Yup the article was pretty bogus but let me ask you a question about the rest of your statement.
If we assume that people will pay more for a superior product then why hasn't the superior automobile you are now able to design led to greater profits? And if improving quality doesn't lead to greater profits where is the incentive to improve quality?
Here's two possible answers for you:
1) GM and Chrysler have both posted record profits in recent years.
2) In a near monopolistic market there is little incentive to improve quality, but when you have high quality asian and european competitors a failure to improve quality will result in a LOSS of profits, so you find yourself running faster just to stand still.
What you are talking about is manufacturing capacity, and you're right, overcapacity is bad, what Solow was talking about was profit per employee which should be the final result of productivity gains. That we want as much of as we can get.
By the way, I'd bet dollars to donuts that China's problems don't stem from a lack of market controls but from their top down command economy.
In both manufacturing and service economists measure productivity in profit dollars. It is not any harder to measure the productivity of UPS or FedEx than of GM or Boeing.
The real difference between the extraordinary productivity gains brought about by industrialization and those that are hoped for but seldom seen in office automation is that when you put a new machine in a factory you damned well make sure that you have trained operators and that the factory manager understands what he or she is managing. We stick increassingly complex machines on the desks of people who can't comprehend drag n' drop and put them under a manager who can figure out why he loses his T1 speed when he takes his laptop home. Its like giving 6 axis milling machines and robot welders to freaking monkeys. Computers are treated in many ways like a magic totem, give each employee a new totem and things will magically get better.
Alright, this "its the net" stuff is bogus. Most companies haven't had net access for more than three years. Really progressive companies may have had computers with web browsers and net access since 1995, but Solow's work is based on data that goes back at least to the seventies, if not earlier. I remember this being a topic of debate on public television in 1990 when there was no Web. In short, The Web is still new, the Solow paradox is much older.
I've seen some serious Exchange installations. 25000 users is far too much for one exchange machine to handle. I've seen 2000+ users on a quad processor Alhpaserver 4100 with an exchange database over 300GB and growing quickly. You can do 25000 users but you will need many servers each with their own RAID array and some serious, dedicated, professional management. POP3 might be a better option since the server requirements per user are lower.
At work we were hearing from Engineering and technical marketing folks that they had Intel dead to rights on the patent violations. The suits in upper management had been interested in finding a manufacturing partner for several years and saw an opportunity to sell their fab, grab some quick cash and get a stable manufacturing partner. The suits apparently never intended to engage in a long battle with Intel, they just wanted to dump the expensive fab so Digital would be a more attractive buy for Compaq.
Your comparison of ADSL to @home cable modem service is innaccurate. @home's service let users consume as much bandwidth as they wanted until the 128K download cap was instituted. DSL specifies guaranteed transfer rates and charges a monthly fee based on the level of service in the contract. This means they can put limits or fees on volume (i.e. >2GB per month) but they can't lower your download or upload cap like @home did because its specified in the service contract.
Funny thing, freeBSD's TCP/IP stack suffers from the same malady as Linux's and would have fared no better in the Mindcraft benchmarks. Serveral major, modern commercial Unices also have or until recently had this problem. Reason is that under real world operating conditions the TCP/IP stack is not the bottleneck, ethernet bandwidth is. Only many separate simultaneous requests and the file size requested is vanishingly small can you overwelm the TCP/IP stack and take a performance hit. This is why Microsoft insisted on very small static pages and multiple 100Mbit cards for the test, that was the only way they could get the number of requests high enough to hit the limits of the TCP/IP stack before they hit the limits of the ethernet protocol (or some other weakness in NT itself).
Sounds like you're slingin' alot of bull cookies there pardner. Opening Mozilla didn't slow it down. Netscape spending the first six months of what should have been Mozilla development time trying to sqeeze communicator 4.5 out the door instead is what slowed Mozilla down. Mozilla has been public just over a year, it should be done by now but its about six months late, Netscape wasted six months screwing around with communicator - you do the math. Your claim that setting up public CVS and other support systems cost netscape six months of development is total nonsense. Are we supposed to believe that dozens of skilled programmers lost six full months of good coding time to setting up a couple of web servers?!? Please.
Re:what's with all the self-appointed censors...
on
Rasterman Goes to VA
·
· Score: 1
Nobody has censored you, they have criticized you. It sounds to me like you want to criticize while remaining free from criticism yourself. If you can't handle the heat...
Certainly some public outlets with an anti-linux agenda will take flames and use them against Linux but the fact remains that they can only do that if we give them the ammunition in the first place. If we don't give them flames they can't use them against us. Also remember this isn't jsut about Mindcraft. Many journalists have been flamed to a crisp for their Linux articles. Most of them were simply ignorant or were working with inaccurate sources. Flaming these people does direct, measuable public harm to the reputation of the Linux community. Polite, courteous, well informed responses may help or may be wasted but at least they do no harm.
The author used his voice to remind us that flaming is rude, immature and EXTREMELY harmful to Linux. Supposedly the flamer's goal is to defend their beloved operating system, if this is so then I would think they'd want to know if they're causing more harm than good.
Freedom of speech does not include freedom from criticism of your speech.
$200 per month, Youch! Down here in GTE territory its $55.00 per month for 384Kbps each way plus ISP fees which I've been quoted at $45.00 per month. I was debating whether to go with this or go through Time Warner for Roadrunner cable modem access. I haven't found any useful info on Roadrunner yet though.
Yes, I know. Lets see, NT does not have any of the Microkernel's alleged stability, so to argue that a microkernel is inherently more stable and then site NT as an example is rather dumb. Mac OS X is only out in a server edition and looks OK but hasn't proven itself. BeOS is single user only and still beta though quite impressive. Show me a serious operating system that A) is a microkernel and B) has superior stability as a result.
So who has any real track record with microkenels? Microsoft. Hmmmmm....
Understand I don't mean any direspect to you, BeOS or Mac OS X but there's a whole lot more out there than just the desktop and right now microkernels are not necessary or worth the porting effort just to get enterprise stability. Linux, *BSD, AIX, Solaris, VMS, Tru64 Unix, HP/UX, OS400, OS390, they're all rock stable; NT, a microkernel, is good 23x6. Thats not much of a track record.
Pretty much everybody has eschewed the microkernel model at this point, not just Linux. For a while HP, IBM, Sun and DEC were all paying lip service to microkernels and saying that they would one day move to microkernels, but none of them have. (Yes, I know Digital Unix is based on Mach, but its Mach kernel 2.5 not the Mach microkernel v3.?) While microkernels may be more stable in theory, in practice good APIs and coding count for alot more.
My understanding is that the IP stack was not multi-threaded until service pack three. Is this correct? In any case its not simple multi-threading that gave NT the boost here, it was the ability to tie each card to a specific CPU and bond all the cards into one channel. Off the shelf NT is not that fast nor is Redhat 6.0.
Regarding ESR's comments on passwords, I did not interpret that as you did:
ESR gives the example of passwords as "information that does not want to be free". However, if that's the case, then why do we have to go to such lengths to keep our passwords secure? Why encrypt them? Why tediously remind users not to write them on Post-It notes on their monitors? It is precisely because unsecured information leaks around easily that we have to take security measures.
I took this as simply an example of information that should not have a clearing cost of zero just because the marginal cost to distribute or reproduce it approaches zero. Not as an example of "information that doesn't want to be free". Obviously information doesn't "want" anything and just as obviously treasure maps and important passwords tend to get distributed widely if not guarded closely. Remember the goal here is to provide economic justification for open source software and the readers ESR most wants to influence are probably the least likely to accept that "information wants to be free". ESR divorces his case from that argument, thus his readers don't have to accept that argument to accept ESR's point.
Nor is there a big developer following behind closed source CD player applets for Windows or the Mac. What you are missng is that many of those submissions on Freshmeat are small projects that require few members and in most cases add to or complement existing projects. There are hundreds of applets out there using GTK and designed for Gnome integration, the same can be said of QT/KDE or Gimp plugins or Apache modules. ESR's point was that open source development is not inherently unstable as some people claim and that the growth of open source submissions is proof of that.
I don't see many of these forking problems you talk about in any of the major open source projects. The Linux kernel is not forked, Apache is not, the was a fork in gcc which was handled well and eventually rolled back in as the main version once it had proven itself. Just not that big a deal.
ESR addressed both of your counter-examples in the paper.
These two quotes come from the following section near the bottom: http://www.tuxedo.org/~esr/writings/magic-cauldron /magic-cauldron-3.html
"In the short run, one can escape this trap by making bug-fix releases pose as new products with a new price attached, but consumers quickly tire of this. In the long run, therefore, the only way to escape is to have no competitors -- that is, to have an effective monopoly on one's market. In the end, there can be only one. "
"And, indeed, we have repeatedly seen this support-starvation failure mode kill off even strong second-place competitors in a market niche. (The pattern should be particularly clear to anyone who has ever surveyed the history of proprietary PC operating systems, word processors, accounting programs or business software in general.) The perverse incentives set up by the factory model fuel to a winner-take-all market dynamic in which even the winner's customers end up losing."
So according to Raymond, one can in the short run call bug fixes and patches new versions and charge for them but this leads to a condition where you tend towards a single monopoly player and customers dissatisfied with poorly supported sofware - sound familiar?
Re:Tough times for Intel?
on
1GHz Alphas
·
· Score: 1
I expect that the K7 has a good chance at grabbing SMP server marketshare from intel. The EV6 bus use a crossbar switch design for full processor clock speed communication between multiple processors providing much better performance than intel's model which requires that the processors communicate through the much slower 100MHz memory bus.
Slashdot Mirror
Yup the article was pretty bogus but let me ask you a question about the rest of your statement.
If we assume that people will pay more for a superior product then why hasn't the superior automobile you are now able to design led to greater profits? And if improving quality doesn't lead to greater profits where is the incentive to improve quality?
Here's two possible answers for you:
1) GM and Chrysler have both posted record profits in recent years.
2) In a near monopolistic market there is little incentive to improve quality, but when you have high quality asian and european competitors a failure to improve quality will result in a LOSS of profits, so you find yourself running faster just to stand still.
What you are talking about is manufacturing capacity, and you're right, overcapacity is bad, what Solow was talking about was profit per employee which should be the final result of productivity gains. That we want as much of as we can get.
By the way, I'd bet dollars to donuts that China's problems don't stem from a lack of market controls but from their top down command economy.
I quote from Hemos' posting "Of course, that doesn't explain why productivity wasn't rising pre-Windows, but c'est la vie."
In both manufacturing and service economists measure productivity in profit dollars. It is not any harder to measure the productivity of UPS or FedEx than of GM or Boeing.
The real difference between the extraordinary productivity gains brought about by industrialization and those that are hoped for but seldom seen in office automation is that when you put a new machine in a factory you damned well make sure that you have trained operators and that the factory manager understands what he or she is managing. We stick increassingly complex machines on the desks of people who can't comprehend drag n' drop and put them under a manager who can figure out why he loses his T1 speed when he takes his laptop home. Its like giving 6 axis milling machines and robot welders to freaking monkeys. Computers are treated in many ways like a magic totem, give each employee a new totem and things will magically get better.
Alright, this "its the net" stuff is bogus. Most companies haven't had net access for more than three years. Really progressive companies may have had computers with web browsers and net access since 1995, but Solow's work is based on data that goes back at least to the seventies, if not earlier. I remember this being a topic of debate on public television in 1990 when there was no Web. In short, The Web is still new, the Solow paradox is much older.
I've seen some serious Exchange installations. 25000 users is far too much for one exchange machine to handle. I've seen 2000+ users on a quad processor Alhpaserver 4100 with an exchange database over 300GB and growing quickly. You can do 25000 users but you will need many servers each with their own RAID array and some serious, dedicated, professional management. POP3 might be a better option since the server requirements per user are lower.
So if we add up TurboLinux, Debian, Redhat and Slackware what do we get?
I thought SGS Thomson had made the move away from defense and into consumer and entertainment markets, but I could be wrong.
At work we were hearing from Engineering and technical marketing folks that they had Intel dead to rights on the patent violations. The suits in upper management had been interested in finding a manufacturing partner for several years and saw an opportunity to sell their fab, grab some quick cash and get a stable manufacturing partner.
The suits apparently never intended to engage in a long battle with Intel, they just wanted to dump the expensive fab so Digital would be a more attractive buy for Compaq.
Your comparison of ADSL to @home cable modem service is innaccurate. @home's service let users consume as much bandwidth as they wanted until the 128K download cap was instituted. DSL specifies guaranteed transfer rates and charges a monthly fee based on the level of service in the contract. This means they can put limits or fees on volume (i.e. >2GB per month) but they can't lower your download or upload cap like @home did because its specified in the service contract.
Funny thing, freeBSD's TCP/IP stack suffers from the same malady as Linux's and would have fared no better in the Mindcraft benchmarks. Serveral major, modern commercial Unices also have or until recently had this problem. Reason is that under real world operating conditions the TCP/IP stack is not the bottleneck, ethernet bandwidth is. Only many separate simultaneous requests and the file size requested is vanishingly small can you overwelm the TCP/IP stack and take a performance hit. This is why Microsoft insisted on very small static pages and multiple 100Mbit cards for the test, that was the only way they could get the number of requests high enough to hit the limits of the TCP/IP stack before they hit the limits of the ethernet protocol (or some other weakness in NT itself).
Sounds like you're slingin' alot of bull cookies there pardner.
Opening Mozilla didn't slow it down. Netscape spending the first six months of what should have been Mozilla development time trying to sqeeze communicator 4.5 out the door instead is what slowed Mozilla down. Mozilla has been public just over a year, it should be done by now but its about six months late, Netscape wasted six months screwing around with communicator - you do the math.
Your claim that setting up public CVS and other support systems cost netscape six months of development is total nonsense. Are we supposed to believe that dozens of skilled programmers lost six full months of good coding time to setting up a couple of web servers?!? Please.
Nobody has censored you, they have criticized you. It sounds to me like you want to criticize while remaining free from criticism yourself. If you can't handle the heat...
Certainly some public outlets with an anti-linux agenda will take flames and use them against Linux but the fact remains that they can only do that if we give them the ammunition in the first place. If we don't give them flames they can't use them against us.
Also remember this isn't jsut about Mindcraft. Many journalists have been flamed to a crisp for their Linux articles. Most of them were simply ignorant or were working with inaccurate sources. Flaming these people does direct, measuable public harm to the reputation of the Linux community. Polite, courteous, well informed responses may help or may be wasted but at least they do no harm.
Hey, why did this get moderated down? The more info we get out on positive advocacy the better.
The author used his voice to remind us that flaming is rude, immature and EXTREMELY harmful to Linux. Supposedly the flamer's goal is to defend their beloved operating system, if this is so then I would think they'd want to know if they're causing more harm than good.
Freedom of speech does not include freedom from criticism of your speech.
$200 per month, Youch! Down here in GTE territory its $55.00 per month for 384Kbps each way plus ISP fees which I've been quoted at $45.00 per month. I was debating whether to go with this or go through Time Warner for Roadrunner cable modem access. I haven't found any useful info on Roadrunner yet though.
Yes, I know. Lets see, NT does not have any of the Microkernel's alleged stability, so to argue that a microkernel is inherently more stable and then site NT as an example is rather dumb. Mac OS X is only out in a server edition and looks OK but hasn't proven itself. BeOS is single user only and still beta though quite impressive. Show me a serious operating system that A) is a microkernel and B) has superior stability as a result.
So who has any real track record with microkenels? Microsoft. Hmmmmm....
Understand I don't mean any direspect to you, BeOS or Mac OS X but there's a whole lot more out there than just the desktop and right now microkernels are not necessary or worth the porting effort just to get enterprise stability. Linux, *BSD, AIX, Solaris, VMS, Tru64 Unix, HP/UX, OS400, OS390, they're all rock stable; NT, a microkernel, is good 23x6. Thats not much of a track record.
Pretty much everybody has eschewed the microkernel model at this point, not just Linux. For a while HP, IBM, Sun and DEC were all paying lip service to microkernels and saying that they would one day move to microkernels, but none of them have. (Yes, I know Digital Unix is based on Mach, but its Mach kernel 2.5 not the Mach microkernel v3.?)
While microkernels may be more stable in theory, in practice good APIs and coding count for alot more.
My understanding is that the IP stack was not multi-threaded until service pack three. Is this correct? In any case its not simple multi-threading that gave NT the boost here, it was the ability to tie each card to a specific CPU and bond all the cards into one channel. Off the shelf NT is not that fast nor is Redhat 6.0.
---------
Nor is there a big developer following behind closed source CD player applets for Windows or the Mac. What you are missng is that many of those submissions on Freshmeat are small projects that require few members and in most cases add to or complement existing projects. There are hundreds of applets out there using GTK and designed for Gnome integration, the same can be said of QT/KDE or Gimp plugins or Apache modules. ESR's point was that open source development is not inherently unstable as some people claim and that the growth of open source submissions is proof of that.
I don't see many of these forking problems you talk about in any of the major open source projects. The Linux kernel is not forked, Apache is not, the was a fork in gcc which was handled well and eventually rolled back in as the main version once it had proven itself. Just not that big a deal.
ESR addressed both of your counter-examples in the paper.
n /magic-cauldron-3.html
These two quotes come from the following section near the bottom: http://www.tuxedo.org/~esr/writings/magic-cauldro
"In the short run, one can escape this trap by making bug-fix releases pose as new products with a new price attached, but consumers quickly tire of this. In the long run, therefore, the only way to escape is to have no competitors -- that is, to have an effective monopoly on one's market. In the end, there can be only one. "
"And, indeed, we have repeatedly seen this support-starvation failure mode kill off even strong second-place competitors in a market niche. (The pattern should be particularly clear to anyone who has ever surveyed the history of proprietary PC operating systems, word processors, accounting programs or business software in general.) The perverse incentives set up by the factory model fuel to a winner-take-all market dynamic in which even the winner's customers end up losing."
So according to Raymond, one can in the short run call bug fixes and patches new versions and charge for them but this leads to a condition where you tend towards a single monopoly player and customers dissatisfied with poorly supported sofware - sound familiar?
I expect that the K7 has a good chance at grabbing SMP server marketshare from intel. The EV6 bus use a crossbar switch design for full processor clock speed communication between multiple processors providing much better performance than intel's model which requires that the processors communicate through the much slower 100MHz memory bus.