I totally agree with your comment that a segment of the high-end market *demands* it. They have demanded it in the past and Microsoft has provided it on a one-on-one ad-hoc basis. That's nothing new. But by packaging up a formal program for anyone with 1500+ licences, Microsoft is beginning to move to making source available to very large companies who *want* it, not just *demand* it as a precondition. And my point was that Microsoft has assessed that demand as being as large as 3% of its customers. Obviously this program still doesn't come close to reaching that large a group of people.
If one assumes that Microsoft wants to give out as little of their source as possible (maximizing their remaining control,) yet at the same time they want to minimize incursions of open source code into their base, then a set of incremental increases in source availability is the least-risk way for them to put their toe in the waters and see how much half-open-source-solutions satisfy people. This small incremental increase in source may not mean much, but to me, it does signal that they are feeling some pain from not making their source available to a their top licensing customers (1500+ licenses is a lot, but that covers a lot more companies than the 25000+ license type of companies I've heard of them giving source to before.)
A couple years ago when MS was first looking at Open Source, Steve Ballmer mentioned that Microsoft had done a study suggesting that something like 1-3% of Windows developers wanted source code access. (Back then you had to pretty much be an OEM, Wall St firm, or Fortune 50 client IIRC to get it.) I was intrigued by this at the time, since A) Microsoft had attempted to get hard data on the demand for this and B) that's a lot of developers. Obviously MS is finally responding to that demand, albeit in a go-slow manner.
I've written a couple a-life programs and have the following observations: one can definitely program systems that demonstrate some behavior that was not explicitly programmed into them. For a coder, this is a cool thing, especially when it works and especially when it surprises you with what it does. But emergent behavior based on programmed-in attributes rather than procedural instructions does not "life" make.
As far as I could tell, the a-life crowd has (to date) failed. All the things we agree on in the real world as being "life" are enormously more complex than these digital abstractions. The definitional issues are hard to agree on ('what is life? what is intelligence? what is consciousness?') And an honest practitioner recognizes the huge amount of work required to set up "just the right initial conditions and assumptions for something 'interesting' to happen". So one way for them to succeed is to dumb down their definition of life. IMHO, the field is in significant danger of doing this. When you hear people proclaiming that they are gods (sounds ridiculous, i know, but the word was even applied in this slashdot review) and have created digital life, take it with a pound of salt as the hubris and grant-inducing-hook that it really is. IMHO.
This particular approach is bogus, since as others have pointed out, the DMCA is not about methods of encryption but about methods of copy control.
This does raise an interesting question. How can we consumers use methods of copy control to prevent excessive and DMCA-illegal snooping by the new corporate thought police?
Much has been said about consumers losing in the past and present congressional debates due to lack of organization, entertainment industry lobbying spending, etc. But I think there's another "follow-the-money" trail we should be prepared to understand and figure out how to route around:
The U.S. runs almost a trillion dollar trade deficit every year. Money is flowing out of the U.S. fast. For now, who cares? But longer-term it's a problem. To reduce this trade deficit, the U.S. needs to increase its exports. There are really only 3-4 big industries where the U.S. is competitive enough that they export much more than they import. One is the software (and to a lesser degree hardware) industry. A second is agriculture. A third is aerospace/military equipment. And the fourth big one is the entertainment industry. At the end of the day, the U.S. Congress is going to do what most seems to help American economic competitiveness, and at the moment, that means giving the entertainment industry whatever protection they need.
Is the DMCA really a hidden trade barrier then? Good question...
--LP
Disclaimer: I am not an economist. The argument above lacks obvious caveats in an attempt to retain simplicity and clarity. Harder statistics on exports by industry might confirm or refute the above analysis.
Exactly what is included in the courses and the degree varies from school to school. There is no academia-wide consensus that I know of with really well-defined boundaries.
My advice would be the following: You don't need to decide your major now when you haven't even graduated high school; wait till you get to campus and see and talk to fellow students about the specific courses involved. Pick the major that has the courses you find most stimulating (or has the fewest courses you find annoying/irrelevant).
Now in general, CS guys will never be hired for hardware/EE jobs, while the reverse will happen. This implies that an EE degree might be better (more flexible), but if you're going to work in software at the end of the day, you might as well get the extra software and algorithm experience from the CS degree. As a counter-example, I once interviewed someone with an E.E. for an internet startup I was working at and after looking at the mostly hardware courses he listed on his resume I found myself wondering how much coding experience he really had (we didn't hire him). So an E.E. degree without much software can be a minus.
Personally, I started out as and E.E. and switched to C.S. after taking and hating the first of four analog electronics courses (RLC circuits anyone?) I would have needed to get the E.E. degree. No regrets.
--LP
Re:One GUI from the "Movie OS"...
on
MUD Shell
·
· Score: 1
That 3D file manager was an actual shipping demo tool that came on all SGI IRIX systems. I don't remember the precise name of it... Nobody I knew really used it of course, but it made for a funny 'in' joke in the movie...
You've got Mr. Schneier's high-level message but you seem to be misquoting him in a way that ignores a very fundamental distinction.
"Acording to Bruce Schneier it is impossible to prove the unbreakability of a cryptographic algorithm. "
Find me that quote. It *is* possible to prove the breakability or unbreakability of an algorithm, as Bruce well knows but your quote of him denies. Proving the unbreakability of a product, of an *implementation* of that algorithm is practically impossible as Mr. Schneier has repeatedly said. (Although one could claim that NCSA/NSA-rated A1 products constitute a potential counter-example for highly-limited problem domains.)
I'm not claiming that you're a phony. But I sure as hell wouldn't trust your quote from Mr. Schneier just because you said it on Slashdot and it got a 5 rating.
random number stream must be so large and sent so fast that a third party must not have the resources to store it all
yet both sender and recipient must have bandwidths large enough to read such a stream in realtime
This is not a technology for 56k modems or even current DSL bandwidths.
Perhaps a series of distributed random number generator servers around the net a la napster/gnutella could make this feasible, but even that seems a bit of a stretch.
One solution is just to start calling open source code approaches "freedom software" when this type of argument is advanced. I always wondered why Stallman picked "free" over "freedom" as an adjective, since "freedom" lacks the confusion and properly conveys the right connotations.
"Freedom software" nicely outlines the one advantage Microsoft will never give you (freedom (from lock-in)), and what could be more pro-American than "freedom software"?
If this FUD-slewing argument becomes a serious (rather than bogus) threat, I think we should just start calling free/open source software "freedom software" instead. That would put the burden back on Mr. Allchin. After all, what could be more American than "freedom software"?
I can see it now: "freedom is an intellectual property destroyer!"
While we're offtopic and proposing alternatives to "free software" as a term... I've always thought "freedom software" would have the closest match of clarity and connotations.
Freedom is the fundamental secular American virtue. It's very hard in American culture to argue that people or products should have less freedom. What is more American than more freedom in your software?
(Whereas in a capitalistic society, connotations around 'free' are more mixed. 'Free' can imply cheap/low-quality, no-free-lunch, little-profit-opporunity, etc. Freedom on the other hand is worth any price...;)
As far as the term "Open" goes, it's less awkward, verbally. Still, I think "Freedom Software" has harder-to-argue-against connotations, culturally. "Open" is kind of a vague word that needs plenty of explaining.
I believe the crime Napster is being accused of committing is called "contributory copyright infringement". Sort of like being an accessory to a murder, but worse.
I find it interesting that the earliest NVIDIA engineers were from Sun. To me, this explains A) their pre-occupation with surfaces as a primitive, long a tendency within Sun due to their MCAD focus, and B) their appreciation for the value of short cycle times. 3dfx, by contrast, was founded in part by engineering from SGI (e.g. Gary Tarolli) who had a stronger appreciation of how to optimize great 3D technology and astutely balance technological tradeoffs, but perhaps less cultural appreciation for obscure business issues like cycle times and ability to execute.
Just as in the workstation market, Sun was able to muscle-out SGI by steadily raising the bar and focusing on good-enough-for-my-core-markets improvements, so NVIDIA beat 3dfx. Although one could always argue that 3dfx beat itself. Still, in my mind, the turning point was when NVIDIA hired a lot of key engineers fleeing SGI which was in the process of spiralling downward. If 3dfx had gotten that team, we might be looking at a different story.
In any event, it's remarkable that engineers from either Sun or SGI were able to shift their mindset from product cycles of 3 years (workstation graphics) to product cycles of 6 months (PC 2D graphics a la Cirrus Logic). If you've grown up building projects the slow way, a faster tempo is not easy to adjust to! I've always thought that besides his brilliance, the psychological development of John Carmack's early exposure to 1 month product cycles helped him learn the focus needed to keep his current 18-month(?) product cycles going, something his competitors have really struggled to match.
It'll be interesting to see if the lack of discipline and business focus so evident within SGI engineering begins reasserting itself at NVIDIA as pressures on them ease and the early SGI engineers look at the comfortable value of their options which should be fairly vested by now. As a consumer, all I can say is that thank goodness ATI is around to keep the pressure on them.:)
People matter. There's an interesting book about how the 3D hardware market was won and lost just waiting to be written.
3D deposition printers have been around for a decade. The key question in my mind is: what is the rate of improvement in this technology? The printers still seem to be very slow and very expensive. 3D printers are not semiconductor-based, they do not necessarily follow Moore's Law (which works out to ~60% "improvement" per year). So at what rate (in either cost or speed of printing) are these 'fabbers' improving? There should be enough historical data out there for someone knowledgeable to take a stab at explaining that. That'd seems to be a rather key question...
I wonder how much of the article was press release fodder. The claim that transporting objects typically takes 21 days is just BS. That probably includes red tape and loading, which the 3 day figure probably doesn't. My one experience in transatlantic travel took me 6 days on the fairly hefty QE2-- looking it up, it weighs 70,000 tons. Transporting 10,000 tons in 3 days is definitely an advance but I don't quite see anything revolutionary about it.
--LP
earliest Video to ASCII tool?
on
The ASCII Cam
·
· Score: 2
I first saw Video to ASCII back in 1995 done by a free tool called ttyvideo, described and downloadable here written by Chris Pirazzi at SGI.
Nice to see a full open source release though rather than just an IRIX binary.
Why is the NSA publishing this? You know, reading the first nine tenths of that piece (and the slashdot one-line summary), one might think that the U.S. Intelligence community (the NSA wasn't around, right?) pulled off quite a coup, developing revolutionary, ahead-of-its-time-digital-telephony-technology. But the appendix buried at the end comes to quite a different conclusion based on the testimony of someone responsible for the project at the time:
"All of the elements of this system were developed by the Bell Telephone Laboratories in the interests of advancing the art of telephony. . . . When it had reached the point where its principles could be demonstrated, the Signal Corps took prompt steps for its procurement."
It appears to me that the NSA has, to a limited degree, tried to make their mammoth budgets and spare-no-expense approach more palatable to the American taxpayer by claiming to develop technology way ahead of its time. "Hey, our job is American security but even if you think that's not a big problem these days, look at the tech spinoffs and innovation we create." To me this seems a little disingenuous.
As this article, among others indicates, the NSA et al may be a nice mechanism *funding* innovation (although VCs are arguably better, with a more attractive upside economically) they don't actually *do* much innovation (outside of say, crypto research).
At some point, I'd be curious what made you pick the $40 price target.
Your assumption that CD-RW drives are expensive is a false one. You can get internal CD-RW drives for as little $94 (+S&H) nowdays, which seems relatively close to your target price. (As with most components, buying the rock-bottom cheapest isn't necessarily a good idea, but hey, if that's what you can afford, that's what you get.) And media costs have dropped dramatically too, with CD-RW disks being basically $1 a piece for 650 MB capacity. For low-end prices, check out Pricewatch (no affiliation).
There is another slightly older alternative, the ZIP drive. OEM internal drives are as little as $34 +S&H at pricewatch, with media costs running ~$5-10. There are three significant problems with this approach: 1) reliability, 2) single source issues, and 3) obsolescence. In my experience, ZIP drives are not particularly reliable. There's a fairly well-known phenomena called the "Click of Death" (do a google.com search to find out more) that plagued drives during one period (my sister's ZIP drive had this) and there was a huge class action lawsuit against the ZIP maker Iomega. Second, the ZIP drive standard is essentially owned by one company, Iomega, so your ability to switch to alternatives is limited if you run into problems or if Iomega jacks up prices and gives up competing on the merits to optimize their profitability (as they should). With CD-RW you have a variety of drive manufacturers competing voraciously for marketshare and prices will continue to drop substantially. And third, ZIP is a standard on its way out. People used it when CD-RW drives were $300+, but with CD-RW drives now under $100, the alternative fits a much broader set of consumer needs. ZIP media has smaller capacity and is less versatile: you can't just take it to any student or faculty or employer's PC unless they too buy a ZIP drive. Every computer is built with at least a CD-ROM reading device... the power of network effects is all on the side of CD-R(W).
There are two basic uses of removable media: 1) moving files between PCs and 2) backing up your PC. For a drive standard to be widely adopted you have to meet both of those reasonably well. Backing up a 10 GB drive with a 100 MB ZIP is obviously a return to the problems of swapping floppies and is one reason CD-RW is picking up steam over ZIP. The other is the rising interest of people in 3) making audio CDs, something that CD-RW has made very popular with the teenage and college crowd as well as the mainstream public. Wannabe successors to the CD-RW drive (cough, DVD, cough) ignore consumers' interest in doing so at their peril.
Buying ZIP and trying to get 20,000 students to go along with your choice would be penny wise and pound foolish. You'd end up having to support the ZIP standard for the next 15 years when its already on its way out and has about 5 more years of life left. (Insert wild hand waving gestures here...;) We may never have something as completely ubiquitous as the floppy was. But with steadily dropping prices, the CD-RW drive is coming increasingly close. ZIP won't be the next floppy. CD-RW will.
The full Washington Post article on Gore's grades can be found <a href="http://newsweek.washingtonpost.com/nw-srv/w<nobr>e<wbr></wbr></nobr> b/special/campaign2000/sr03_032700.htm"& gt; here</a>.
Gore's intelligence advantage over Bush is overrated. As evidence, I recommend to you this fascinatingly counter-intuitive article to you that appealed to the under-performing student in me.
Gore actually got worse grades in college than Bush ever did! The Washington Post did an front-page article comparing the two on March 19, 2000. The Post charges $2.50 to dig stuff out of their archives but a little searching turned up a copy here.
I was referring to DEC graphics-card engineers, not the microprocessor guys. Also the timeframe of the discussion was much, much earlier than the EV7 design timeframe.
Why the hell is it so hard to make non-buggy products with Rambus? I first heard about Rambus when evaluating some new SGI graphics adapters (High Impact and Maximum Impact) that were among the first devices to use Rambus DRAM. Guess what? The products were announced in mid-1995 but ran into various bugs and supply problems which led them not to really ship till beginning-1996. When those supply problems became apparent, their stock hit an all-time high (~45) never to return and it has since dropped to around ~4.
I talked to a competitor (DEC)'s engineers around that time and they said that while they'd looked at Rambus, it was not a very stable memory technology; the complexities it introduced into their engineering were not worth the performance gain and cost hit.
The fact that Intel has had the same problems as SGI, albeit on a much, much larger scale, really leads me to wonder... what is it that makes Rambus memory controllers or interfacing chipsets so damn difficult to get working properly?
If someone could answer this question, I'd be really obliged. The "toll-keeper" problems were obvious from day 1 with Rambus... the techical problems were not.
I read through Karl's policy statement on his website and a couple things made me wonder if he understands the full set of tradeoffs involved.
He did a fine job of explaining why the technical stability of the name system is not a significant problem that should inhibit change, but I kept wondering if by "stability", the ICANN traditionalists were concerned with non-technical stability issues that were at stake, such as stability of brand name and trademark recognition online. Isn't that the heart of the issue? Is anyone really arguing about name system uptime?
I think the ability of big companies to squash any use of any particular brandname or trademark they come up with should be limited, but I don't think that companies and their trademarks, whether "McDonalds" or "Slashdot," should be without protection from looter/moocher types who attempt to register recognized names under other global TLDs, whether ICANN adds 20 more or 20,000 more. I know that aspects of this problem are dealt with in current ICANN policies, albeit not fully satisfactorily, but in my limited examination of Karl's writings, I haven't seen Karl articulate that he either recognizes this issue or has a balanced solution in mind. Corrections or pointers welcome.
--LP
(This is not an attack on Karl. This is an attempt to point out an issue of concern and ask for informational pointers or responses.)
Slashdot Mirror
On a side note, they have quite a few "Celebrity's Bookmarks" lists there and while many are trite or stupid, a few stars have interesting lists.
--LP
I totally agree with your comment that a segment of the high-end market *demands* it. They have demanded it in the past and Microsoft has provided it on a one-on-one ad-hoc basis. That's nothing new. But by packaging up a formal program for anyone with 1500+ licences, Microsoft is beginning to move to making source available to very large companies who *want* it, not just *demand* it as a precondition. And my point was that Microsoft has assessed that demand as being as large as 3% of its customers. Obviously this program still doesn't come close to reaching that large a group of people.
If one assumes that Microsoft wants to give out as little of their source as possible (maximizing their remaining control,) yet at the same time they want to minimize incursions of open source code into their base, then a set of incremental increases in source availability is the least-risk way for them to put their toe in the waters and see how much half-open-source-solutions satisfy people. This small incremental increase in source may not mean much, but to me, it does signal that they are feeling some pain from not making their source available to a their top licensing customers (1500+ licenses is a lot, but that covers a lot more companies than the 25000+ license type of companies I've heard of them giving source to before.)
--LP
A couple years ago when MS was first looking at Open Source, Steve Ballmer mentioned that Microsoft had done a study suggesting that something like 1-3% of Windows developers wanted source code access. (Back then you had to pretty much be an OEM, Wall St firm, or Fortune 50 client IIRC to get it.) I was intrigued by this at the time, since A) Microsoft had attempted to get hard data on the demand for this and B) that's a lot of developers. Obviously MS is finally responding to that demand, albeit in a go-slow manner.
--LP
I've written a couple a-life programs and have the following observations: one can definitely program systems that demonstrate some behavior that was not explicitly programmed into them. For a coder, this is a cool thing, especially when it works and especially when it surprises you with what it does. But emergent behavior based on programmed-in attributes rather than procedural instructions does not "life" make.
As far as I could tell, the a-life crowd has (to date) failed. All the things we agree on in the real world as being "life" are enormously more complex than these digital abstractions. The definitional issues are hard to agree on ('what is life? what is intelligence? what is consciousness?') And an honest practitioner recognizes the huge amount of work required to set up "just the right initial conditions and assumptions for something 'interesting' to happen". So one way for them to succeed is to dumb down their definition of life. IMHO, the field is in significant danger of doing this. When you hear people proclaiming that they are gods (sounds ridiculous, i know, but the word was even applied in this slashdot review) and have created digital life, take it with a pound of salt as the hubris and grant-inducing-hook that it really is. IMHO.
--LP
This particular approach is bogus, since as others have pointed out, the DMCA is not about methods of encryption but about methods of copy control.
This does raise an interesting question. How can we consumers use methods of copy control to prevent excessive and DMCA-illegal snooping by the new corporate thought police?
--LP
Much has been said about consumers losing in the past and present congressional debates due to lack of organization, entertainment industry lobbying spending, etc. But I think there's another "follow-the-money" trail we should be prepared to understand and figure out how to route around:
The U.S. runs almost a trillion dollar trade deficit every year. Money is flowing out of the U.S. fast. For now, who cares? But longer-term it's a problem. To reduce this trade deficit, the U.S. needs to increase its exports. There are really only 3-4 big industries where the U.S. is competitive enough that they export much more than they import. One is the software (and to a lesser degree hardware) industry. A second is agriculture. A third is aerospace/military equipment. And the fourth big one is the entertainment industry. At the end of the day, the U.S. Congress is going to do what most seems to help American economic competitiveness, and at the moment, that means giving the entertainment industry whatever protection they need.
Is the DMCA really a hidden trade barrier then? Good question...
--LP
Disclaimer: I am not an economist. The argument above lacks obvious caveats in an attempt to retain simplicity and clarity. Harder statistics on exports by industry might confirm or refute the above analysis.
My advice would be the following: You don't need to decide your major now when you haven't even graduated high school; wait till you get to campus and see and talk to fellow students about the specific courses involved. Pick the major that has the courses you find most stimulating (or has the fewest courses you find annoying/irrelevant).
Now in general, CS guys will never be hired for hardware/EE jobs, while the reverse will happen. This implies that an EE degree might be better (more flexible), but if you're going to work in software at the end of the day, you might as well get the extra software and algorithm experience from the CS degree. As a counter-example, I once interviewed someone with an E.E. for an internet startup I was working at and after looking at the mostly hardware courses he listed on his resume I found myself wondering how much coding experience he really had (we didn't hire him). So an E.E. degree without much software can be a minus.
Personally, I started out as and E.E. and switched to C.S. after taking and hating the first of four analog electronics courses (RLC circuits anyone?) I would have needed to get the E.E. degree. No regrets.
--LP
--LP
You've got Mr. Schneier's high-level message but you seem to be misquoting him in a way that ignores a very fundamental distinction.
"Acording to Bruce Schneier it is impossible to prove the unbreakability of a cryptographic algorithm. "
Find me that quote. It *is* possible to prove the breakability or unbreakability of an algorithm, as Bruce well knows but your quote of him denies. Proving the unbreakability of a product, of an *implementation* of that algorithm is practically impossible as Mr. Schneier has repeatedly said. (Although one could claim that NCSA/NSA-rated A1 products constitute a potential counter-example for highly-limited problem domains.)
I'm not claiming that you're a phony. But I sure as hell wouldn't trust your quote from Mr. Schneier just because you said it on Slashdot and it got a 5 rating.
--LP
- random number stream must be so large and sent so fast that a third party must not have the resources to store it all
- yet both sender and recipient must have bandwidths large enough to read such a stream in realtime
This is not a technology for 56k modems or even current DSL bandwidths.Perhaps a series of distributed random number generator servers around the net a la napster/gnutella could make this feasible, but even that seems a bit of a stretch.
--LP
One solution is just to start calling open source code approaches "freedom software" when this type of argument is advanced. I always wondered why Stallman picked "free" over "freedom" as an adjective, since "freedom" lacks the confusion and properly conveys the right connotations.
"Freedom software" nicely outlines the one advantage Microsoft will never give you (freedom (from lock-in)), and what could be more pro-American than "freedom software"?
--LP
If this FUD-slewing argument becomes a serious (rather than bogus) threat, I think we should just start calling free/open source software "freedom software" instead. That would put the burden back on Mr. Allchin. After all, what could be more American than "freedom software"?
I can see it now: "freedom is an intellectual property destroyer!"
Cynically yours,
--LP
While we're offtopic and proposing alternatives to "free software" as a term... I've always thought "freedom software" would have the closest match of clarity and connotations.
;)
Freedom is the fundamental secular American virtue. It's very hard in American culture to argue that people or products should have less freedom. What is more American than more freedom in your software?
(Whereas in a capitalistic society, connotations around 'free' are more mixed. 'Free' can imply cheap/low-quality, no-free-lunch, little-profit-opporunity, etc. Freedom on the other hand is worth any price...
As far as the term "Open" goes, it's less awkward, verbally. Still, I think "Freedom Software" has harder-to-argue-against connotations, culturally. "Open" is kind of a vague word that needs plenty of explaining.
--LP
I believe the crime Napster is being accused of committing is called "contributory copyright infringement". Sort of like being an accessory to a murder, but worse.
--LP
(Oops, I left out the <SARCASM> </SARCASM> tags!)
I find it interesting that the earliest NVIDIA engineers were from Sun. To me, this explains A) their pre-occupation with surfaces as a primitive, long a tendency within Sun due to their MCAD focus, and B) their appreciation for the value of short cycle times. 3dfx, by contrast, was founded in part by engineering from SGI (e.g. Gary Tarolli) who had a stronger appreciation of how to optimize great 3D technology and astutely balance technological tradeoffs, but perhaps less cultural appreciation for obscure business issues like cycle times and ability to execute.
:)
Just as in the workstation market, Sun was able to muscle-out SGI by steadily raising the bar and focusing on good-enough-for-my-core-markets improvements, so NVIDIA beat 3dfx. Although one could always argue that 3dfx beat itself. Still, in my mind, the turning point was when NVIDIA hired a lot of key engineers fleeing SGI which was in the process of spiralling downward. If 3dfx had gotten that team, we might be looking at a different story.
In any event, it's remarkable that engineers from either Sun or SGI were able to shift their mindset from product cycles of 3 years (workstation graphics) to product cycles of 6 months (PC 2D graphics a la Cirrus Logic). If you've grown up building projects the slow way, a faster tempo is not easy to adjust to! I've always thought that besides his brilliance, the psychological development of John Carmack's early exposure to 1 month product cycles helped him learn the focus needed to keep his current 18-month(?) product cycles going, something his competitors have really struggled to match.
It'll be interesting to see if the lack of discipline and business focus so evident within SGI engineering begins reasserting itself at NVIDIA as pressures on them ease and the early SGI engineers look at the comfortable value of their options which should be fairly vested by now. As a consumer, all I can say is that thank goodness ATI is around to keep the pressure on them.
People matter. There's an interesting book about how the 3D hardware market was won and lost just waiting to be written.
--Greg
3D deposition printers have been around for a decade. The key question in my mind is: what is the rate of improvement in this technology? The printers still seem to be very slow and very expensive. 3D printers are not semiconductor-based, they do not necessarily follow Moore's Law (which works out to ~60% "improvement" per year). So at what rate (in either cost or speed of printing) are these 'fabbers' improving? There should be enough historical data out there for someone knowledgeable to take a stab at explaining that. That'd seems to be a rather key question...
--LP
I wonder how much of the article was press release fodder. The claim that transporting objects typically takes 21 days is just BS. That probably includes red tape and loading, which the 3 day figure probably doesn't. My one experience in transatlantic travel took me 6 days on the fairly hefty QE2-- looking it up, it weighs 70,000 tons. Transporting 10,000 tons in 3 days is definitely an advance but I don't quite see anything revolutionary about it.
--LP
Nice to see a full open source release though rather than just an IRIX binary.
--LP
"All of the elements of this system were developed by the Bell Telephone Laboratories in the interests of advancing the art of telephony. . . . When it had reached the point where its principles could be demonstrated, the Signal Corps took prompt steps for its procurement."
It appears to me that the NSA has, to a limited degree, tried to make their mammoth budgets and spare-no-expense approach more palatable to the American taxpayer by claiming to develop technology way ahead of its time. "Hey, our job is American security but even if you think that's not a big problem these days, look at the tech spinoffs and innovation we create." To me this seems a little disingenuous.
As this article, among others indicates, the NSA et al may be a nice mechanism *funding* innovation (although VCs are arguably better, with a more attractive upside economically) they don't actually *do* much innovation (outside of say, crypto research).
--LP
Your assumption that CD-RW drives are expensive is a false one. You can get internal CD-RW drives for as little $94 (+S&H) nowdays, which seems relatively close to your target price. (As with most components, buying the rock-bottom cheapest isn't necessarily a good idea, but hey, if that's what you can afford, that's what you get.) And media costs have dropped dramatically too, with CD-RW disks being basically $1 a piece for 650 MB capacity. For low-end prices, check out Pricewatch (no affiliation).
There is another slightly older alternative, the ZIP drive. OEM internal drives are as little as $34 +S&H at pricewatch, with media costs running ~$5-10. There are three significant problems with this approach: 1) reliability, 2) single source issues, and 3) obsolescence. In my experience, ZIP drives are not particularly reliable. There's a fairly well-known phenomena called the "Click of Death" (do a google.com search to find out more) that plagued drives during one period (my sister's ZIP drive had this) and there was a huge class action lawsuit against the ZIP maker Iomega. Second, the ZIP drive standard is essentially owned by one company, Iomega, so your ability to switch to alternatives is limited if you run into problems or if Iomega jacks up prices and gives up competing on the merits to optimize their profitability (as they should). With CD-RW you have a variety of drive manufacturers competing voraciously for marketshare and prices will continue to drop substantially. And third, ZIP is a standard on its way out. People used it when CD-RW drives were $300+, but with CD-RW drives now under $100, the alternative fits a much broader set of consumer needs. ZIP media has smaller capacity and is less versatile: you can't just take it to any student or faculty or employer's PC unless they too buy a ZIP drive. Every computer is built with at least a CD-ROM reading device... the power of network effects is all on the side of CD-R(W).
There are two basic uses of removable media: 1) moving files between PCs and 2) backing up your PC. For a drive standard to be widely adopted you have to meet both of those reasonably well. Backing up a 10 GB drive with a 100 MB ZIP is obviously a return to the problems of swapping floppies and is one reason CD-RW is picking up steam over ZIP. The other is the rising interest of people in 3) making audio CDs, something that CD-RW has made very popular with the teenage and college crowd as well as the mainstream public. Wannabe successors to the CD-RW drive (cough, DVD, cough) ignore consumers' interest in doing so at their peril.
Buying ZIP and trying to get 20,000 students to go along with your choice would be penny wise and pound foolish. You'd end up having to support the ZIP standard for the next 15 years when its already on its way out and has about 5 more years of life left. (Insert wild hand waving gestures here... ;) We may never have something as completely ubiquitous as the floppy was. But with steadily dropping prices, the CD-RW drive is coming increasingly close. ZIP won't be the next floppy. CD-RW will.
--LP
The full Washington Post article on Gore's grades can be found <a href="http://newsweek.washingtonpost.com/nw-srv/w<nobr>e<wbr></wbr></nobr> b/special/campaign2000/sr03_032700.htm"& gt; here</a>.
Gore actually got worse grades in college than Bush ever did! The Washington Post did an front-page article comparing the two on March 19, 2000. The Post charges $2.50 to dig stuff out of their archives but a little searching turned up a copy here.
--LP
I was referring to DEC graphics-card engineers, not the microprocessor guys. Also the timeframe of the discussion was much, much earlier than the EV7 design timeframe.
Why the hell is it so hard to make non-buggy products with Rambus? I first heard about Rambus when evaluating some new SGI graphics adapters (High Impact and Maximum Impact) that were among the first devices to use Rambus DRAM. Guess what? The products were announced in mid-1995 but ran into various bugs and supply problems which led them not to really ship till beginning-1996. When those supply problems became apparent, their stock hit an all-time high (~45) never to return and it has since dropped to around ~4.
I talked to a competitor (DEC)'s engineers around that time and they said that while they'd looked at Rambus, it was not a very stable memory technology; the complexities it introduced into their engineering were not worth the performance gain and cost hit.
The fact that Intel has had the same problems as SGI, albeit on a much, much larger scale, really leads me to wonder... what is it that makes Rambus memory controllers or interfacing chipsets so damn difficult to get working properly?
If someone could answer this question, I'd be really obliged. The "toll-keeper" problems were obvious from day 1 with Rambus... the techical problems were not.
--LP
I read through Karl's policy statement on his website and a couple things made me wonder if he understands the full set of tradeoffs involved.
He did a fine job of explaining why the technical stability of the name system is not a significant problem that should inhibit change, but I kept wondering if by "stability", the ICANN traditionalists were concerned with non-technical stability issues that were at stake, such as stability of brand name and trademark recognition online. Isn't that the heart of the issue? Is anyone really arguing about name system uptime?
I think the ability of big companies to squash any use of any particular brandname or trademark they come up with should be limited, but I don't think that companies and their trademarks, whether "McDonalds" or "Slashdot," should be without protection from looter/moocher types who attempt to register recognized names under other global TLDs, whether ICANN adds 20 more or 20,000 more. I know that aspects of this problem are dealt with in current ICANN policies, albeit not fully satisfactorily, but in my limited examination of Karl's writings, I haven't seen Karl articulate that he either recognizes this issue or has a balanced solution in mind. Corrections or pointers welcome.
--LP
(This is not an attack on Karl. This is an attempt to point out an issue of concern and ask for informational pointers or responses.)