of The US Military-Industrial Complex.
The validity of US export laws is debatable (at least if you don't take the First Amendment seriously, which the government doesn't, but until it gets totally thrown out we're stuck dealing with it.) Some people are challenging it, and some people take the First Amendment as their defense and ignore it. However, if you're a company that's subject to other regulations, as all corporations and many non-incorporated businesses are, you have to take a conservative approach to avoid being either shut down directly or bankrupted by the cost of a legal defense.
Some companies or private individuals, like C2.Net software and John Gilmore's FreeS/WAN Linux IPSec project, take the approach of hiring non-US-citizens to develop the product outside the US - it's legal to import the crypto into the US, and the people providing the money aren't providing technical assistance to foreigners, they're customers getting technical assistance from foreigners. FreeS/WAN has taken an especially careful approach with this, because they want their product to be unquestionably legal for anybody to use, whether inside or outside the US.
Whether the export is detectable or not is a separate issue - Not Getting Caught is a different problem than Not Violating Bogus Rules:-)
Cable ISPs route your IP packets to the Internet, where they can reach any information providers you want. An ISP isn't a monolithic service, if in fact it ever was. ISPs provide several important services: - Getting your packets to and from the Internet - Sending you a bill every month covering the costs of service plus a profit - Selling eyeballs of happy customers to advertisers and content providers. Other service providers use the Internet to provide content, web hosting, or mailbox hosting. AOL, in fact, provides AOL content service to people who use other ISPs to get dial access, and charges less for it than for content-plus-dial, and you can use this even faster over cable than over dial. The problem is whether users think of them selves as happy AOL users, which AOL can sell to advertisers, or whether they think of themselves as Internet users or hotmail users or Geocities users or Metricom users.
The right way* to provide cable modem service, technically, is to use the cable as a neighborhood Ethernet and use some combination of bridges and routers to get the IP packets to the outside world. How close the routers get to the end users and how much bridging you tolerate in the network are engineering and/or religious issues. How much concentration happens before you connect to something other than the Cable ISP's router network is a service quality issue, and there's probably a market need for different service quality / price options. For anything other than a local cable company, it makes sense to have multiple peering points, whether private or NAP-based peering, to provide better performance and reliability.
But AOL and its friends aren't arguing about technology - they're not trying to get separate frequency channels to run their own cable modems over, and they're not arguing about how much bandwidth exists between end users and AOL's peering points or saying that they should provide the wires to provide regional concentration to every cable head end. They're arguing about whether Excite@Home provides you an email mailbox that doesn't have advertising banners on it along with your IP service (as opposed to Excite providing you an email mailbox with advertising banners) because that makes you a happy (or disgruntled) @Home customer and a couple bucks from your cable modem bill pay for it, rather than you being a Happy AOL Customer that uses uninteresting cable modem the way other Happy AOL Customers use the uninteresting phone company to reach AOL.
[* There are some people who think that running ATM from the cable modem to the head end or from the head end to the various service-provider routers is a better technical approach, similar to the way DSL services run. I don't agree with them, but they're not part of the fundamental AOL-vs-Cable squabble here, though their technology is a better match to a multi-provider model.]
[Disclaimer: I work for one of the affected companies, though not in that part of the organization. If this were their opinion, they'd have herds of lawyers and lobbyists and PR people saying it, instead of the apparently less convincing things they've been saying over the past six months:-)]
At 07:38 PM 7/13/99 -0400, Seth Finkelstein wrote: > THERE IS NO GLOBAL EMAIL TAX IN THE WORKS! .... > I hope I've helped stop an urban-legend-in-the-making, but I'm >scared that the meme is going to be just too attractive.
Rumor-squelching is a valuable activity:-) Based on the later press release from the UN about how they're not planning to do a global email tax, it sounds like they've decided they need to squelch the rumor also. And somebody's posted your Cyberia posting to Slashdot.
> This looks likes the sort of thing which will get vectored by >the Libertarian and the gullible (by no means disjoint sets!),
This, on the other hand, was in bad taste, the sort of thing I'd only expect to see from people who engage in obsessed flamewars. (Which doesn't mean I _haven't_ seen some noisy discussion of it on cypherpunks, or that it wasn't on slashdot on July 13, or that I don't expect to get several forwarded copies this week.) Gullibility is also popular among Socialists, Liberals, Statists, MIT students, and whatever other categories it'll take to be sure I include Seth's friends:-)
The study's author, ostensibly an economist, probably should have known that 10KB isn't a very long document, but certainly should have noticed that any tax that can extract $10B out of Belgium alone is not a small tax, and that anything of that magnitude is also grossly market-distorting. On the other hand, that was one of a variety of suggestions for funding internet development that she mentioned, so perhaps that wasn't her preference.
Re:THERE IS NO GLOBAL EMAIL TAX... bandwidth tax?
on
UN Proposes Email Tax
·
· Score: 1
No, you did the math wrong - it was a tax of 1 cent per 10KB, which is $1/MB. Of course, the so-called economist who mentioned this (as one of many possible funding mechanisms) should have noticed that any tax capable of extracting $10B/year out of Belgium alone is a huge tax, not a small one, and that it would cause substantial market distortions. However, as Seth pointed out, that was just one suggestion, not a proposal.
Graphics Performance vs. Number-Crunching?
on
Carmack on the K7
·
· Score: 1
Carmack's article appeared to focus on the speed the K7 provided to gamers. Given his background, that's no surprise:-) But gamer speed really depends on the graphics cards, motherboards, and connectivity between them and the CPU more than it does on raw CPU speed. So how fast is the K7 (vs.PIII) at integers and floating point? Are they roughly balanced, or is this one of the chips that's faster at integers at the expense of floating point? Can it do MMX or other parallel execution tricks? Are there specmarks or other cpu/memory intensive benchmarks?
My V7 manual being in a box somewhere, I looked at my copy of "Unix User's Manual, Release 3.0", from June 1980, which is approximately the AT&T System III stuff. sdiff is there.
The patent lists 22 claims, in the traditional "write the software patent as if it were a hardware device patent" style of obfuscation. Some of those claims are for the diff algorithms, some are for the display stucture, and some are for the mechanism for getting from the diff to the screen.
According to the article, "Advanced Software said a 1989 patent and 1998 patent reissue cover an invention by employee Cary Queen that offers comparison of documents in original and modified versions in split-screen format."
While that's not an explicit description of their complaint, it sounds much more like they're complaining about the display format than the diff algorithms (which aren't clearly explained in the patent claims) and there were certainly many methods for diffs by the mid 80s, most unpatented.
The original Mac was a Motorola 68000 with a small screen, ROMs, floppy drive, keyboard, and mouse. The original Palm Pilot was a Motorola 68000 with a screen, more memory, ROMs, and a touchscreen instead of a keyboard+mouse. It might need a slightly better screen, and you'd have to add some extra programming for the Graffiti drivers, and a compact flash instead of floppies, but if you can make a Pilot run Linux, you should be able to make it run MacOS:-)
There have been hoax rumors going around the US and some other countries (Canada or Australia, I think) about the Post Office planning to impose a fee on email. The Telecom New Zealand thing could be a hoax or generic bogus rumor, or it could be real. Telecom providers who offer flat rate calling and have engineered for typical voice traffic don't like model-breaking things like modem usage - but they do get to sell people extra phone lines for their modems, which they do like.
I've been running the Mersenne Prime search http://www.mersenne.org for a couple of years now. All of these systems are fine if you're running on AC power, but CPU-burners do Really Bad Things to batteries! I'd guess it cost my employer one or two NiMHs before I figured this out:-) Running ofr short times is fine, or draining your NiCad to the bone if you want, but NiMH memory doesn't seem to like that kind of abuse. I don't know how the newer Lithium batteries feel about it (but Lithium is no longer available on credit:-)
Also, if you're using a background-version program, rather than a screen-saver type, remember not to use a CPU-intensive screen-saver - pick something that draws or moves a picture every few minutes rather than something that's constantly moving, especially if it's doing complex graphics transforms.
Slashdot Mirror
I'm assuming that somewhere in North America, some power plant will have missed something that matters - maybe not on Jan 1 (e.g. the oiler-widget that's scheduled to oil hourly is expecting to oil something at 12:30am 1/1/100 which is a long time after 1/1/00, but the valve it oils doesn't seize up for a few weeks), but eventually the plant will crash or degrade. The question is how much of the power grid it takes down with it, and how good each power company is at isolating its working plants from the dead parts of the grid. Northern California's PG&E promises it understands that, and has some reassuring words about how this isn't happening during the summer air-conditioning crunch so the system should be pretty stable anyway.
:-)
Does your power plant have good plans for that? Or only for not crashing?
I also expect some electic company's billing system to get confused, but shutting off people's power is almost always manual, and any power company that sends around trucks to turn off anybody's power on Jan 2 for non-payment will deserve their Darwin Awards
>Y2K movie, or new X-Files?
This week's X-Files was just Brain-Eating Monster Of The Week. But the Preview of NeXT Week'S X-Files has them meeting Frank Black of Millenium, Chris Carter's Other Show!
John Walker's Millenium Screen Saver
Two Digits for a Date. by Paschal S Hammond. (to the tune of "Gilligan's Island," more or less)
can be easily found at http://www.hamline.edu/~math/ wnk/humor/twodigits.html, or at other places without the author's name.
Just sit right back and you'll hear a tale
Of the doom that is our fate.
That started when programmers used
Two digits for a date.
Two digits for a date.
and a page or two more.
> When you restore power to a computer, it is immediately ready to go - no booting required.
:-)
A friend of a friend had a PDP-8 in his garage.
When he had it all put back together and powered it up, he didn't need to boot it - the operating system was still sitting happily in genuine magnetic-donut core memory
The offical PR is at http://www.att.com/press/item/0,1 354,2228,00.html
Sure, it may cost you some small amount of money to receive the spam, but for most people that's really minor - a few minutes of connect time a week, which for most Americans is free.
It also costs some money for your ISP to receive the spam, since they're receiving larger quantities of it, but most ISPs have fixed-size data lines and there's near-zero incremental cost for more traffic on them. Disk drives are also nearly free.
The real cost of spam is the amount of time you spend dealing with it. It's those mental cycles you have to spin looking at mail on your mailing lists that's just spam, or building filters to auto-trash things marked M8K3 M0NEY FA$T, and the sheer annoyance that the spammers don't mind wasting your personal time with their junk, because some small percentage of the readers are suckers who will buy what they're selling.
Some spam is worth hunting down - particularly anything selling spamware or pyramid scams that will lead to lots more spam. It helps if there's an obvious internet address that the spammers are using to have suckers contact them - traceroute makes it easy to find the service providers and ask them to shut down the accounts.
> Attach Bricks/etc to Business Reply Mail
It's been said since the 70s that the Post Office doesn't actually return bricks and similar junk attached to reply envelopes.
Also, the Post Office no longer accepts packages heavier than 16 ounces in mailboxes - you have to physically hand them to a post office person, because otherwise they might be bombs. So no more 69-pound reply mails.
Yup.
That was certainly my first reaction.
My favorite two monitors have higher resolution
of *course* there's a dark side.
The Labs has been spreading its focus to topics other than nuclear weapons design and the scientific research behind it (and the environmental research into how to clean up places like Livermore), but they're still fundamentally a research institute for the military. So as the author said, "deal with it".
We knew this sort of technology would be around eventually anyway. It's interesting that it's this close to reality already. For some fictional treatment of things to do with smart dust, see "A Deepness In The Sky" by Vinge, as well as "The Diamond Age" by Stephenson.
We didn't install it on the Macintosh (where the vision problems are more important) - there wasn't enough physical space on the desk to install the Fresnel and still sit in the chair with the keyboard on the desk and get reasonable viewing distance. Also, the Mac's screen wasn't bright enough, and the monitor keeps going semi-blue, so I set her screen fonts bigger, and printer fonts bigger, and eventually a new monitor will help.
So we installed it on the TV. It's just a big curved Fresnel with some metal-rod hangers that you hang 6-12 inches in front of the TV screen. Works pretty well - you have to sit straight in front of the TV to see it, but the picture does get about 20-30% bigger for reasonable adjustments. It's still bright enough, and being a bit fuzzy around the edges is ok for TV, even though it would be really annoying for reading text. Vision's funny about things like that.
Another option is radio - WebGear Aviator is a ~2Mbps short-distance radio system that plugs into a parallel port, and they have other models for USB and PCMCIA. I don't know if there are Linux drivers yet or not. Their older models tend to be on sale at Fry's for about $100-150/pair. This means you can have the real PC in your computer room, but sit somewhere comfortable around the house with your laptop.
There are also phone-line and power-line systems for short-distance communications. Most of them are ~19200 instead of megabit speeds, but that's fine for terminal emulation and many programming activities.
Last time I saw a dumb terminal at a computer fair or the physical stores for Halted Specialties or Weird Stuff, it was about $10-20 for a wyse or equivalent.
Old Macintoshes are about $25. They're not as portable as a laptop, but they're a bit smaller than most dumb terminals, and they're cute.
There are all sorts of small portable computers, laptop-like things, industrial portables, penpads, etc. which can be cheap because they're obsolete but are still good enough to use as a remote terminal while sitting on your back porch; some of the industrial models can even be used in your hot tub :-)
Unfortunately, most older portables used specialized batteries, typically NiCad, and the batteries are often dead or near-dead and new replacements can't be found, so you either have to do battery hacking or run them off A/C, and eithe r way you tend to void the watertightness that lets you use them in the tub, as well as being less convenient.
My Psion 3A pocket organizer has a real keyboard and a screen that's legible at 25x80 (though it's pretty small print), and there's free vt100 terminal software available. It's particularly nice for those times that you need to connect to a router console or server computer that you'd normally access with telnet. The Psion is a wonderful beast, with an interesting operating system, good physical design, and a small but fanatical bunch of supporters putting out free and commercial software.
Psion Web Ring
The 3A has a weird I/O requiring a somewhat expensive adapter cable, but the 3C uses a standard RS-232 cable (and also infrared!) and has a backlit screen. The 5 is a major redesign and can do lots of Internet things on its own. You can pick up a used 3-series model for about $150-200.
Of course, you could get a Palm Pilot and one of the keyboards that goes with it; just because I'm a Psion fan doesn't mean I don't know where the market leader is
Unfortunately, Anonymous Coward is correct that, if an American citizen does this, it counts as technical assistance and is therefore as illegal as any other means of providing unlicensed crypto to foreigners (assuming the work is made available for access by foreigners, as opposed to fixing the crypto software at a foreign office of a US company, or fixing a product that was legally exported to a foreign bank or Friend Of The US Military-Industrial Complex.
The validity of US export laws is debatable (at least if you don't take the First Amendment seriously, which the government doesn't, but until it gets totally thrown out we're stuck dealing with it.) Some people are challenging it, and some people take the First Amendment as their defense and ignore it. However, if you're a company that's subject to other regulations, as all corporations and many non-incorporated businesses are, you have to take a conservative approach to avoid being either shut down directly or bankrupted by the cost of a legal defense.
Some companies or private individuals, like C2.Net software and John Gilmore's FreeS/WAN Linux IPSec project, take the approach of hiring non-US-citizens to develop the product outside the US - it's legal to import the crypto into the US, and the people providing the money aren't providing technical assistance to foreigners, they're customers getting technical assistance from foreigners. FreeS/WAN has taken an especially careful approach with this, because they want their product to be unquestionably legal for anybody to use, whether inside or outside the US.
Whether the export is detectable or not is a separate issue - :-)
Not Getting Caught is a different problem than Not Violating Bogus Rules
Cable ISPs route your IP packets to the Internet, where they can reach any information providers you want.
:-)]
An ISP isn't a monolithic service, if in fact it ever was. ISPs provide several important services:
- Getting your packets to and from the Internet
- Sending you a bill every month covering the costs of service plus a profit
- Selling eyeballs of happy customers to advertisers and content providers.
Other service providers use the Internet to provide content, web hosting, or mailbox hosting.
AOL, in fact, provides AOL content service to people who use other ISPs to get dial access, and charges less for it than for content-plus-dial, and you can use this even faster over cable than over dial. The problem is whether users think of them selves as happy AOL users, which AOL can sell to advertisers, or whether they think of themselves as Internet users or hotmail users or Geocities users or Metricom users.
The right way* to provide cable modem service, technically, is to use the cable as a neighborhood Ethernet and use some combination of bridges and routers to get the IP packets to the outside world. How close the routers get to the end users and how much bridging you tolerate in the network are engineering and/or religious issues.
How much concentration happens before you connect to something other than the Cable ISP's router network is a service quality issue, and there's probably a market need for different service quality / price options. For anything other than a local cable company, it makes sense to have multiple peering points, whether private or NAP-based peering, to provide better performance and reliability.
But AOL and its friends aren't arguing about technology - they're not trying to get separate frequency channels to run their own cable modems over, and they're not arguing about how much bandwidth exists between end users and AOL's peering points or saying that they should provide the wires to provide regional concentration to every cable head end. They're arguing about whether Excite@Home provides you an email mailbox that doesn't have advertising banners on it along with your IP service (as opposed to Excite providing you an email mailbox with advertising banners) because that makes you a happy (or disgruntled) @Home customer and a couple bucks from your cable modem bill pay for it, rather than you being a Happy AOL Customer that uses uninteresting cable modem the way other Happy AOL Customers use the uninteresting phone company to reach AOL.
[* There are some people who think that running ATM from the cable modem to the head end or from the head end to the various service-provider routers is a better technical approach, similar to the way DSL services run. I don't agree with them, but they're not part of the fundamental AOL-vs-Cable squabble here, though their technology is a better match to a multi-provider model.]
[Disclaimer: I work for one of the affected companies, though not in that part of the organization. If this were their opinion, they'd have herds of lawyers and lobbyists and PR people saying it, instead of the apparently less convincing things they've been saying over the past six months
At 07:38 PM 7/13/99 -0400, Seth Finkelstein wrote:
:-)
:-)
> THERE IS NO GLOBAL EMAIL TAX IN THE WORKS!
....
> I hope I've helped stop an urban-legend-in-the-making, but I'm
>scared that the meme is going to be just too attractive.
Rumor-squelching is a valuable activity
Based on the later press release from the UN about how they're
not planning to do a global email tax, it sounds like they've
decided they need to squelch the rumor also.
And somebody's posted your Cyberia posting to Slashdot.
> This looks likes the sort of thing which will get vectored by
>the Libertarian and the gullible (by no means disjoint sets!),
This, on the other hand, was in bad taste, the sort of thing
I'd only expect to see from people who engage in obsessed flamewars.
(Which doesn't mean I _haven't_ seen some noisy discussion of it
on cypherpunks, or that it wasn't on slashdot on July 13,
or that I don't expect to get several forwarded copies this week.)
Gullibility is also popular among Socialists, Liberals, Statists, MIT students,
and whatever other categories it'll take to be sure I include Seth's friends
The study's author, ostensibly an economist, probably should have known
that 10KB isn't a very long document, but certainly should have noticed
that any tax that can extract $10B out of Belgium alone is not a small tax,
and that anything of that magnitude is also grossly market-distorting.
On the other hand, that was one of a variety of suggestions for funding
internet development that she mentioned, so perhaps that wasn't
her preference.
No, you did the math wrong - it was a tax of 1 cent per 10KB, which is $1/MB. Of course, the so-called economist who mentioned this (as one of many possible funding mechanisms) should have noticed that any tax capable of extracting $10B/year out of Belgium alone is a huge tax, not a small one, and that it would cause substantial market distortions. However, as Seth pointed out, that was just one suggestion, not a proposal.
Carmack's article appeared to focus on the speed the K7 provided to gamers. Given his background, that's no surprise :-)
But gamer speed really depends on the graphics cards, motherboards, and connectivity between them and the CPU more than it does on raw CPU speed.
So how fast is the K7 (vs.PIII) at integers and floating point? Are they roughly balanced, or is this one of the chips that's faster at integers at the expense of floating point? Can it do MMX or other parallel execution tricks?
Are there specmarks or other cpu/memory intensive benchmarks?
My V7 manual being in a box somewhere, I looked at my copy of "Unix User's Manual, Release 3.0", from June 1980, which is approximately the AT&T System III stuff. sdiff is there.
The patent lists 22 claims, in the traditional "write the software patent as if it were a hardware device patent" style of obfuscation. Some of those claims are for the diff algorithms, some are for the display stucture, and some are for the mechanism for getting from the diff to the screen.
According to the article,
"Advanced Software said a 1989 patent and 1998 patent reissue cover an invention by employee Cary Queen that offers comparison of documents in original and modified versions in split-screen format."
While that's not an explicit description of their complaint, it sounds much more like they're complaining about the display format than the diff algorithms (which aren't clearly explained in the patent claims) and there were certainly many methods for diffs by the mid 80s, most unpatented.
The original Mac was a Motorola 68000 with a small screen, ROMs, floppy drive, keyboard, and mouse. :-)
The original Palm Pilot was a Motorola 68000 with a screen, more memory, ROMs, and a touchscreen instead of a keyboard+mouse.
It might need a slightly better screen, and you'd have to add some extra programming for the Graffiti drivers, and a compact flash instead of floppies, but if you can make a Pilot run Linux, you should be able to make it run MacOS
There have been hoax rumors going around the US
and some other countries (Canada or Australia, I think)
about the Post Office planning to impose a fee on email.
The Telecom New Zealand thing could be a hoax or generic bogus rumor, or it could be real.
Telecom providers who offer flat rate calling and have engineered for typical voice traffic don't like model-breaking things like modem usage - but they do get to sell people extra phone lines for their modems, which they do like.
I've been running the Mersenne Prime search :-) :-)
http://www.mersenne.org
for a couple of years now. All of these systems
are fine if you're running on AC power,
but CPU-burners do Really Bad Things to batteries!
I'd guess it cost my employer one or two NiMHs
before I figured this out
Running ofr short times is fine, or draining your
NiCad to the bone if you want, but NiMH memory
doesn't seem to like that kind of abuse.
I don't know how the newer Lithium batteries feel about it
(but Lithium is no longer available on credit
Also, if you're using a background-version program,
rather than a screen-saver type, remember not to
use a CPU-intensive screen-saver - pick something
that draws or moves a picture every few minutes rather than something that's constantly moving, especially if it's doing complex graphics transforms.