Ah. Just read the article. It appears that the prize was the idea of the one J. Craig Venter. My initial impression that it is all the work of self-serving, egotistical, self-publicists seems to have been born out.
Because the people who are attempting to do this already, could almost certainly not care less about 5million dollars either way. If you the X foundation were venture capital offering 5 million dollars no one would have noticed either way. Or better still a philanthropic research charity. They should try sponsering research, rather than living in it's reflected glory.
You can not sequence a genome this cheap. The reagents cost more than this. The point is that it would require new technology, new chemistry (or better, no chemistry). There are at least 4 or 5 methods on the horizon though.
Yes, of course, it would cost much more than this. 5 million dollars is nothing at all in R&D. People are very expensive. And, unlike human space travel, there are quite a few fairly immediate applications for this technology and lots of people who will pay for it.
It's great advertising for the X foundation though. Someone else does all the work, they get to appear visionary and it only costs 5 million. Pretty clever.
I have no problems with the idea that evolution should be open to question, and investigation. It is, after all, what evolutionary biologists do for a living.
What I find irritating is "evolution is wrong, it says so in the bible". When faced with this, I have no real problem in ridicule. There is, after all, no mechanism for arguing against it.
"I'm against all public education systems. I don't believe they've worked."
How interesting. I guess from this statement that you've made some efforts to look at "all public education systems". As it happens, the public education system in the UK is pretty good.
"That being said, if we must have them, let's focus on pure education -- facts, repetition, useful classes: how to read, write and perform basic math. At most, some basic scientific theory might be OK."
This is not "pure" education, it's a lack of education. If we focus on "facts", then this leaves us almost nothing to actually teach. Even reading is pretty useless, unless you have something to read about.
As for expecting parents to teach, I really can't see your point. It's like suggesting that we should expect parents to provide health care. We have professionalised health workers, why not professional teachers as well. If parents want to teach, of course, they can.
"What amazes me as a Swede is that all Anglo-saxon countries I've been to build so incredibly flimsy and energy-inefficient houses. England, Australia, and from what I've heard, the US as well. I mean, you are rich countries, why build like third world?"
It depends where in the country you are. Quite a lot of our cities have very integrated buildings--Edinburgh and Glasgow are obvious examples--where the buildings are 100+ years old. Are really really cold. The windows don't fit properly and there are no cavity walls. My current flat in Newcastle which is an old industrial terrace suffers from the same problem. Fixing the problem would, pretty much, require knocking down the entire street.
Newer builds are much better though, with roof and wall insulation. My parents house, for example, is probably three times the inner area of mine and has heating costs about 1/3 mine. I have to open the window there at night to cool the bedroom to a comfortable temperature, even in winter when it's -5 outside.
Of course, in Sweden, you don't have this problem; after all, the winters there are cold enough, that anything older than 10 years has freeze fractured to the point of falling down.
More seriously, though, it is very mild in Britain. It doesn't get hot in summer nor really cold in winter.
Rains a lot though; check out our damp proof coursings if you want to see expert enginneering.
It's probably a little bit more than that, as it's managed. Also, it's doubling in size every 10 months; a problem as the rate of increase of hard drive size is something like once every 18 months. This means the cost of providing this storage will increase exponentially.
Incidentally, it's not "genetic data"--that is sequence data. It's trace data which is then interpreted to produce sequence data. So actually, the data storage requirements for each base takes more than 2 bits. Moreover it's redundant (DNA is sequencing at least 10 times). So you're probably several orders of magnitude out with your calculations.
"I have to wonder how many of these entries are quasi-redundant."
All of them, pretty much. This is a trace archive. It stores the traces as they come of the sequencing machine. Given that DNA is normally sequenced to 10x (five times in both directions), most of the data in this database will be replicates.
"Sadly, the trend here seems to be more of 'sequence it, upload it, and patent it' instead of 'sequence it, upload it, figure out what it does/makes, do something useful with it'."
Data collection is the bed rock of all good science. You can sit and think of clever hypotheses all that you like, but it's all junk unless you actually have the data to test it against.
People are already doing interesting things with this sort of trace data that were not thought of when it first came out (predication of polymorphisms is possible for instance). More will come over time.
Gather, storing and archiving data is vital. The trace archive is a modern day Library of Alexandria.
"By "real" job, I mean a job in which you are required to meet deadlines imposed by customers and to produce end products specified by customers, otherwise your business fails."
RMS is no doubt aware of the pressures that people are under. As he says, if you send someone a word doc, you are pressuring them to give up their freedom. Which is true enough. If you are working in the environment then you may indeed end up giving up some of your freedom.
Most of us are willing to make these compromises, because they make the rest of our life easier. Stallman has remade his life so that he does not.
"Frankly, if I was running a business and I had to keep my family fed, there's no way in hell I'd be relying on GNU to develop the tools I needed. Their track record is simply appalling."
You give two examples, XEmacs and Hurd. One of which produced some perfectly usable software in common usage--the fork is unfortunate, but there you have it. Hurd you are right about, of course. Two examples? The rate of failed software projects in industry is at least 50% -- that is 50% of software developed never gets used. Most of it disappoints the customers. Pretty much all of it breaks at one stage or another.
In this light, your two examples fail to convince me that Gnu's "track record is appalling".
Besides, if you read the interview, he wasn't suggesting that you donate cash to Gnu. It is not, and never has been, a software house. If you ask them, "will you write this piece of software that I really need", then they are likely to say no. There are many other programmers who will do this for you.
Phil
Re:Emacs vs Eclipse: A losing battle
on
The Future of Emacs
·
· Score: 2, Insightful
Actually, I think that you are missing the major advantage that Eclipse has over Emacs, which is that it's written in Java. While I like lisp, and think it's a very nice language, there is a problem with library support (for Emacs-Lisp, rather than Lisp in general). Writing packages for Emacs is quite hard. When you get down to it, this means that the rate at which new and usable packages come out in slower than it needs to be. And there are fewer people who are able to fix the bugs with it.
Having said that, the internet is a large place, and people do keep on writing packages for it. Look at something like semantic, or ECB, or muse or auctex. These are large packages, many of which are less than 4 years old.
I think that Emacs will exist for quite a while yet. It may be a minority sport these days, but that still means a lot of users.
Emacs has a fairly sharp learning curve. But once you have got there it's still one of the quickest editors out there. The user interface is very fast to interact with.
It's also very functional, doing most of what you want and quite a few things that you have never imagined. It's also not an IDE--it's a general purpose editing environment.
Finally, it's very configurable, and all of the files that configure it are transparent. This means that my emacs setup works on different operating systems and I can sync the set up between different machines with the same program that I use for all my synchronisation needs.
I used to spend 90% of my day in emacs. More recently, I've been microsofted; I have to use outlook for email (which is really, really horrible), and write more word docs than latex. I really miss being emacs. My wrists are starting to give in already.
I'm quite pleased with it actually. I seem to have amused and annoyed in equal measure. I wish I'd see it as a +5 troll though. I'm down to +2 flamebait now.
I guess the point is that many people are finding this story highly amusing and feel that it will give DRM a bad name. But there is another interpretation as well, which might well be the one that ends up happening.
It is clear that DRM software is going to be as open to bugs as any other software, and some of these will constitute a security threat.
Surely the solution is obvious. If they built DRM software directly into the operating system, then it could be happily updated with all the rest of the software, using whatever update mechanisms your OS provides.
I'm sure that the security minded folks on slashdot will be the first to support a legal requirement for DRM in all OS'es, so that we can solve this problem before it becomes really serious.
They damn sys admins switched it off. By only supporting one email client and nobbling all the others, they ensure that they have to do the minimum amount of work necessary, while their users have to put up with a crap client.
If only Gnus did exchange servers, I would be such a happy man. Since I've been forced onto an exchange server, nothing has ever been the same again. Click, click, point, crawl through email in outlook that I would have flown through in Gnus.
Left hand drive, right hand drive. Doesn't make any difference to be honest. The theory about automatics is bogus. Australians also prefer automatics.
I think it has more to do with the shape of the roads. In the UK, you go through lots of towns, the roads curve more, they are narrower and harder to overtake. In the US, they are straight for miles at a time. You can overtake without changing gear, which is generally not possible in the UK.
The real reason the the US drives on the right is probably much the same as the reason they spell different. A deliberate attempt to differentiate themselves from an old colonial power. Perfectly reasonable in some ways,
Actually, it took them about 5 months to make them nearly ready. The "Status" page went to nearly done that quickly and then has sat there for 5 years.
Having just arrived in the US, you wouldn't have a social security number. So no one would give you credit for anything. So you can't get anything which you pay for in arreas.
Of course being unable to get broadband would not be a problem. In the absence of electricity, what would you plug your computer into?
"The answer to this is no, because not all blind people want some impersonal robot. A dog is much nicer as it's alive and can make decisions in the external environment that a robot cannot make."
Most blind people don't use dogs. They use sticks. Quite why anyone would swap a low tech but highly effective device for one the size of a washing machine and costing as much as a car, I can't understand.
Why not have a small device, something like a bluetooth ear piece which tells you where in the shops or airport are? This sort of aid in navigation is generally enough. There is no need for something which also won't bump into walls. Blind people don't walk into walls anyway!
"Bitkeeper nonwithstanding, their argument will still be use the best tool for the job." "most people nowadays (including me, for that matter) will take the practical way over the idealistic way."
The irony for me, is that free software often is the most pragmatic, practical decision. How useful a tool is does include the licensing conditions. If it is illegal to use in the way that I want, then it's not useful.
Linus should know this. When SCO were busy with their tricks many potential users of linux were saying "hmm, this is a bit of a worry".
BitKeeper was not the best available tool for the job. For a version control system it's future proofing is critical. With a license that could be revoked at any time, it did not appear at any time in the past to be future proof. So it has proved to be now.
This is incorrect. Journals like PLOS are open access. But they are also tightly peer reviewed and publication ready.
This works because they charge the authors for publications, as do many closed access journals. PLOS do not have to support the expensive business of managing subscriptions, however, so they are no more expensive than any one else.
In fact, all of the open access journals I know are peer reviewed. All that we are talking about here in the discussion of "camera ready" is the final presentation format, not the science.
"I don't understand how you can claim to work in academic publishing and you don't even seem to know what LaTeX is"
This is fair enough. In computer science where I work everyone knows what latex is, and many people use it. In biological sciences, where I also work, people only use latex for making surgical gloves.
(Previous poster said) "Normally you submit your article text in a fairly non-structured form, and submit your graphics and charts separately in a different format."
Even more amusingly, some journals actually have latex style files which produce this unstructured author copy.
As a reviewer, it's really annoying. Text formatted in this way is really hard to read. You end up juggling text and figures. In my experience, this is getting rarer nowadays. Thank heavens for small mercies.
This will be on of the interesting things with open access publication. PLOS is going for a fairly straight "we look like a journal" publication style. BioMedCentral is not; there articles don't have page numbers but use DOI's instead. The world is changing. It's not just in open access that scientific publishing needs to catch up.
"In the biological sciences (where I work), as well as the chemical and ecological sciences (in which I have close colleagues), you are completely incorrect. "
I'm a cross disciplinarian. So I publish in computer science (where camera ready is 95% the rule) and biology (where it is not).
"Preparing a paper for print requires professionals that have training and experience at their jobs. Someday you should take a look at a pile of grant applications, which tend to represent the best writing most scientists can do."
Actually, my father was a typesetter in his time. I agree, that many articles look much better when professionally set. None the less, this process does not happen for many articles.
I agree with your point about grants, although grants and papers are different things. The former do not hang around to haunt you, and so the presentational standard is lower. This is true in computer science also, where 95% of your papers are produced camera ready.
"Who reads and evaluates the 50-60% of papers that are not even sent to peer review, but are rejected and returned to the authors?"
That's only really true for the big journals. For smaller journals and conferences everything gets sent out to program committee or editorial board. Some of the complete crap that I have been sent to review in my time is so bad it's amusing. Most of it is just bad.
"Who finds reviewers in the first place? Is there some magic list somewhere? Who vets their qualifications? "
You get an editor. They know the field. Alternatively, you just look for people who have previously successfully submitted papers. In short, the magic list is pubmed. I don't edit any journals, but I am chair of a conference. This is the way we do it. It's not that hard.
Besides I am not disagreeing with anything you say. Journals cost money. They cost time and effort. But the publishers currently make vast profits from the process and then use restrictive copyright laws to prevent the scientists who produce the work from actually using it. As a bioinformatician I really am capable of using the full text of several million journal papers. There is no way I can do this if there is even a micropayment for access.
I am happy for publishers to still exist. But not for them to have control over the main representation of scientific results. These should be given out freely. More fool us, as those who generate the results for giving these away.
Slashdot Mirror
Ah. Just read the article. It appears that the prize was the idea of the one J. Craig Venter. My initial impression that it is all the work of self-serving, egotistical, self-publicists seems to have been born out.
Phil
Because the people who are attempting to do this already, could almost
certainly not care less about 5million dollars either way. If you the X foundation
were venture capital offering 5 million dollars no one would have noticed either
way. Or better still a philanthropic research charity. They should try sponsering
research, rather than living in it's reflected glory.
Phil
You can not sequence a genome this cheap. The reagents cost more than this. The point is that it would require new technology, new chemistry (or better, no chemistry). There are at least 4 or 5 methods on the horizon though.
Phil
Yes, of course, it would cost much more than this. 5 million dollars is nothing at
all in R&D. People are very expensive. And, unlike human space travel, there are
quite a few fairly immediate applications for this technology and lots of people
who will pay for it.
It's great advertising for the X foundation though. Someone else does all the work,
they get to appear visionary and it only costs 5 million. Pretty clever.
Phil
It's a question of how evolution is questioned.
I have no problems with the idea that evolution should be open to question, and investigation. It is, after all, what evolutionary biologists do for a living.
What I find irritating is "evolution is wrong, it says so in the bible". When faced with this, I have no real problem in ridicule. There is, after all, no mechanism for arguing against it.
Phil
No it wasn't.
Most of the hereditary element was abolished, but some of it is still there.
Phil
"I'm against all public education systems. I don't believe they've worked."
How interesting. I guess from this statement that you've made some
efforts to look at "all public education systems". As it happens,
the public education system in the UK is pretty good.
"That being said, if we must have them, let's focus on pure education -- facts, repetition, useful classes: how to read, write and perform basic math. At most, some basic scientific theory might be OK."
This is not "pure" education, it's a lack of education. If we focus
on "facts", then this leaves us almost nothing to actually teach.
Even reading is pretty useless, unless you have something to read about.
As for expecting parents to teach, I really can't see your point.
It's like suggesting that we should expect parents to provide health
care. We have professionalised health workers, why not professional
teachers as well. If parents want to teach, of course, they can.
Phil
"What amazes me as a Swede is that all Anglo-saxon countries I've been to build so incredibly flimsy and energy-inefficient houses. England, Australia, and from what I've heard, the US as well. I mean, you are rich countries, why build like third world?"
It depends where in the country you are. Quite a lot of our cities have very integrated
buildings--Edinburgh and Glasgow are obvious examples--where the buildings are 100+ years old.
Are really really cold. The windows don't fit properly and there are no cavity walls. My
current flat in Newcastle which is an old industrial terrace suffers from the same problem.
Fixing the problem would, pretty much, require knocking down the entire street.
Newer builds are much better though, with roof and wall insulation. My parents house, for
example, is probably three times the inner area of mine and has heating costs about 1/3 mine.
I have to open the window there at night to cool the bedroom to a comfortable temperature,
even in winter when it's -5 outside.
Of course, in Sweden, you don't have this problem; after all, the winters there are cold
enough, that anything older than 10 years has freeze fractured to the point of falling
down.
More seriously, though, it is very mild in Britain. It doesn't get hot in summer nor really
cold in winter.
Rains a lot though; check out our damp proof coursings if you want to see expert enginneering.
Phil
It's probably a little bit more than that, as it's managed. Also, it's doubling in
size every 10 months; a problem as the rate of increase of hard drive size is
something like once every 18 months. This means the cost of providing this storage
will increase exponentially.
Incidentally, it's not "genetic data"--that is sequence data. It's trace data
which is then interpreted to produce sequence data. So actually, the data storage
requirements for each base takes more than 2 bits. Moreover it's redundant (DNA is
sequencing at least 10 times). So you're probably several orders of magnitude out
with your calculations.
Phil
"I have to wonder how many of these entries are quasi-redundant."
All of them, pretty much. This is a trace archive. It stores the traces as they
come of the sequencing machine. Given that DNA is normally sequenced to 10x
(five times in both directions), most of the data in this database will be
replicates.
"Sadly, the trend here seems to be more of 'sequence it, upload it, and patent it' instead of 'sequence it, upload it, figure out what it does/makes, do something useful with it'."
Data collection is the bed rock of all good science. You can sit and think
of clever hypotheses all that you like, but it's all junk unless you actually
have the data to test it against.
People are already doing interesting things with this sort of trace data that
were not thought of when it first came out (predication of polymorphisms
is possible for instance). More will come over time.
Gather, storing and archiving data is vital. The trace archive is a modern day
Library of Alexandria.
Phil
"By "real" job, I mean a job in which you are required to meet deadlines imposed by customers and to produce end products specified by customers, otherwise your business fails."
RMS is no doubt aware of the pressures that people are under. As he says, if you
send someone a word doc, you are pressuring them to give up their freedom. Which
is true enough. If you are working in the environment then you may indeed end up
giving up some of your freedom.
Most of us are willing to make these compromises, because they make the rest of
our life easier. Stallman has remade his life so that he does not.
"Frankly, if I was running a business and I had to keep my family fed, there's no way in hell I'd be relying on GNU to develop the tools I needed. Their track record is simply appalling."
You give two examples, XEmacs and Hurd. One of which produced some perfectly usable software in common usage--the fork is unfortunate, but there you have it.
Hurd you are right about, of course. Two examples? The rate of failed software projects in industry is at least 50% -- that is 50% of software developed never gets used. Most of it disappoints the customers. Pretty much all of it breaks at one stage or another.
In this light, your two examples fail to convince me that Gnu's "track record is appalling".
Besides, if you read the interview, he wasn't suggesting that you donate cash to Gnu. It is not, and never has been, a software house. If you ask them, "will you write this piece of software that I really need", then they are likely to say no.
There are many other programmers who will do this for you.
Phil
Actually, I think that you are missing the major advantage
that Eclipse has over Emacs, which is that it's written in Java. While
I like lisp, and think it's a very nice language, there is a problem
with library support (for Emacs-Lisp, rather than Lisp in general). Writing
packages for Emacs is quite hard. When you get down to it, this means that
the rate at which new and usable packages come out in slower than it
needs to be. And there are fewer people who are able to fix the bugs with
it.
Having said that, the internet is a large place, and people do keep on
writing packages for it. Look at something like semantic, or ECB, or muse
or auctex. These are large packages, many of which are less than 4 years old.
I think that Emacs will exist for quite a while yet. It may be a minority
sport these days, but that still means a lot of users.
Phil
Emacs has a fairly sharp learning curve. But once you have got there
it's still one of the quickest editors out there. The user interface
is very fast to interact with.
It's also very functional, doing most of what you want and quite a few
things that you have never imagined. It's also not an IDE--it's a general
purpose editing environment.
Finally, it's very configurable, and all of the files that configure
it are transparent. This means that my emacs setup works on different
operating systems and I can sync the set up between different machines with
the same program that I use for all my synchronisation needs.
I used to spend 90% of my day in emacs. More recently, I've been microsofted;
I have to use outlook for email (which is really, really horrible), and write
more word docs than latex. I really miss being emacs. My wrists are starting
to give in already.
Phil
I'm quite pleased with it actually. I seem to have amused and
annoyed in equal measure. I wish I'd see it as a +5 troll
though. I'm down to +2 flamebait now.
I guess the point is that many people are finding this story
highly amusing and feel that it will give DRM a bad name. But
there is another interpretation as well, which might well be
the one that ends up happening.
Phil
It is clear that DRM software is going to be as open to bugs as any other
software, and some of these will constitute a security threat.
Surely the solution is obvious. If they built DRM software directly into the
operating system, then it could be happily updated with all the rest of the
software, using whatever update mechanisms your OS provides.
I'm sure that the security minded folks on slashdot will be the first to
support a legal requirement for DRM in all OS'es, so that we can solve this
problem before it becomes really serious.
Phil
They damn sys admins switched it off. By only supporting
one email client and nobbling all the others, they ensure that
they have to do the minimum amount of work necessary, while
their users have to put up with a crap client.
Great plan. For them.
Phil
If only Gnus did exchange servers, I would be such a happy man. Since
I've been forced onto an exchange server, nothing has ever been the
same again. Click, click, point, crawl through email in outlook that I would
have flown through in Gnus.
I really miss it.
Phil
Left hand drive, right hand drive. Doesn't make any difference to be honest. The theory about automatics is bogus. Australians also prefer automatics.
I think it has more to do with the shape of the roads. In the UK, you go through lots of towns, the roads curve more, they are narrower and harder to overtake. In the US, they are straight for miles at a time. You can overtake without changing gear, which is generally not possible in the UK.
The real reason the the US drives on the right is probably much the same as the reason they spell different. A deliberate attempt to differentiate themselves from an old colonial power. Perfectly reasonable in some ways,
Actually, it took them about 5 months to make them nearly ready. The "Status" page went to nearly done that quickly and then has sat there for 5 years.
Very odd.
Phil
In the US, this wouldn't be a problem.
Having just arrived in the US, you wouldn't have a social security number. So no one would give you credit for anything. So you can't get anything which you pay for in arreas.
Of course being unable to get broadband would not be a problem. In the absence of electricity, what would you plug your computer into?
Phil
"The answer to this is no, because not all blind people want some impersonal robot. A dog is much nicer as it's alive and can make decisions in the external environment that a robot cannot make."
Most blind people don't use dogs. They use sticks. Quite why anyone would swap a low tech but highly effective device for one the size of a washing machine and costing as much as a car, I can't understand.
Why not have a small device, something like a bluetooth ear piece which tells you where in the shops or airport are? This sort of aid in navigation is generally enough. There is no need for something which also won't bump into walls. Blind people don't walk into walls anyway!
Phil
"Bitkeeper nonwithstanding, their argument will still be use the best tool for the job."
"most people nowadays (including me, for that matter) will take the practical way over the idealistic way."
The irony for me, is that free software often is the most pragmatic, practical decision. How useful a tool is does include the licensing conditions. If it is illegal to use in the way that I want, then it's not useful.
Linus should know this. When SCO were busy with their tricks many potential users of linux were saying "hmm, this is a bit of a worry".
BitKeeper was not the best available tool for the job. For a version control system it's future proofing is critical. With a license that could be revoked at any time, it did not appear at any time in the past to be future proof. So it has proved to be now.
Phil
This is incorrect. Journals like PLOS are open access. But they are also tightly peer reviewed and publication ready.
This works because they charge the authors for publications, as do many closed access journals. PLOS do not have to support the expensive business of managing subscriptions, however, so they are no more expensive than any one else.
In fact, all of the open access journals I know are peer reviewed. All that we are talking about here in the discussion of "camera ready" is the final presentation format, not the science.
Phil
"I don't understand how you can claim to work in academic publishing and you don't even seem to know what LaTeX is"
This is fair enough. In computer science where I work everyone knows what latex is, and many people use it. In biological sciences, where I also work, people only use latex for making surgical gloves.
(Previous poster said)
"Normally you submit your article text in a fairly non-structured form, and submit your graphics and charts separately in a different format."
Even more amusingly, some journals actually have latex style files which produce this unstructured author copy.
As a reviewer, it's really annoying. Text formatted in this way is really hard to read. You end up juggling text and figures. In my experience, this is getting rarer nowadays. Thank heavens for small mercies.
This will be on of the interesting things with open access publication. PLOS is going for a fairly straight "we look like a journal" publication style. BioMedCentral is not; there articles don't have page numbers but use DOI's instead. The world is changing. It's not just in open access that scientific publishing needs to catch up.
Phil
"In the biological sciences (where I work), as well as the chemical and ecological sciences (in which I have close colleagues), you are completely incorrect. "
I'm a cross disciplinarian. So I publish in computer science (where camera ready is 95% the rule) and biology (where it is not).
"Preparing a paper for print requires professionals that have training and experience at their jobs. Someday you should take a look at a pile of grant applications, which tend to represent the best writing most scientists can do."
Actually, my father was a typesetter in his time. I agree, that many articles look much better when professionally set. None the less, this process does not happen for many articles.
I agree with your point about grants, although grants and papers are different things. The former do not hang around to haunt you, and so the presentational standard is lower. This is true in computer science also, where 95% of your papers are produced camera ready.
"Who reads and evaluates the 50-60% of papers that are not even sent to peer review, but are rejected and returned to the authors?"
That's only really true for the big journals. For smaller journals and conferences everything gets sent out to program committee or editorial board. Some of the complete crap that I have been sent to review in my time is so bad it's amusing. Most of it is just bad.
"Who finds reviewers in the first place? Is there some magic list somewhere? Who vets their qualifications? "
You get an editor. They know the field. Alternatively, you just look for people who have previously successfully submitted papers. In short, the magic list is pubmed. I don't edit any journals, but I am chair of a conference. This is the way we do it. It's not that hard.
Besides I am not disagreeing with anything you say. Journals cost money. They cost time and effort. But the publishers currently make vast profits from the process and then use restrictive copyright laws to prevent the scientists who produce the work from actually using it. As a bioinformatician I really am capable of using the full text of several million journal papers. There is no way I can do this if there is even a micropayment for access.
I am happy for publishers to still exist. But not for them to have control over the main representation of scientific results. These should be given out freely. More fool us, as those who generate the results for giving these away.
Phil