Ah I see I'm guilty of wrongthink. Pointing out clearly specious arguments is "defending authority". I shal report myself to your SJW commission for reeducation shotly, comrade.
Yes, and? Peer review is one of the pillars of science and sometimes it gives bad results. Often in fact. This isn't news.
like deciding whether hundreds of millions of NSF dollars or billions of NASA or NOAA dollars should go to pay for a new weather satellite, or probe to Jupiter.
Er what? When has a hnudred million dollars been spent on the NSF on the basis of a feshly published piece of research that hasn't stood any test of time?
Imagine if we predicated the procurerment of GPS satellites or weather satellites on the decisions of people who have no problem letting a flat earth paper through the door in the same journal
Most of the things GPS is based on like special relativity, precision clocks, kalman filters, RF engineering, DSSS, and so on were well past the point of being mere journal publications. Those were all things which started that way but were known to work though repeated verification. GPS was primarily an engineering challenge, not a scientific one.
Now imagine if we were to entrust our legal system or medical system to people who think men can get pregnant and women are on average just as fit to be loggers and firefighters as men, with any difference automatically being proof of criminal levels of harassment
Imagine we entrust our legal system to an army of evil cloned commie-nazi elves!
I mean sure we can imagine hypotheticals about things that don't exist, but eally what's the point in this case?
Well you've got a bunch of headless chickens running ound here shitting themselves because they don't seem to understand it.
Then they should take the occasional prank in stride.
some dickheads waste a bunch of people's time re-proving a well known fact. Lol your time was wasted laugh you should have a send of humour! lolz
It's an uninformative prank. And academia does take it in its stride: it will continue just fine.
The whole argument against him
He faked a bnuch of data, put in a huge amount of work and got published. He's now axe ginding about how the whole field is broken, despite targeting a well known defect of peer review that covers his field of study as well.
He's a time-wasting hypocrite who has added nothing else new to the discussion. That's the criticism of him.
Yup. Heck, there's no corner case of crazy that hasn't happened somewhere in the US university system recently, and some schools have political indoctrination mixed in with their STEM classes now.
Do you have a link?
If it's happened the one time, well, that's a lot fewer times than lecture hall shootings. Clearly gun ownership is way more statistically important to worry about.
Well ok, if you find that failure rate acceptable, I can't argue.
Well you clearly are. Except instead of cogent arguments, you're just flinging lightly disguised shit while claiming to take the moral high ground.
Peer review is not robust to fraud because no one has figued out a practical way of making it so.
I simply expected better from people who claim to know so much more than the rest of us.
Well, then frankly you're an idiot. You no nothing about how it works, and have never engaged in it, yet expect it to meet some arbitrary high standard that just popped into you head. Tell you what propose something better, rather than just smugly sitting on a high horse implying you're oh-so-smart.
Bonus points if whatever you propose is even vaguely practical. I won't hold my breath, because there's an approximately 0% chance that what you propose is workable.
And no having every review replicate the original results is not practical, for the same reason that you don't go and attempt to replicate published results.
No one except idiots with an axe to grind have ever made such a claim. Peer review checks if a paper passes basic smell tests, not that it's 100% correct and not that it's fraudlent.
The only thing that checks its real corectness is ongoing scientific inquiry after publication.
No, it's like getting a flat earth paper published in Nature.
Nature has had its fair share of utter shit published. See for example Hendrik Schoen, for actual fraud. And it has an uh slight peference for fashion to put it mildly and a real bias against computer based disciplines (they are really bad at picking competent reviewes when a paper has a heavy computational component, generally they only pick people to review the science based parts) and so on.
Peer review is awful and full of holes. And it's easy to fool. And it's completely overloaded with too many papes for the number of reviewers. It's particularly suceptible to taked papes.
Yet the world keeps turning and papers get published and science measurably advance.
Yes, it does. This is how you expose fraud in the system. You have a better way?
What fraud? They got papers in by falsifying data. News flash: everyone who actually knows about peer review already knows it's not robust to false data.
It's a general (and somewhat noisy) check to see if the methods sound legit and to see if the conclusions ae warranted by the data. And as it was they managed only a moderate hit rate with 5 out of 18 being accepted at legitimate journals.
Notive how they use that and the axe-grindes over here to declae the entire field is broken rather than just those 5 jounals.
"No amount of experimentation can ever prove me right; a single experiment can prove me wrong." - Albert Einstein Edge cases have a way of doing that.
Think, for example, of Black Body Radiation.
That would work if the agument was that univesities and the academic process is perfect. Thing is no one is arguing that. Few people would even ague that pee review is robust to malice (such as faked data).
Nonetheless it is the best system we have by a very long way.
Are you under the impression that the Earth is infinite?/em.
We ae a long way from using up the entie Earth. What is going on in our lifetimes and many after is not a zero sum game. the wold is moe complex than a simple asymptotic analysis.
As if the left hasn't operated like they've at least had the moral imperative if not the constitutional right up until this point.
Well, then the left understand the reasons behind the first amendment then, and if you haven't realised why then you really really don't.
Here's a partisan-free clue for you. Do not treat the party you voted for as "your team" to win no matter the costs. Do not treat the party you voted for as beyond reproach. If you're naive enough to think that the person you voted for is perfect then for the good of everyone please don't vote next time.
You have effectively two choices to vote for. Neither of them is going to represent you on everything across the board. Neither is going to stick to their promises. Neither is even going to hold a terribly coherent philosophical position. Both will listen to special interest groups of some sort. Neither will be especially well informed.
So yes, you should hold your representative to account. And sometimes that means you will wind them up enough to get a response.
IOW you have a much stronger moral imperative to troll whoever is currently the president than you do to wave a political party flag. How come you didn't realise this already? What do you think the first amendment is for if not saying stuff that will piss off those on power?
Oh definitely. It's not overpriced and unreliable by nature, that's all down to the way its being run currently, ie badly. If it works for you though then all the better.
It works for more than just me. Waterloo alone works for 100,000,000 people per year. Imagine trying to get that many cars through.
I've heard of moving goalposts. But you could at least keep them in the same stadium.
Quite, so I'd appreciate it if you came back to the same stadium that I started it. It's hilarious that you're accusing me of moving the goalposts when you're the one banging on about designing from scratch. Here's a very brief summary of the post histrory:
What I originally said was this:
A modern cheapass class D amp outperforms probably 99% of systems if not more from that era.
in the context of the comment:
You forgot how crappy sound systems were in the old days.
Now you're insisting that we can only discuss designing amplifiers entirely from scratch? WTF how on earth does that have any bearing on sound systems. A clue: it doesn't. This whole "desgning from scratch" thing is something you brought to the thread. If you're building a sound system and designing everything from scratch the chances are you're doing it wrong.
And speaking of moving the goalposts:
Don't worry though, actual high end companies know which is why they design a unique amplifier for each driver on their speakers with different switching characteristics, different filtering, different feedback loops etc.
How is restricting the discussion to only the very highest of the high end remorely relevant to my original statement about the 99% of systems?
Basically you've taken a of mine about sound common systems, disagreed and to "win", you're sticking to designing only the very highest end systems from discretes. In other words the only arguments you can make are utterly irrelevant to my original claim.
I didn't say 4-6 discrete components.
You said "simple". You semeed to say shoving a long tailed pair on the input to biased complementary MOSFETs on the output. You can do tht crapily in 4-6 components. But you said simply, right? You kept on saying it, in fact, so I assumed you meant it. To build a good analogue amp is as you're demonstrating, not simple.
In fact there's more components than that in just a long-tailed pair input stage, and more in a decent CCS. It's quite clear you have no idea what you're talking about if you somehow think I said 4-6 component, or for some reason think that what I listed isn't the fundamental setup of every amplifier even you fancy $15k monoblocks.
Shows how little you know if you think it's fundamental to every amp. You can go from I2S to PWM or delta-sigma without every touching an analogue component. Why pick up noise in an analogue state when you can go digital right to the filter just before the speaker?
I did nothing of the sort. You never specified small or cool in your original design criteria.
Now you're just desparate. I did not specify small or cool, but if you're building a class A, then it's so inefficient that you have to put the heat somewhere. And you forgot to talk about the heatsinks which YOU will need in order for your design to work without literally burning itself up.
But apparently all this is "simple" while a software sigma-selta modulator is complex. Right-o
I don't think you know what that means, and if you're talking about high PSRR typical designs based on the blocks I mentioned will quite happily have >90dB PSRR which is also why when you rip apart a Class A amplifier you won't find a voltage regulator anywhere. A simple CRC filter and the resulting system noise is happily down in the irrelevant regime.
Now you need to design a CRC filter o the PSU? Your simple design is getting really rather complex, I must say. I think I'll stick with a chip from TI.
Oh no I do. You don't realise just how difficult it is to make a switching system work well across the audio bandwidth.
3HMz is sufficient. That's not especially onerous. It's not like youre into transmission lines, ferrite magic and 4 terminal transistors.
Now if you'll excuse me I'm going to Bestbuy to "design myself" an amplifier.
If you're designing an amp from discretes when one from best buy meets your customer's spec then you're doing a crappy job as an engineer.
Yeah and the rail network is overpriced and unreliable.
Yes, it is indeed overpriced and unreliable but it's still infinitely better than not having one. For a start, the road capacity does not and pretty much cannot exist to replace the rail network, and that's before parking is taken into account. And the trains run faster too for a lot of journeys, both long distance (due to the train speed) and urban (due to the heavy road traffic).
I use the trains a lot. I can also drive (and preversely quite enjoy it as long as I'm not on a motorway, as long as I don't have to do it too often) and have many cars and vans I can use within walking distance. I generally choose t ouse the overpriced, unreliable train network in preference because a deeply imprefect system can still be worlds better than a different system.
A large amount of the problem is that parliament is incompetent. Despite the evidence, they are convinced that repeatedly franchising it out will work better even though whenever a franchise collapses and the civil service takes over, reliablility goes up. But it's quasi-religious fanatacism on the part of our elected reprehensibles.
Slashdot Mirror
Thanks man, that's the nicest thing anyone has said to me all year.
Waiittt it's only 8 days in!
Is this really a thing?
Is what a thing? Not quoting accuratly apparently...
Sorry, you're just defending authority.
Ah I see I'm guilty of wrongthink. Pointing out clearly specious arguments is "defending authority". I shal report myself to your SJW commission for reeducation shotly, comrade.
Except when it is consequential,
Yes, and? Peer review is one of the pillars of science and sometimes it gives bad results. Often in fact. This isn't news.
like deciding whether hundreds of millions of NSF dollars or billions of NASA or NOAA dollars should go to pay for a new weather satellite, or probe to Jupiter.
Er what? When has a hnudred million dollars been spent on the NSF on the basis of a feshly published piece of research that hasn't stood any test of time?
Imagine if we predicated the procurerment of GPS satellites or weather satellites on the decisions of people who have no problem letting a flat earth paper through the door in the same journal
Most of the things GPS is based on like special relativity, precision clocks, kalman filters, RF engineering, DSSS, and so on were well past the point of being mere journal publications. Those were all things which started that way but were known to work though repeated verification. GPS was primarily an engineering challenge, not a scientific one.
Now imagine if we were to entrust our legal system or medical system to people who think men can get pregnant and women are on average just as fit to be loggers and firefighters as men, with any difference automatically being proof of criminal levels of harassment
Imagine we entrust our legal system to an army of evil cloned commie-nazi elves!
I mean sure we can imagine hypotheticals about things that don't exist, but eally what's the point in this case?
You got me. What kind of asshat brings FACTS into these things?
A liberal progessive SJW that's who.
So what?
Well you've got a bunch of headless chickens running ound here shitting themselves because they don't seem to understand it.
Then they should take the occasional prank in stride.
some dickheads waste a bunch of people's time re-proving a well known fact. Lol your time was wasted laugh you should have a send of humour! lolz
It's an uninformative prank. And academia does take it in its stride: it will continue just fine.
The whole argument against him
He faked a bnuch of data, put in a huge amount of work and got published. He's now axe ginding about how the whole field is broken, despite targeting a well known defect of peer review that covers his field of study as well.
He's a time-wasting hypocrite who has added nothing else new to the discussion. That's the criticism of him.
Yup. Heck, there's no corner case of crazy that hasn't happened somewhere in the US university system recently, and some schools have political indoctrination mixed in with their STEM classes now.
Do you have a link?
If it's happened the one time, well, that's a lot fewer times than lecture hall shootings. Clearly gun ownership is way more statistically important to worry about.
And in why the down votes too. DO NOT CONTRADICT THE NARRATIVE!
If you quote actual bits from TFA (and not out of context either), that conuts as trolling.
Well ok, if you find that failure rate acceptable, I can't argue.
Well you clearly are. Except instead of cogent arguments, you're just flinging lightly disguised shit while claiming to take the moral high ground.
Peer review is not robust to fraud because no one has figued out a practical way of making it so.
I simply expected better from people who claim to know so much more than the rest of us.
Well, then frankly you're an idiot. You no nothing about how it works, and have never engaged in it, yet expect it to meet some arbitrary high standard that just popped into you head. Tell you what propose something better, rather than just smugly sitting on a high horse implying you're oh-so-smart.
Bonus points if whatever you propose is even vaguely practical. I won't hold my breath, because there's an approximately 0% chance that what you propose is workable.
And no having every review replicate the original results is not practical, for the same reason that you don't go and attempt to replicate published results.
Well played, sir!
[first time that came out as Well played si! I think my keyboard might need some wok with an ai duste]
Bullshit, that what a "review" is supposed to do.
Bullshit! No it's not.
No one except idiots with an axe to grind have ever made such a claim. Peer review checks if a paper passes basic smell tests, not that it's 100% correct and not that it's fraudlent.
The only thing that checks its real corectness is ongoing scientific inquiry after publication.
No, it's like getting a flat earth paper published in Nature.
Nature has had its fair share of utter shit published. See for example Hendrik Schoen, for actual fraud. And it has an uh slight peference for fashion to put it mildly and a real bias against computer based disciplines (they are really bad at picking competent reviewes when a paper has a heavy computational component, generally they only pick people to review the science based parts) and so on.
Peer review is awful and full of holes. And it's easy to fool. And it's completely overloaded with too many papes for the number of reviewers. It's particularly suceptible to taked papes.
Yet the world keeps turning and papers get published and science measurably advance.
And when he finally managed to get 7 out of 20 through, he needed both funding and 90 hour weeks working on it:
It's pointed out late that 2 of those do not seem like legit journals, making it moe like 5 out of 18.
Yes, it does. This is how you expose fraud in the system. You have a better way?
What fraud? They got papers in by falsifying data. News flash: everyone who actually knows about peer review already knows it's not robust to false data.
It's a general (and somewhat noisy) check to see if the methods sound legit and to see if the conclusions ae warranted by the data. And as it was they managed only a moderate hit rate with 5 out of 18 being accepted at legitimate journals.
Notive how they use that and the axe-grindes over here to declae the entire field is broken rather than just those 5 jounals.
It is a problem if you go to a physics class, and get a lecture from a women's studies class instead
Has that ever actually happened?
"No amount of experimentation can ever prove me right; a single experiment can prove me wrong." - Albert Einstein Edge cases have a way of doing that.
Think, for example, of Black Body Radiation.
That would work if the agument was that univesities and the academic process is perfect. Thing is no one is arguing that. Few people would even ague that pee review is robust to malice (such as faked data).
Nonetheless it is the best system we have by a very long way.
That applies to ANY SJW.
Wait it applies to you too?
Are you under the impression that the Earth is infinite?/em.
We ae a long way from using up the entie Earth. What is going on in our lifetimes and many after is not a zero sum game. the wold is moe complex than a simple asymptotic analysis.
the traditional media outlets shut down comments once they realized that people were allowed to post differing viewpoints
Well, yes it is so very you to believe that aggressivly spammy shitposting counts as a "differing viewpoint".
As if the left hasn't operated like they've at least had the moral imperative if not the constitutional right up until this point.
Well, then the left understand the reasons behind the first amendment then, and if you haven't realised why then you really really don't.
Here's a partisan-free clue for you. Do not treat the party you voted for as "your team" to win no matter the costs. Do not treat the party you voted for as beyond reproach. If you're naive enough to think that the person you voted for is perfect then for the good of everyone please don't vote next time.
You have effectively two choices to vote for. Neither of them is going to represent you on everything across the board. Neither is going to stick to their promises. Neither is even going to hold a terribly coherent philosophical position. Both will listen to special interest groups of some sort. Neither will be especially well informed.
So yes, you should hold your representative to account. And sometimes that means you will wind them up enough to get a response.
IOW you have a much stronger moral imperative to troll whoever is currently the president than you do to wave a political party flag. How come you didn't realise this already? What do you think the first amendment is for if not saying stuff that will piss off those on power?
Oh definitely. It's not overpriced and unreliable by nature, that's all down to the way its being run currently, ie badly. If it works for you though then all the better.
It works for more than just me. Waterloo alone works for 100,000,000 people per year. Imagine trying to get that many cars through.
It could work substantially better though.
I've heard of moving goalposts. But you could at least keep them in the same stadium.
Quite, so I'd appreciate it if you came back to the same stadium that I started it. It's hilarious that you're accusing me of moving the goalposts when you're the one banging on about designing from scratch. Here's a very brief summary of the post histrory:
What I originally said was this:
A modern cheapass class D amp outperforms probably 99% of systems if not more from that era.
in the context of the comment:
You forgot how crappy sound systems were in the old days.
Now you're insisting that we can only discuss designing amplifiers entirely from scratch? WTF how on earth does that have any bearing on sound systems. A clue: it doesn't. This whole "desgning from scratch" thing is something you brought to the thread. If you're building a sound system and designing everything from scratch the chances are you're doing it wrong.
And speaking of moving the goalposts:
Don't worry though, actual high end companies know which is why they design a unique amplifier for each driver on their speakers with different switching characteristics, different filtering, different feedback loops etc.
How is restricting the discussion to only the very highest of the high end remorely relevant to my original statement about the 99% of systems?
Basically you've taken a of mine about sound common systems, disagreed and to "win", you're sticking to designing only the very highest end systems from discretes. In other words the only arguments you can make are utterly irrelevant to my original claim.
I didn't say 4-6 discrete components.
You said "simple". You semeed to say shoving a long tailed pair on the input to biased complementary MOSFETs on the output. You can do tht crapily in 4-6 components. But you said simply, right? You kept on saying it, in fact, so I assumed you meant it. To build a good analogue amp is as you're demonstrating, not simple.
In fact there's more components than that in just a long-tailed pair input stage, and more in a decent CCS. It's quite clear you have no idea what you're talking about if you somehow think I said 4-6 component, or for some reason think that what I listed isn't the fundamental setup of every amplifier even you fancy $15k monoblocks.
Shows how little you know if you think it's fundamental to every amp. You can go from I2S to PWM or delta-sigma without every touching an analogue component. Why pick up noise in an analogue state when you can go digital right to the filter just before the speaker?
I did nothing of the sort. You never specified small or cool in your original design criteria.
Now you're just desparate. I did not specify small or cool, but if you're building a class A, then it's so inefficient that you have to put the heat somewhere. And you forgot to talk about the heatsinks which YOU will need in order for your design to work without literally burning itself up.
But apparently all this is "simple" while a software sigma-selta modulator is complex. Right-o
I don't think you know what that means, and if you're talking about high PSRR typical designs based on the blocks I mentioned will quite happily have >90dB PSRR which is also why when you rip apart a Class A amplifier you won't find a voltage regulator anywhere. A simple CRC filter and the resulting system noise is happily down in the irrelevant regime.
Now you need to design a CRC filter o the PSU? Your simple design is getting really rather complex, I must say. I think I'll stick with a chip from TI.
Oh no I do. You don't realise just how difficult it is to make a switching system work well across the audio bandwidth.
3HMz is sufficient. That's not especially onerous. It's not like youre into transmission lines, ferrite magic and 4 terminal transistors.
Now if you'll excuse me I'm going to Bestbuy to "design myself" an amplifier.
If you're designing an amp from discretes when one from best buy meets your customer's spec then you're doing a crappy job as an engineer.
Yeah and the rail network is overpriced and unreliable.
Yes, it is indeed overpriced and unreliable but it's still infinitely better than not having one. For a start, the road capacity does not and pretty much cannot exist to replace the rail network, and that's before parking is taken into account. And the trains run faster too for a lot of journeys, both long distance (due to the train speed) and urban (due to the heavy road traffic).
I use the trains a lot. I can also drive (and preversely quite enjoy it as long as I'm not on a motorway, as long as I don't have to do it too often) and have many cars and vans I can use within walking distance. I generally choose t ouse the overpriced, unreliable train network in preference because a deeply imprefect system can still be worlds better than a different system.
A large amount of the problem is that parliament is incompetent. Despite the evidence, they are convinced that repeatedly franchising it out will work better even though whenever a franchise collapses and the civil service takes over, reliablility goes up. But it's quasi-religious fanatacism on the part of our elected reprehensibles.
If the concept of limited resources and zero sum games are alien to you
The world isn't a zero sum game.
dumb buffoon is better than a smart well established crook that is in bed with the media.
On wht grounds is Trump less of a crook than Hillary? He's done fucktons of skeezy shit.