Slashdot Mirror

O.K.,

Go to their web site and click on Groups. Then click on clubs and then Team Slashdot and then Join this group.

Better yet, just click HERE

O.K.,

Go to their web site and click on Groups. Then click on clubs and then Team Slashdot and then Join this group.

Better yet, just click HERE

Just go to http://setiathome.ssl.berkeley.edu/tea m.html. The rest is easy.

If you run OS/2, there is a command-line client available via http://www.os2ss.com/seti/. Like the Unix client, this version has no GUI and runs at a low priority so that it doesn't interfere with your regular work. It does have the option to run at regular priority.

There is also an OS/2 Warp group, available via http://setiathome.ssl.berkeley.edu/cgi-bin/cgi?cmd =team_lookup&name=OS%2F2+Warp. We currently have 211 members and we're in 32nd place.

Of course, you don't need to be running the OS/2 client to join the group. In fact, I have one Mac, 2 OS/2 PC's, and 1 NT PC all running the client under the same name.

--
Timur Tabi
Remove "nospam_" from email address

Some of the more interesting pieces, like LSF, are only licensed by Sun, thus will not be included in this `deal.' (For a free improvement over LSF, check out GNU Queue. If it doesn't do something you want, you can support the community and extend it.) If you read the announcement carefully, you'll see that the only new codes to which it applies are the parallel file system (the Sun CTO thinks distributed file systems are dead, anyways), the Prism debugger, and the parallel run-time environment.

Of those, the only with no available substitute is the debugger. The ROMIO library is a good place to start for the MPI file I/O stuff (a good database would be a better place, imho). I already mentioned queue management software. The Ptools Consortium and the Globus Project have links to other HPC cluster tools.

Many of the pieces for debugging are available (combine ddd and gnuplot), but some notable ones are missing. The ability to control multiple GDBs easily from one processes and the visualization of parallel execution are needed, and quite difficult to implement. There seems to be interest in making GDB easier to use from other processes, which is a good start towards solving the larger problem of general, distributed debugging. And both the mpich and LAM MPI implementation have some profiling information, but few tools to dig through it.

To be fair, Sun has contributed (and supported contributions) to the original packages. Why they are releasing the rest under their Exploit the Community license is beyond me.

Jason, ejr@cs.berkeley.edu

If it is the electrical bill you worried about, don't. Torn off monitor, wich is the biggest power drawer and let you're box help find aliens while it is idleing. I have mine installed as a screensvaer and whenever I am not around I am looking for aliens.... To get the program (in many flavors unix nt 9x) goto http://setiathome.ssl.berkeley.edu

Mirrored here.

http://www.ocf.berkeley.edu/~ fhsu/104_linuxcare_s.jpg

h ttp://setiathome.ssl.berkeley.edu/cgi-bin/cgi?cmd= team_lookup&name=team+slashduh

Currently, there are 4 members:

Name Email Work units sent Results received Total CPU time Average CPU time per work unit

Michael Dolan mtdolan@mtu.edu 1801 1533 14585 hr 49 min 16.2 sec ( 1.67 years) 9 hr 30 min 52.4 sec

UC Berkeley Millennium seti@mmnt.millennium.berkeley.edu 1028 826 12124 hr 45 min 26.7 sec ( 1.38 years) 14 hr 40 min 44.0 sec

xilef felix.rauch@nice.ch 938 641 7105 hr 38 min 53.5 sec 11 hr 05 min 06.9 sec

ARDWGC seti@ircache.net 719 557 5791 hr 55 min 57.8 sec 10 hr 23 min 54.4 sec

The concept just has a certain perfection to it. ;-)

Join the team here

We joined the Seti-At-Home project two years ago, for what that's worth, but the project itself has only just begun...They have problems with server overload fairly often; please be patient as they figure out how to deal with these typical new-project problems.

I wouldn't put too much value in the current ranking of the teams. The work units are not being added up correctly. Take a look at Team Sweden to see what I mean. You can see the stats for Team Sweden here. For the team it shows the total work units as 62, but the the top team member it shows the (correct) value of 80. I have already reported this bug using there bug report form, so it will be interesting to see how much the stats change once this is fixed.
---

For anyone from or interested in the beautiful country of Sweden, you may want to consider joining Team Sweden. At the moment I'm the first one registered for the team, but there is a group of us that was already working as a "team" before they setup real teams. Check out the stats for Team Sweden here. There are currently about 80 machines happily processing data.
---

http://setiathome.ssl.berkeley.edu/software/setiat home_win_1_0.exe

Similarly, the Mac client is at

http://setiathome.ssl.berkeley.edu/software/setiat home_mac_1_0.sit

Um. Have you read the background materials at http://setiathome.ssl.berkeley.edu at all? The SETI@Home project is being carried out by a team at the same university that is doing the Serendip IV project, which is the main SETI project previously underway. They are funded, they have published (very interesting) technical details about their methodology, and they are every bit as much a 'part of' SETI as is anyone else.

Are you upset because Jodie Foster didn't use SETI@Home on her personal computer in Contact or something? Who do you imagine decides who is 'affiliated with' SETI?

According to the published papers on their site, SETI@Home will examine as many possible signals as the Serendip IV project. Yes, SETI@Home will process a smaller frequency range, but it will examine it in much greater detail, which much more expensive computational analysis thanks to your computer and mine.

Incidentally, I'd recommend taking a look at the scientific papers linked to the SETI@Home site.. what they are doing to perform reasonable data analysis on signals picked up by their piggy-back receivers while the Areceibo telescope is in use and even in motion for other projects doing direct observation for traditional radio astronomy is fascinating. No wonder it took them so much longer to get the SETI@Home client out than it took the distributed.net people to get their network running.

Since they've reset the work stats, Windows 95 is now the leading OS for completed work units.

We can't let this stand for even one minute can we?!?!? C'mon /.'ers, crank up your boxen and make sure Mickeysoft gets shoved down far on the list where they belong!

Here is a link to the Windows EXE. Or, for the full directory, go here.

Although how long that will last I can't say ;)

/*He who controls Purple controls the Universe. *

Here is a link to the Windows EXE. Or, for the full directory, go here.

Although how long that will last I can't say ;)

/*He who controls Purple controls the Universe. *

Looks like they've reset the usage statistics like I thought I remembered they said they would. The top user now has a whole 4 work units. All my previous stats are gone (my pc at home must not be running it right now as I am still at zero). Hmm, guess i'll have to run home at lunch and see what the deal is.

The stats page is here.

I guess this is an incentive for new ppl to run this on as many machines as possible so they show up in the top users as before the to 20 ppl had something like, oh, 50 ga-zillion work units processed.

To get its hand on a good x86 design

Last year AMD had sale of something like 2.5 billion dollars. Last _quarter_, Motorola had sales of 7.2 billion. And Intel? 7.1 billion last quarter.

Precisely. AMD is too small to be taking on Intel. Motorola could take the K7 design and run with it.

The rumour mill has Compaq buying AMD. I think they need to be bought by someone who knows how to run a fab and who won't be in competition with AMD's customers.

x86 is a dying technology

Merced is rumoured to be a flop, McKinley is rumoured to be late, Alpha is still not taking off. Sure x86 is dying, but it has been for years and that hasn't stopped Intel making a fortune on it. Remember, Windows 2000 is either going to be W98/DOS-based or late or both, so where does that leave non-x86 designs for the mainstream? Some years off.

The PowerPC 750 is a damn fine chip.

I'm sure it is, and Motorola should keep building it, but its not where the volume is, and that's not going to change. And while they are doing OK, they don't have any sort of performance lead over the `dying' x86 chips: According to The CPU Info Centre they get maximally 17.6 SPECint95s at 400MHz in Motorola's 0.22um process, while Intel is well into the twenties at 0.25um (I think).

-- Give him Head? Be a Beacon?

Adam Beberg (duncan) wrote:
it has become apparent that the goals of DCTI have changed considerably over the years, and are no longer the same as what they were.

David McNett (nugget) wrote:
It has also become clear to us that Adam's goals for Cosm and distributed.net's vision of its future differ enough to justify this parting of ways. Adam is very motivated about seeing the system he's designed, Cosm, implemented and put into production. True to its name, distributed.net is more focused on seeing what can result from a truly open and distributed continuing development effort. While each of these respective approaches is viable in its own right, attempting to co-mingle them has proven to be counterproductive.

What's the division about? It's clear from these carefully worded pieces that they've decided to split, and they've put an amiable face on it. However, the announcements are so vague I can't tell what the real issues are.

One of the problems I've always had with the d.net project was their closed decision-making style. On the one hand, they've build this wonderful thing for running the DES/rc contests, and made it fun to participate. I think that's really nifty. On the other hand, they've been advertising 'v3' for a over a year now, with its plug-in architecture promising a wide variety of clients to choose from, and an open interface so one may write one's own. But it never arrived, partly, it seemed, because Beberg wouldn't let anyone else work on it. Sometimes I felt like they didn't want to allow any other clients because they'd lose people from the rc effort, which is what they're really interested it. I think this might not be such a problem given rc5's microscopic bandwidth, footprint, and tolerance for latency compared to alot of other distributed computing projects If you think seti@home is bad, try cg rendering, or scientific simulations! Even the Mersenne prime client is more efficient with a large (>~16MB) memory allocation.

Nugget speaks of "a truly open and distributed continuing development effort." Their hot new client is the OGR project, and still closed source. Beberg is at least publishing a programming interface, but hasn't specified a license yet.

I've always been bothered by d.net's interest in using my processor for their particular project, paying only lip service to giving (control) back to the community they created. I've always been bothered by their failure to grok open source development.

I guess what all this boils down to is that I'd like to think that either Beberg or others at d.net have seen some light in this vein, but I can't tell which of them it might be. Comments?

i took a lakoff class at berkeley (intro to cogsci) and his position boils down to this: the mind is embodied i.e. abstract thought is based on image schemas that arise from our physical interaction with the world. these basic schemas (see mark johnson's book the body in the mind) are NON-MODULAR in that they need to be seen as a gestalt to be understood i.e. the containment schema only makes sense when there is a containing structure and a contained/noncontained structure. these schemas are extended up via metaphor to form abstract thought. the levels are physical, social, epistemic, then speech-action. programmers tend to operate in the epistemic realm. metaphor is not just something from shakespeare, but is how we extend ALL knowledge from one realm to another. a list of the common metaphors in the western world is found here . This stuff is interesting because it turns out that artificial intelligence is not possible. lakoff and crew are doing interesting work called the neural theory of language project (NTL) where they are building software models of thought this is the page. more metaphor stuff is found here . lakoff and mark johnson have a new book called philosophy in the flesh , the premise is that western philosophy is based on the false premise that the mind exists independently of the body, that there is a gods-eye view. this stuff is NOT post modern however, as social facts can impact life etc. okay i need to get back to work, this is internet after all

i took a lakoff class at berkeley (intro to cogsci) and his position boils down to this: the mind is embodied i.e. abstract thought is based on image schemas that arise from our physical interaction with the world. these basic schemas (see mark johnson's book the body in the mind) are NON-MODULAR in that they need to be seen as a gestalt to be understood i.e. the containment schema only makes sense when there is a containing structure and a contained/noncontained structure. these schemas are extended up via metaphor to form abstract thought. the levels are physical, social, epistemic, then speech-action. programmers tend to operate in the epistemic realm. metaphor is not just something from shakespeare, but is how we extend ALL knowledge from one realm to another. a list of the common metaphors in the western world is found here . This stuff is interesting because it turns out that artificial intelligence is not possible. lakoff and crew are doing interesting work called the neural theory of language project (NTL) where they are building software models of thought this is the page. more metaphor stuff is found here . lakoff and mark johnson have a new book called philosophy in the flesh , the premise is that western philosophy is based on the false premise that the mind exists independently of the body, that there is a gods-eye view. this stuff is NOT post modern however, as social facts can impact life etc. okay i need to get back to work, this is internet after all

http://www.cs.berkeley.edu/~healey/MURI/ It may depend on this screen or not.. I'm not sure. It is interesting though.

Want to help look for proof of life in outer space? Check out http://setiathome.ssl.berkeley.edu/ and http://www.setileague.org/ for two ways that you can use your computer at home to speed the search.

"Starting an OpenSource project" isn't enough these days. We are seeing too much of duplicated clones of commercial software+sites, and many "projects" with a low ratio code/wishes.

It would be nice to have an OpenSource wavelet project, but you got to be aware that video/audio compression is a black art, that there is the beginnings of a GNU Wavelet Image Compression somewhere by a PhD student in Finland, and that for mono/multicast audio/video the Mash project is doing a good job (yes they have hybrid wavelet/DCT video codec).

Apple MrC with PowerMathLib:
NUMERIC SORT: Iterations/sec.: 107.164610 Index: 2.769469
STRING SORT: Iterations/sec.: 9.606252 Index: 4.222528
BITFIELD: Iterations/sec.: 3571869546.392807 Index: 612.688667
FP EMULATION: Iterations/sec.: 20.041185 Index: 9.635185
FOURIER: Iterations/sec.: 1602.031493 Index: 1.813977
ASSIGNMENT: Iterations/sec.: 1.510908 Index: 5.756497
IDEA: Iterations/sec.: 462.599684 Index: 7.077719
HUFFMAN: Iterations/sec.: 168.458324 Index: 4.681348
NEURAL NET: Iterations/sec.: 3.053514 Index: 5.166690
LU DECOMPOSITION: Iterations/sec.: 52.623917 Index: 3.106855

This from the bytemark of a PowerMac 604e/180Mhz, compared to a P90 (which would get index=1.0). They both came from a page at Apple, that I linked to slashdot and that seems to have been removed since (no longer available from Altavista). Do some calculations: take the BITFIELDS, iterations per second column ; compute the number of clock cycles taken by iteration considering the 180Mhz frequency of the processor ; conclude.

The "index" column shows that the PPC604e was 600.0 times faster than a P90 (code that you don't run is very fast indeed). This pretty much killed the benchmark. This is why bytemark is plain bogus.

A SIMILAR BYTEMARK BENCH WAS USED BY APPLE TO CLAIM THAT G3 ARE TWICE FASTER THAN PII. Actually they might have taken another compiler to make this claim, but the outrageously better results for the G3 are for the exact same reason, and Apple perfectly knew that its benchmarks were bogus.

Even PC Magazine, a company that almost no one would argue would be biased in favor of Apple vs. Wintel, admits that a G3/400 beats a Dell Dimension XPS T500 Pentium III on tests using an application *widely reported* to be heavily optimized for MMX and SSE! All that with a 20% slower clock speed!

If it is Photoshop benchmark, it is mostly a loop, so not very characteristic of typical applications. Also last time I saw the people added the time of the different filters, which is stupid because it resulted on the slower filters (i.e. unoptimised) being the very preponderant, independantly of how often they are used in real life. Other debatable benchmarks are http://macspeedzone.com/4.0/Wi nvsMacmathematica.html (we don't know which compiler is used, which optimizations, etc.).

Real benchmarks are SPEC CPU benchmarks (see www.spec.org). G3 are about 15% faster than PII at the same clock rate. And of course, a very important point is that all processors aren't available at the same clock speed at the same time. PIII/500 are commercially available that do 22 SPECint. I don't know if I can buy a G3/500 *now* ; and Alpha have had consistently a higher clock than other CPUs.

Another source of SPEC numbers is http://infopad.eecs.berkeley.ed u/CIC/summary/local/ ; if you want to measure the "advocacy effect", have also a look to http://macspeedzone.com/4.0/WinvsM acSPECint.html, which amusingly have lower SPECint numbers for PII than the best published and verified at spec.org (dig spec.org, maybe they took the very first results for PII, or the worse ; there are several different entries with PIIs).

Perhaps you've heard of Project Appleseed? No?
http://exodus.physics.ucla.edu/appleseed/appleseed .html
Imagine that--an independent study of G3 vs. x86 and other processors...hmm...a PII/300 is over 33% slower than a G3/266 (13% slower clock), and over 17% slower than a Rev. A iMac (29% slower clock).

The funny thing is that the cluster of 8 G3/266 would have the same performance of a single 21264/667Mhz for floating point (or 2 21264 for integer code). The 21264 may be more expensive than 8 G3, but it is sure much more easy to debug a mono-processor code :-).
You should have noted the fine prints:
This investigation was initially motivated by the impressive single node performance we achieved on our well-benchmarked suite of plasma particle-in-cell (PIC) simulation codes [4-5] on the Macintosh G3/266, as shown in Table I. This was due in part to the availability of an excellent optimizing Fortran compiler for the Macintosh produced by the Absoft Corporation [6]
and of course:
Computer Push Time Loop Time
Macintosh G3/300: 1750 nsec. 191.1 sec.

One iteration runs in 1.75 microsecond (i.e. 500 cycles), which means that it is really short code and loop intensive (about 100 millions loops), very likely to be optimiseable. This is a specific application, that couldn't be generalized. The annoying thing is that, contrary to SPEC which allows any submitter to choose its own compiler, we don't know how they decided to choose the others' compilers. They are also daily Mac users, they probably know very well how to optimize, and choose the compiler for their Mac.
But all in all, it is a good point for G3. Now if only Apple/Motorola released official SPEC numbers for G3, instead of plain bogus bytemarks...

In the beginning there was IBM.

Well who did you expect? Transmeta? Maybe?

Anyhow, back in the mid seventies, their 801 project, which was not really a microprocessor for a while longer, really helped get RISC started. Seymour Cray was involved in an earlier proto-RISC project, but then he was a blue ribbon genius.

Well, IBM did use some of their RISC developments, but it wasn't until the early nineties that they created the POWER1 chip (which is not really all that RISCy, but that's neither here nor there - it did evolve from the 801 project, is all)

In '92, Apple, IBM and Motorola (collectively known as AIM, or the Smith Street Gang ;) got together and began developing the PowerPC design, which was based off of POWER and was intended to replace the Motorola 68000 series (which it basically did) and the Intel x86 series, which is still crawling along, much to everyone's surprise.

The first PowerPC chip was the 601, and it can be considered a Generation 1 chip in this family (G1).

The 603 chip, intended primarily for laptops, but also used in low-end systems due to it being quite cheap, was more or less a G2 chip. The version that Apple used more frequently was the 603e, which had a larger cache - critical for laptops, where a L2 cache would prove detrimental.

The big name G2 chip was the 604. Rather than being a heavily POWER influenced design it was the first 'real' PowerPC chip. Its big brother is the rarely seen 620, which expanded the design to 64 bits, but was only really used for the IBM RS/6000 machines, and maybe some others. It was kind of slow, and arrived late, and is quite large and hot, IIRC.

There was also a rumored 615 chip which would have an x86 core as well as a 604. This never materialized however, but it's a neat, if useless (probably too expensive and unpredictable) idea.

The G3 that Apple touts is really the 750 chip. I suspect that they call it the G3, because it sounds better when compared to the PII. ;)

Well, to be quite frank, although the 750 is damned fast, and exceptionally cheap (it destroyed the market for used macs) it is actually pretty weak in the FPU department as compared to the 604, and is more of a successor to the 603. It's a cheap laptop chip that's popular in desktops due to low cost and reasonable performance.

Copper wiring is also now being used in the 750's (my G3 Blue and White has a copper chip) which not only boosts clock speeds by approx 33%, but also is an excellent folk debugging remedy. ;)

A little later this year, hopefully by July (the NY Macworld show) we'll see the first Macs with the G4 processor which will have several innovations.

First, it'll have additional instructions (collectively known as Altivec) which are somewhat like the MMX extensions. They're supposed to speed up a number of 128 bit 'multimedia' operations by operating in parallel with the int and fpu. While they have been reported to speed a lot of stuff up a great deal, I think that Motorola's management heard about the MMX announcement some years back and told their designers to one-up them. It would explain why they're so late.

Also, although we're unlikely to see this on anything that actual people can afford, AIM is also going to be making multiprocessor G4 chips. That is, multiple processors on the same piece of silicon. That should be hella fast, but i doubt they'll be used in much outside of servers and the worlds' most wicked pissah photoshop box (1 GB RAM, natch)

For more information on Grank Funk... er, PowerPCs, check out http://infopad.eecs.b erkeley.edu/CIC/archive/cpu_history.html, http://www.mot.com/SPS/PowerPC/ and http://www.chips.ibm.com/products/ppc/

And it looks so nifty!!

Mike
--

As a counterpoint to all the libertarian cheerleading around here, let me offer the following excerpt from an essay by Brad DeLong, an economics professor at Berkeley. (Click here for the complete essay and click here for DeLong's home page.)

This was written in response to Ira Magaziner's recommendations for government regulation, or lack thereof, on electronic commerce. In the introduction, which I snipped, DeLong gave Michael Froomkin, Hal Varian, and Paul Romer credit for most of the ideas in the essay.

As I read over the Magaziner report, and think about how what it says and leaves unsaid interacts with the other pressures on government policy, I find myself more worried about the future than most of the speakers at the conference. Look at the principles of the Magaziner report: "the private sector should lead," "avoid undue government restrictions," "government should provide a predictable, minimalist, consistent, and simple legal environment," "recognize unique qualities," and "facilitate global electronic commerce." Look at how they are applied: No internet taxes, but also no pools of government money to help provide the public-goods commons for our global electronic village. An information superhighway, as the Vice President used to say, but one without federally-funding. A heavy push to embrace and extend private intellectual property rights. A push to end, worldwide governments' ability to require compulsory licensing as a matter of course. Extension of the property rights of current trademark holders, at least for those with deep pockets. Privacy principles which seem to be honored in the breach because the private sector has not yet led.

It seems to me that Ira Magaziner and his political masters have a view that government is the surveyor of the electronic frontier. The government's job is to draw the property lines--the north boundary of parcel 24 runs from the cottonwood tree to the waterhole--set up rules for selling off the plots, make sure that the railroads get their share of the land, and provide a judge to rule on disputes and a sheriff to enforce the judge's orders.

Now when you are settling a real frontier, this kind of "letting the private sector lead" works pretty well. We may not like what happens to the Indians, or what happens if the judge decides that no witness born in Mexico is credible, or how much land the railroads get, or what happens when the cattle baron has his hired hands homestead all the waterholes in the county. But in the main letting the private sector lead works very well. The Invisible Hand of the marketplace does a good job at guiding people to reach productive and fruitful decisions as to how to use resources as they settle the frontier.

But I suspect that the information economy is going to be different. I may be wrong, but I think it is going to be different enough that the market economy is going to work much less well than we are used to. I suspect that going down the road marked by the Magaziner report is going to leave us suffering from carpal tunnel syndrome of the Invisible Hand.

For one example, consider the push to embrace and extend intellectual property rights. The idea is that by making more information appropriable, we are making incentives better. After all, who is going to finance work if you cannot make money off of it? But when I look at current stock market valuations, I find it hard to believe that many internet enterprises today cannot find financing because investors fear that they will not be able to profit from the consumer value they create. And the dangers of providing broad rights to intellectual property are great.

You see, information goods are what economist Paul Romer calls non-rival. You can sell it more than once. Just because one of your customers is "using" a piece of information doesn't mean that another--or many others--cannot be. This non-rivalry gives the largest producer the potential of unlimited economies of scale. It means that, as Carl Shapiro and Hal Varian write in their Information Rules book, information goods markets will not, cannot look like the competitive markets in which the Invisible Hand works well.

So do we break up every very successful company once a decade? Do we learn to live with natural monopoly and be happy about it? Bear in mind that this time the economies of scale or so large that it is monopoly, and not the early twentieth-century oligopolies that we face. I suspect that in many cases in the future we will find that in market after market the most powerful competitor of the dominant firm is its own installed base, the products that it sold to end users as it was becoming dominant. It seems to me that some leakage or slippage in control over intellectual property may well be desirable.

As a counterpoint to all the libertarian cheerleading around here, let me offer the following excerpt from an essay by Brad DeLong, an economics professor at Berkeley. (Click here for the complete essay and click here for DeLong's home page.)

This was written in response to Ira Magaziner's recommendations for government regulation, or lack thereof, on electronic commerce. In the introduction, which I snipped, DeLong gave Michael Froomkin, Hal Varian, and Paul Romer credit for most of the ideas in the essay.

As I read over the Magaziner report, and think about how what it says and leaves unsaid interacts with the other pressures on government policy, I find myself more worried about the future than most of the speakers at the conference. Look at the principles of the Magaziner report: "the private sector should lead," "avoid undue government restrictions," "government should provide a predictable, minimalist, consistent, and simple legal environment," "recognize unique qualities," and "facilitate global electronic commerce." Look at how they are applied: No internet taxes, but also no pools of government money to help provide the public-goods commons for our global electronic village. An information superhighway, as the Vice President used to say, but one without federally-funding. A heavy push to embrace and extend private intellectual property rights. A push to end, worldwide governments' ability to require compulsory licensing as a matter of course. Extension of the property rights of current trademark holders, at least for those with deep pockets. Privacy principles which seem to be honored in the breach because the private sector has not yet led.

It seems to me that Ira Magaziner and his political masters have a view that government is the surveyor of the electronic frontier. The government's job is to draw the property lines--the north boundary of parcel 24 runs from the cottonwood tree to the waterhole--set up rules for selling off the plots, make sure that the railroads get their share of the land, and provide a judge to rule on disputes and a sheriff to enforce the judge's orders.

Now when you are settling a real frontier, this kind of "letting the private sector lead" works pretty well. We may not like what happens to the Indians, or what happens if the judge decides that no witness born in Mexico is credible, or how much land the railroads get, or what happens when the cattle baron has his hired hands homestead all the waterholes in the county. But in the main letting the private sector lead works very well. The Invisible Hand of the marketplace does a good job at guiding people to reach productive and fruitful decisions as to how to use resources as they settle the frontier.

But I suspect that the information economy is going to be different. I may be wrong, but I think it is going to be different enough that the market economy is going to work much less well than we are used to. I suspect that going down the road marked by the Magaziner report is going to leave us suffering from carpal tunnel syndrome of the Invisible Hand.

For one example, consider the push to embrace and extend intellectual property rights. The idea is that by making more information appropriable, we are making incentives better. After all, who is going to finance work if you cannot make money off of it? But when I look at current stock market valuations, I find it hard to believe that many internet enterprises today cannot find financing because investors fear that they will not be able to profit from the consumer value they create. And the dangers of providing broad rights to intellectual property are great.

You see, information goods are what economist Paul Romer calls non-rival. You can sell it more than once. Just because one of your customers is "using" a piece of information doesn't mean that another--or many others--cannot be. This non-rivalry gives the largest producer the potential of unlimited economies of scale. It means that, as Carl Shapiro and Hal Varian write in their Information Rules book, information goods markets will not, cannot look like the competitive markets in which the Invisible Hand works well.

So do we break up every very successful company once a decade? Do we learn to live with natural monopoly and be happy about it? Bear in mind that this time the economies of scale or so large that it is monopoly, and not the early twentieth-century oligopolies that we face. I suspect that in many cases in the future we will find that in market after market the most powerful competitor of the dominant firm is its own installed base, the products that it sold to end users as it was becoming dominant. It seems to me that some leakage or slippage in control over intellectual property may well be desirable.

• Collaborative Filtering Workshop (Berkeley, 1996)
• Adding Comments to the Web (read it all, including the FAQ )
• W3C's Collaboration/Annotation page (seems dormant).

From CPU info center

450 Mhz Pentium II 17.2 - 12.9
450 Mhz Xeon 18.9 - 14.7
400 Mhz PowerPC G3 17.6 - 12.2
200 Mhz IBM POWER3 13.2 - 30.1
450 Mhz UltraSparcII 19.6 - 27.1
600 Mhz UltraSparcIII 35+ - 60+ (est)
575 Mhz Alpha 21264 30.3 - 47.7
667 Mhz Alpha 21264 44 - 66
1000 Mhz Alpha 21364 ~70 - ~120 (est)
250 Mhz MIPS R10000 14.7 - 24.5

Featuring: The Performance Database Server. Start with the bottom of the Dhrystone results. Then find the older Whetstone database.

STREAM graph of Memory speed vs MFLOPS. (The STREAM standard results has speed numbers. The CPU Info Center has assorted historical CPU info. Here is Intel's Moore's Law graph. Here is a computer timeline. Here are the top 500 supercomputers since 1983.

• Seven parallel specialized execution units
• Multiple sophistocated x86-to-RISC86 instruction decoders
• Advanced two-level branch prediction
• Speculative execution
• Out-of-order execution
• Register renaming and data forwarding
• Issues up to six RISC86 instructions per clock

• 32-Kbyte instruction cache with additional predecode cache
• 32-Kbyte writeback dual-ported data cache
• MESI protocol support

• Seven parallel specialized execution units
• Multiple sophistocated x86-to-RISC86 instruction decoders
• Advanced two-level branch prediction
• Speculative execution
• Out-of-order execution
• Register renaming and data forwarding
• Issues up to six RISC86 instructions per clock

• 32-Kbyte instruction cache with additional predecode cache
• 32-Kbyte writeback dual-ported data cache
• MESI protocol support

Andrea di Sessa and colleagues at Berkeley Grad School of Education developed a system called Boxer which might be of interest to you. I don't know that much development has been going on recently, but you might check with them further. For more info see the Boxer Project website and their FTP server.

Andrea di Sessa and colleagues at Berkeley Grad School of Education developed a system called Boxer which might be of interest to you. I don't know that much development has been going on recently, but you might check with them further. For more info see the Boxer Project website and their FTP server.

For example, in 1906, the average car cost--in 1993 dollars--about $53,000; by 1918, the inflation-adjusted cost was less than half and the quality was about twice as good. Some intellectuals of that era saw the Ford Motor Company as the model for a better America. (See this essay by DeLong for more details.)

Alas, this trend has not continued for the past 80 years. Perhaps computer technology will maintain its growth rate for as long as Kurzweil etal. predict, but I wouldn't bet my 401(k) on it.

For example, in 1906, the average car cost--in 1993 dollars--about $53,000; by 1918, the inflation-adjusted cost was less than half and the quality was about twice as good. Some intellectuals of that era saw the Ford Motor Company as the model for a better America. (See this essay by DeLong for more details.)

Alas, this trend has not continued for the past 80 years. Perhaps computer technology will maintain its growth rate for as long as Kurzweil etal. predict, but I wouldn't bet my 401(k) on it.