Each time they send you an email, just FAX it back to them, together with a polite note explaining that you have carefully considered their offer, and are not interested at this time.
1. Conclusions
Warranties don't make sense for software. Vendors don't provide them,
and consumers don't rely on them.
2. Manufacturing defects
Manufactured goods--e.g. garments, refrigerators, automobiles--are
subject to defects in materials and workmanship: manufacturing
defects. These occur essentially at random. Manufacturers can
typically reduce defect rates by spending more money; however, defects
can never be entirely eliminated.
3. Selling and buying
On this analysis, a manufacturer does not sell a product per se;
rather, a manufacturer sells a product with some defect rate.
Similarly, a buyer does not buy simply a product, but a product with
some defect rate.
4. Large quantities
This situation causes no special problems as long as the buyer is
purchasing a product in large quantities. A manufacturer sells a
product with some defect rate; the buyer discovers through experience
what that defect rate is. Knowing the defect rate, the buyer can
assess the true cost of the product, and take that into account when
budgeting or making competitive purchase decisions. In short, the
buyer can buy what the seller is selling.
5. Small quantities
This model breaks down when a buyer is purchasing a product in small
quantities. For example, consumers typically purchase automobiles in
quantity one. A manufacturer may sell automobiles with a 1% defect
rate, but a consumer who buys one automobile does not get a 1% defect
rate--they get one automobile, and it either works or it doesn't. In
an important sense, the buyer cannot buy what the seller is selling.
This is a kind of market failure.
6. Warranties
Warranties remedy this market failure. A warranty binds the seller to
repair or replace defective units. This allows the seller to sell a
product with an effective defect rate of zero, even though it is not
physically possible to manufacture such a product. By providing a
warranty, the seller can sell something that the buyer can buy, even
in quantity one.
7. Non-purposes
It is important to emphasize that the sole purpose of a warranty is to
remedy this market failure. A warranty does not force manufacturers to
bear the cost of their manufacturing defects: the cost of warranty
service is passed to the buyer in the original purchase price. Neither
does a warranty force manufacturers to produce high-quality products.
In fact, a warranty allows manufacturers to produce low-quality
products, if that turns out to minimize the total of their
manufacturing and warranty costs. (Sofa-bed manufacturers appear to
take this approach.)
8. Software
Software is not tangible, it is not a manufactured good, and it is not
subject to manufacturing defects. Furthermore, each copy of a software
product is, by definition, identical.
9. Distribution media
Some software products are distributed on physical media, such as
floppy disks or CD-ROM. These distribution media are tangible,
manufactured goods, and are subject to defects. The actual defect rate
is quite small; nonetheless, virtually every commercial software
product warrants its distribution media. These warranties are simple,
straightforward, and uncontroversial; buyers and sellers have a common
understanding of them. Warranty service usually consists of exchanging
defective media for a working copy, usually at the place of sale.
10. No warranty
Virtually no commercial software product warrants its functionality.
The reason for this is that there is no market failure for a warranty
to remedy.
11. No market failure
As discussed above, there is a market failure when a consumer buys a
manufactured product: they cannot know whether the one they get will
be defective. This cannot happen with software. Every copy of a
software product is identical: every buyer will experience exactly the
same benefits and suffer exactly the same deficiencies, whatever they
may be. Good or bad; working or broken: every buyer gets exactly what
the seller is selling, even in quantity one. The market failure that a
warranty remedies simply does not occur with software products.
12. The grapevine
In principle, a consumer takes a risk when they buy a software
product, because they do not know whether it will do what the seller
claims. In practice, information about the performance of software
products is available from many sources: vendor demonstrations,
consumer testing organizations, reviews in industry publications, web
sites, on-line discussion groups, mailing lists, friends, colleagues.
In any case, there is no reason to think that the markets cannot
provide whatever information about software products consumers may
demand. If consumers have information about the performance of a
software product, and every copy of the product performs identically,
then consumers know exactly what they are buying, and again there is
no need for a warranty.
13. Structural differences
Many commercial software products are unreliable--they have bugs. Their
is some view that software vendors do not warrant their software
because they are, variously, too inept to product reliable products
or too greedy to bear the cost of warranty service. However, this view
is inconsistent with the low defect rates and near universal
warranties provided for distribution media. A better explanation is
that there are fundamental structural differences between distribution
media (tangible) and software (intangible), and that warranties are
appropriate for the former but not the latter.
I work with computers
I play with computers
I own computers
I read on computers
I write on computers
I write about computers
I program computers
I teach computers to anyone who will stand still long enough to listen.
In the town of Littleton, MA, I vote on a paper ballot, and I like it just fine.
how much longer until we can create a system to mimic the human body closely enough to try our genetic experiements out digitally?
Computing the shape of a single protien from its amino-acid sequence (the folding problem) is still unsolved and generally regarded as computationally intractable.
until that point, I really don't see what good the Genome has done for us.
You don't have to do a complete digital simulation of the human body.
There are (useful) drugs that cannot be marketed because different people have radically different dose-response curves. If you can correlate patient's repsonses with variations in one (or several) genes, then you can genotype the patients and adjust the dose accordingly. We are already within field-goal range on this one.
We've had the patent system in this country for over two hundred and twenty-five years... I think that the arguments you made were made about polymers, they were made about automobiles, they were made about telephones, and they were made about telegraph, and the system survived very well...
The fact that something survives doesn't mean that it is good. Malaria and tuberculosis survive, but we'd really rather they went away. The relevant question isn't whether the patent system survives. The relevant question is whether its benefits are greater than its costs.
Costs
the cost of running the patent office
the costs of dealing with the patent system
writing patents
reading patents
litigating patents
negotiating patent licenses
the burden of license fees for every invention that would have been invented even without a patent system.
The burden isn't the license fee itself: that just moves money from one pocket to another; the overall wealth of the economy is unchanged. The burden is the cost to the economy when individuals forgo the use of an invention because the license fee is more than the invention is worth to them. This decreases the aggregate wealth of the economy.
Benefits
The value of every invention that wouldn't have been invented in the absence of a patent system.
The only way you will become successful is by having a publicity and promotion campaign behind you that elevates what you're doing above what your competition is doing.
This is true in the market that the record companies have created for themselves. Record companies are sales and promotion engines. They have created a competitive distribution model where you need their services to succeed. It works out rather nicely for them.
statistically, for every one band that you hear about, for every one band that a record company helps make successful, they lose their fucking shirt on the nine other ones you never hear about
There is more creative talent out there than the record companies need to run their business. They take what they need and leave the rest.
me and you could form a band together, and we could like, make a demo and then we could put it up on Napster. Who is going to give a fuck? Nobody's going to care, because they don't know anything about what sets my and your band out from the gardener and the guy who cleans my pool's band.
The record companies have defined success as million-selling disks, because they make more money with a small number of big sellers than a large number of small sellers.
The world could be different.
What if the (global) measure of success was not the number of dollars in record execs pockets, but the number of songs written and heard by at least one person.
People writing songs, that voices never shared - Paul Simon
If you view this as a strictly religious question, then it really isn't subject to analysis. You can find verses in the bible (or any other religious text) to support just about any interpretation of anything.
If you view it as a practical issue, then an historical perspective may help.
Hunter/gatherers don't structure time as we do. They don't divide time into periods of work and rest. In fact, they don't make the distinctions that we do between work/play/rest/leisure. They just hang out on the veldt and do stuff. Some is interesting, some is boring; some takes muscles; some doesn't; but it's all just stuff. The only structure on their time is the days and the seasons, and those are immutable, so why worry about it?
Furthermore, hunter/gatherers tend to be nomadic (because the land won't support intensive use). Nomads can't accumulate more than they can carry. So there isn't really any point in working harder, or smarter: you can't take it with you, even in this life.
Anthropological evidence is that hunter/gathers only spend 4 to 6 hours per day doing things that we call work. The balance of their time is spent in sleep, socializing, creative, recreational and ritual activites.
All that changes with the invention of agriculture. Farmers are out there in the fields every day, working now for a harvest that is months away. The more land you farm, the more food you grow, so incremental time spent in the fields pays off later.
Agriculture supports more intensive use of the land, so people can stop wandering and start accumulating stuff, like surplus food, or tools, or housing. Specially ambitious and intense people start accumulating armies and castles.
So everyone is out there working hard, and accumulating stuff, and you can do it 16x7 if you like, and eventually you have to stop and ask: How much is enough? How hard do we work?
And the answer that many societies came up with, going 8000 years back to the invention of agriculture, is that we rest 1 day in N, where N is typically in the range of 4 to 10. Our value of 7 traces back to the Persians.
Religion is central to the organization of pre-industrial societies, so the calandar typically became a religous matter in those societies, with attendant ritual, ceremony, and claims of divine authority.
In our society, business has largely eclipsed religion as the institution that organizes time. People work to the clock; critical services run 24x7; factories run 2nd and 3rd shift; retail runs all weekend. And the people at the top--programmers, lawyers, consultants--can work as many hours as they like.
So we're back to the question that the early farmers faced: how hard do we work? How much is enough? And we have to have an answer. People who work all the time burn out, and the currency of the phrase burn out shows just how real and prevelant this problem is for us.
You can find your answer in different places. Some people manage the problem informally: they know their limits, and take time off when they need it. Some people are more comfortable with rules. You can make up your own rules, or you can adopt someone else's rules and follow them.
The ancient Hebrews had rules, and wrote them down; you can follow them if you like. Keep in mind that their rules were written long, long ago, in a wheat field far, far away. Trying to apply them literally in our society probably isn't useful. Claims of divine authority notwithstanding, their rules were written by people to serve people; the only reason for you to follow them is that they also serve you.
Anyway, I think that the true future of 3d monitors will be something like an Active Holographic Display; i.e. a monitor who's elements can emit different light patterns in different directions. For example, an LCD-type screen that had a mechanism by which one pixel could emit blue light in one direction and green in another. The same concept is what makes holograms work. Surely it couldn't be too hard to make a monitor that does this same kind of thing?
A SMOF (Small Matter Of Fabrication)
To make a holographic display, all you need is
a screen with a dot pitch smaller than the wave length of light
and the pixels have to emit coherent (i.e. laser) light
and you have to control the phase at which each pixel emits
Yeah...that's the ticket: if you control the amplitude and phase of the light emitted at each point on a surface, then you can synthesize any image.
Computing the required video bandwidth is left as an exercise for the reader.
Boycotts are also somewhat repugnant to the free-market philosophy many individualists hold.
Only if they are enforced by law or violence
1....We need family farming
Why? We used to have family farming, and most family farmers were dirt poor. Today, we have economies of scale and division of labor. Should we have family
iron foundries?
car makers?
drug companies?
From an economic standpoint, the reason that we don't have family farmers any more is that they aren't competitive. Why should I pay higher grocery bills to support someone else's vision of family farming?
2....corporate entities should also embrace moral and ideological values -- of their own choosing --apart from pleasing stockholders.
AAAIIIEEEE!!!!!
The need to return a profit to shareholders is one of the few effective constraints on corporate power. Lifting this constraint creates major problems of corporate governance.
As a practical matter, the values that corporate entities choose--apart from pleasing stockholders--are protecting the jobs, perks, and compensation of their own top management.
After that, it only gets worse. Do you really want people like Ross Perot and Bill Gates using their $Billions to advance their own personal moral and ideological values in a corporate--rather than political--arena?
3....Workers are entitled to safe, creative and secure work environments, to freedom from continuous downsizings, re-structurings, layoffs and "re-engineerings."
I have a safe, creative and secure work environment. This has a lot to do with the fact that my labor has high value, which in turn has at least something to do with the fact that corporations are free to lay off workers, so that labor moves to its best and highest use. (I've been laid off, and in a considerably worse job market than the current one.)
I interviewed at a company yesterday. My resume lists substantial open-source software and documentation. All 4 interviewers had my resume in front of them. Not one remarked on the open source stuff.
1. Ask what the false positive rate for the WAVE program is. 2. Ask what the false negative rate is. 3. Ask how they measure these rates. 4. Ask if the phrase "peer reviewed" appears in any of their protocols. 5. If they don't know/won't tell, then posit that the FP and FN rates are both 50%. If they balk at this, go back to question 1...
Slashdot Mirror
Each time they send you an email, just FAX it back to them, together with a polite note explaining that you have carefully considered their offer, and are not interested at this time.
1. Conclusions
Warranties don't make sense for software. Vendors don't provide them, and consumers don't rely on them.
2. Manufacturing defects
Manufactured goods--e.g. garments, refrigerators, automobiles--are subject to defects in materials and workmanship: manufacturing defects. These occur essentially at random. Manufacturers can typically reduce defect rates by spending more money; however, defects can never be entirely eliminated.
3. Selling and buying
On this analysis, a manufacturer does not sell a product per se; rather, a manufacturer sells a product with some defect rate. Similarly, a buyer does not buy simply a product, but a product with some defect rate.
4. Large quantities
This situation causes no special problems as long as the buyer is purchasing a product in large quantities. A manufacturer sells a product with some defect rate; the buyer discovers through experience what that defect rate is. Knowing the defect rate, the buyer can assess the true cost of the product, and take that into account when budgeting or making competitive purchase decisions. In short, the buyer can buy what the seller is selling.
5. Small quantities
This model breaks down when a buyer is purchasing a product in small quantities. For example, consumers typically purchase automobiles in quantity one. A manufacturer may sell automobiles with a 1% defect rate, but a consumer who buys one automobile does not get a 1% defect rate--they get one automobile, and it either works or it doesn't. In an important sense, the buyer cannot buy what the seller is selling. This is a kind of market failure.
6. Warranties
Warranties remedy this market failure. A warranty binds the seller to repair or replace defective units. This allows the seller to sell a product with an effective defect rate of zero, even though it is not physically possible to manufacture such a product. By providing a warranty, the seller can sell something that the buyer can buy, even in quantity one.
7. Non-purposes
It is important to emphasize that the sole purpose of a warranty is to remedy this market failure. A warranty does not force manufacturers to bear the cost of their manufacturing defects: the cost of warranty service is passed to the buyer in the original purchase price. Neither does a warranty force manufacturers to produce high-quality products. In fact, a warranty allows manufacturers to produce low-quality products, if that turns out to minimize the total of their manufacturing and warranty costs. (Sofa-bed manufacturers appear to take this approach.)
8. Software
Software is not tangible, it is not a manufactured good, and it is not subject to manufacturing defects. Furthermore, each copy of a software product is, by definition, identical.
9. Distribution media
Some software products are distributed on physical media, such as floppy disks or CD-ROM. These distribution media are tangible, manufactured goods, and are subject to defects. The actual defect rate is quite small; nonetheless, virtually every commercial software product warrants its distribution media. These warranties are simple, straightforward, and uncontroversial; buyers and sellers have a common understanding of them. Warranty service usually consists of exchanging defective media for a working copy, usually at the place of sale.
10. No warranty
Virtually no commercial software product warrants its functionality. The reason for this is that there is no market failure for a warranty to remedy.
11. No market failure
As discussed above, there is a market failure when a consumer buys a manufactured product: they cannot know whether the one they get will be defective. This cannot happen with software. Every copy of a software product is identical: every buyer will experience exactly the same benefits and suffer exactly the same deficiencies, whatever they may be. Good or bad; working or broken: every buyer gets exactly what the seller is selling, even in quantity one. The market failure that a warranty remedies simply does not occur with software products.
12. The grapevine
In principle, a consumer takes a risk when they buy a software product, because they do not know whether it will do what the seller claims. In practice, information about the performance of software products is available from many sources: vendor demonstrations, consumer testing organizations, reviews in industry publications, web sites, on-line discussion groups, mailing lists, friends, colleagues. In any case, there is no reason to think that the markets cannot provide whatever information about software products consumers may demand. If consumers have information about the performance of a software product, and every copy of the product performs identically, then consumers know exactly what they are buying, and again there is no need for a warranty.
13. Structural differences
Many commercial software products are unreliable--they have bugs. Their is some view that software vendors do not warrant their software because they are, variously, too inept to product reliable products or too greedy to bear the cost of warranty service. However, this view is inconsistent with the low defect rates and near universal warranties provided for distribution media. A better explanation is that there are fundamental structural differences between distribution media (tangible) and software (intangible), and that warranties are appropriate for the former but not the latter.
If you run into 38 engineers who are looking for work, just send them to us.
Thanks.
I play with computers
I own computers
I read on computers
I write on computers
I write about computers
I program computers
I teach computers to anyone who will stand still long enough to listen.
In the town of Littleton, MA, I vote on a paper ballot, and I like it just fine.
Computing the shape of a single protien from its amino-acid sequence (the folding problem) is still unsolved and generally regarded as computationally intractable.
until that point, I really don't see what good the Genome has done for us.
You don't have to do a complete digital simulation of the human body.
There are (useful) drugs that cannot be marketed because different people have radically different dose-response curves. If you can correlate patient's repsonses with variations in one (or several) genes, then you can genotype the patients and adjust the dose accordingly. We are already within field-goal range on this one.
I'm waiting for someone to splice the genes for THC, cocaine, morphine, etc. into yeast, so you can brew up your fix in the privacy of your own home.
If he knew how to do that he'd be a novelist, not a hacker.
You probably lose even more if you move to a different country.
At least they speak english in BC...
The fact that something survives doesn't mean that it is good. Malaria and tuberculosis survive, but we'd really rather they went away. The relevant question isn't whether the patent system survives. The relevant question is whether its benefits are greater than its costs.
Costs
- the cost of running the patent office
- the costs of dealing with the patent system
- writing patents
- reading patents
- litigating patents
- negotiating patent licenses
- the burden of license fees for every invention that would have been invented even without a patent system.
The burden isn't the license fee itself: that just moves money from one pocket to another; the overall wealth of the economy is unchanged. The burden is the cost to the economy when individuals forgo the use of an invention because the license fee is more than the invention is worth to them. This decreases the aggregate wealth of the economy.Benefits
This is true in the market that the record companies have created for themselves. Record companies are sales and promotion engines. They have created a competitive distribution model where you need their services to succeed. It works out rather nicely for them.
statistically, for every one band that you hear about, for every one band that a record company helps make successful, they lose their fucking shirt on the nine other ones you never hear about
There is more creative talent out there than the record companies need to run their business. They take what they need and leave the rest.
me and you could form a band together, and we could like, make a demo and then we could put it up on Napster. Who is going to give a fuck? Nobody's going to care, because they don't know anything about what sets my and your band out from the gardener and the guy who cleans my pool's band.
The record companies have defined success as million-selling disks, because they make more money with a small number of big sellers than a large number of small sellers.
The world could be different.
What if the (global) measure of success was not the number of dollars in record execs pockets, but the number of songs written and heard by at least one person.
People writing songs, that voices never shared
- Paul Simon
If you view it as a practical issue, then an historical perspective may help.
Hunter/gatherers don't structure time as we do. They don't divide time into periods of work and rest. In fact, they don't make the distinctions that we do between work/play/rest/leisure. They just hang out on the veldt and do stuff. Some is interesting, some is boring; some takes muscles; some doesn't; but it's all just stuff. The only structure on their time is the days and the seasons, and those are immutable, so why worry about it?
Furthermore, hunter/gatherers tend to be nomadic (because the land won't support intensive use). Nomads can't accumulate more than they can carry. So there isn't really any point in working harder, or smarter: you can't take it with you, even in this life.
Anthropological evidence is that hunter/gathers only spend 4 to 6 hours per day doing things that we call work. The balance of their time is spent in sleep, socializing, creative, recreational and ritual activites.
All that changes with the invention of agriculture. Farmers are out there in the fields every day, working now for a harvest that is months away. The more land you farm, the more food you grow, so incremental time spent in the fields pays off later.
Agriculture supports more intensive use of the land, so people can stop wandering and start accumulating stuff, like surplus food, or tools, or housing. Specially ambitious and intense people start accumulating armies and castles.
So everyone is out there working hard, and accumulating stuff, and you can do it 16x7 if you like, and eventually you have to stop and ask: How much is enough? How hard do we work?
And the answer that many societies came up with, going 8000 years back to the invention of agriculture, is that we rest 1 day in N, where N is typically in the range of 4 to 10. Our value of 7 traces back to the Persians.
Religion is central to the organization of pre-industrial societies, so the calandar typically became a religous matter in those societies, with attendant ritual, ceremony, and claims of divine authority.
In our society, business has largely eclipsed religion as the institution that organizes time. People work to the clock; critical services run 24x7; factories run 2nd and 3rd shift; retail runs all weekend. And the people at the top--programmers, lawyers, consultants--can work as many hours as they like.
So we're back to the question that the early farmers faced: how hard do we work? How much is enough? And we have to have an answer. People who work all the time burn out, and the currency of the phrase burn out shows just how real and prevelant this problem is for us.
You can find your answer in different places. Some people manage the problem informally: they know their limits, and take time off when they need it. Some people are more comfortable with rules. You can make up your own rules, or you can adopt someone else's rules and follow them.
The ancient Hebrews had rules, and wrote them down; you can follow them if you like. Keep in mind that their rules were written long, long ago, in a wheat field far, far away. Trying to apply them literally in our society probably isn't useful. Claims of divine authority notwithstanding, their rules were written by people to serve people; the only reason for you to follow them is that they also serve you.
A SMOF (Small Matter Of Fabrication)
To make a holographic display, all you need is
- a screen with a dot pitch smaller than the wave length of light
- and the pixels have to emit coherent (i.e. laser) light
- and you have to control the phase at which each pixel emits
Yeah...that's the ticket: if you control the amplitude and phase of the light emitted at each point on a surface, then you can synthesize any image.Computing the required video bandwidth is left as an exercise for the reader.
Sure it is. I'm boycotting Nestle.
Boycotts are also somewhat repugnant to the free-market philosophy many individualists hold.
Only if they are enforced by law or violence
1....We need family farming
Why? We used to have family farming, and most family farmers were dirt poor. Today, we have economies of scale and division of labor. Should we have family
From an economic standpoint, the reason that we don't have family farmers any more is that they aren't competitive. Why should I pay higher grocery bills to support someone else's vision of family farming?
2....corporate entities should also embrace moral and ideological values -- of their own choosing --apart from pleasing stockholders.
AAAIIIEEEE!!!!!
The need to return a profit to shareholders is one of the few effective constraints on corporate power. Lifting this constraint creates major problems of corporate governance.
As a practical matter, the values that corporate entities choose--apart from pleasing stockholders--are protecting the jobs, perks, and compensation of their own top management.
After that, it only gets worse. Do you really want people like Ross Perot and Bill Gates using their $Billions to advance their own personal moral and ideological values in a corporate--rather than political--arena?
3....Workers are entitled to safe, creative and secure work environments, to freedom from continuous downsizings, re-structurings, layoffs and "re-engineerings."
I have a safe, creative and secure work environment. This has a lot to do with the fact that my labor has high value, which in turn has at least something to do with the fact that corporations are free to lay off workers, so that labor moves to its best and highest use. (I've been laid off, and in a considerably worse job market than the current one.)
Read Lessig's piece .
It is better than the sound bite.
I interviewed at a company yesterday.
My resume lists substantial open-source software and documentation.
All 4 interviewers had my resume in front of them.
Not one remarked on the open source stuff.
1 year at 12x6 and you can go straight from school to burnout w/o having to detour through the real world...
At In praise of cheap labor, MIT Economist Paul Krugman argues, "Bad jobs at bad wages are better than no jobs at all..."
1. Ask what the false positive rate for the WAVE program is. 2. Ask what the false negative rate is. 3. Ask how they measure these rates. 4. Ask if the phrase "peer reviewed" appears in any of their protocols. 5. If they don't know/won't tell, then posit that the FP and FN rates are both 50%. If they balk at this, go back to question 1...
Yes. For a more detailed analysis, see How The Internet Will Make The Record Labels Evaporate.