I can understand that, but, if you followed that logic consistently, you'd never buy a computer at all. I bought a 701, which I liked apart from the screen. Then I bought a Windows 900, on which I've installed Kubuntu, and I'm quite happy with it. It's a bit irritating that the next model is out already, but I'll be using mine on a series of train trips next week. If I had done things your way, I'd be reading magazine reviews instead of doing any work...
I think 'revolutionize' is a bit of an overstatement. Most people who are likely to fancy a bit of Linux-based hardware diagnosis can already do that with their favourite boot-from-CD distro, but it's hardly the sort of thing your average Word power user is going to try on the bus on the way to work.
Including an OS on the motherboard makes sense for Asus - at least it is then possible to do basic hardware diagnostics independently of, say, Windows diagnostics.
But, in terms of Linux adoption, it's only exciting if people keep linux once they've finished building the computer, and the precedents here are hardly promising.
And, even if you like Linux (which I do), would you want to keep the version supplied with your m/b? On my first EeePC, I tried to get to like Xandros, I really did, but in the end I wiped it and installed Kubuntu. My Dark Side Brother played with Xandros until he broke it, and then installed XP. And it's going to happen even more with the EeePC 900, since the Linux version has a larger SSD than the Windows version (at least in the UK), so you buy the Linux version in order to install Windows.
Which beam would that be? I own several linux boxes, one Mac and no Windows machines... My point about Vista is that no-one is saying "This is the most wonderful OS ever".
The thing the article and most of the comments miss is that Apple don't sell your a product, they sell you a self-image. Buying a PC isn't a bad choice or bad value for money, it makes you a corporate drone. If you own a Mac, you are a free-thinking, creative, pretty young thing - whatever your age, IQ or silouette. You don't pay Steve Jobs to ship you a computer. You pay him to make you a better person.
So when people criticise a Mac, it hits Mac owners at the same level as comments about their personal appearance. When you say that the AirBook is overpriced, it has the same emotional impact as telling the AirBook's owner to consider plastic surgery.
Actually, price criticism isn't too serious, because it invites the "I care about myself enough to buy the best" defence. The real killer is pointing out that, actually, Apple does far more lock-ins than Microsoft nowadays. Those who bought their self-esteem from Apple read that as "You thought you had found the One True Church, but in reality you joined a sect led by an LA snake-oil salesman."
On that basis, expect Mac Rage to turn thermonuclear if/when the Linux desktop ever gets any sort of market penetration, because even Apple can't make Linux look like an evil empire.
And, yes, Linux geeks do the same sort of thing, except that they do it in a nerdier and cheaper way. TBH, Windows users are probably the least religious group out there right now, especially post-Vista.
Toshiba was found guilty of illegally selling the Soviet Union
Wow, respect! How much did they get for it? With such incredible marketing skills, it seems all the stranger that they couldn't sell the HD-DVD format. I mean, that was legal...
I've lost count of how many times I've seen this rather basic mistake. Perl is compiled before it executes. So there's a compiler overhead with each execution, and that can amount to several seconds if you use loads of libraries, which is why it's best to run non-trivial perl websites using mod_perl. But once it's compiled, it's as compiled as C and rather more compiled than java.
For many text-based tasks, Perl is blindingly fast. The canonical example (from the Camel Book) is implementing grep in perl, which runs faster than some native OS implementations. That's partly because it's optimised for that sort of task, but it's also that it has a ruthlessly efficient implementation of hashes. Sure, you can do better in C if you really try, but if you don't really try a perl nested hash structure will blow C linear search out of the water.
The Camel book acknowledges that perl isn't the fastest option for tight loops, and ray tracing would probably fall into that category. But, more generally, the other thing about perl is that it has an extremely large collection of APIs to C libraries. If I want to parse XML, LibXML/LibXSLT are among the fastest C libraries out there, and those libraries run about as fast when called from perl as they do called from C.
Maybe it's just me but I don't equate an increase in search engine queries regarding programming in python as indicating an increase in the popularity of python.
Indeed. It probably tells you more about the randomness of the documentation. I almost never search for anything connected to perl using a generic search engine, since pretty much everything is on CPAN. OTOH, if your preferred language has a 'right-click' approach to modularisation (which seems to me to be the case with PHP), you spend all your time googling for the right bit of code to right-click.
I'm sorry you are astonished and angry. But your post went
Aren't the benefits of open source, or, generally, hackable hardware very simple to explain?
[explanation that is almost entirely irrelavent to 99% of software users]
There. Was that so difficult?
In future, if you must be patronising, try to be patronising and right. I've explained why I think your entire explanation heads off in the wrong direction, and some of what you say is downright wrong. For example, your claim that open sourcing of code somehow ensures ongoing support of the code is patent nonsense, as a brief look at sourceforge will confirm. There are lots of open source projects out there with an active user base and no developers.
Now, on that point, I'm inclined to agree that open sourcing the code provides better guarantees of ongoing support in most circumstances than closed source code. But posting a gross generalisation and then signing off with "There, was that so difficult" is a way to paint a large target on your butt.
Incidentally, I see from your website that you do write code, so I take back my suggestion to the contrary a few posts up.
A choice you can't afford to make isn't a choice. I understand the benefits of the code being available to be modified. I'm simply pointing out that, in practice, most people have no option but to use open source software as if it couldn't be modified, because they do not have the resources (technical or financial) to modify it. This means the theoretical possibility of them modifying the code isn't any kind of selling point for them.
Also, I think you'll find that a lot of closed source companies are willing to modify or fork their code, for a price. I've seen more than one small software vendor make small fixes, for free, in response to my feedback. OTOH, XFree86 forked after complaints that the project owners weren't as responsive as my small closed-source vendors. The picture is a lot more complex than "Open Good, Closed Bad".
Ok, you win. Open source clearly isn't perfect. Therefore, we should all throw in the towel and go back to closed-source software. After all, since open source software isn't perfect, closed source software is clearly better.
No, we should just stop over-selling open source as if the mere fact that the code is available makes all things possible and solves every problem from bugs to world poverty and acne. Selling points for the general public are more along the lines of long-term availability of the software, a better record on fixing bugs and a culture that encourages interoperability. "You can fix it yourself" isn't a selling point for most people, even if it were true. "This car comes with no warranty, there are no dealers, but you are free to cast your own engine parts when it breaks down, and even to distribute those engine parts to third parties" isn't a sales pitch you are going to see on TV any time soon.
If there is any bug, or desireable feature that is missing, or really any kind of improvement to be made, it can be made by anyone. This includes you, but you don't have to do it yourself - chances are there is somebody who wants the same improvement and will make it and share it with the world.
There ought to be a Slashdot autoresponder for this suggestion. It is not and never has been true of software, and is even less true of hardware. If you really think you can personally add whatever feature you fancy to any and all open-source software in anything like a reasonable timescale, I suspect you've never tried. I'd really like emacs to display all regexes properly (# is a particular problem in both perl and tcl modes). The bug has been around for years, and I'm sure tens of thousands of technical users have noticed. Can you fix that for me by next Wednesday?
If you think it makes sense to even try fixing other people's code in most cases, you are probably still in full-time education. I suspect I could work out a fix for the above problem if I had nothing else to do for a few months, but the reason I use emacs in the first place is because I have other things to do. That's true of the vast majority of software users. The options are to live with the bugs, hold out for a bug fix and find another product that does the same job better.
Cathedral and Bazaar describes (probably, I suspect, in a rather rose-tinted way) the development of fetchmail as a collaborative open source project, but, for heaven's sake, it's part of server-side email infrastructure. The contributors were all übertechies. What proportion of end users of systems that use fetchmail do you think have even heard of fetchmail, let alone being able to improve the source code, let alone persuading whoever handles their mail to install the improved version?
The reality is that modern software is generally extremely complex to maintain, that there's a huge learning curve in unpicking even well-documented projects, that there are politics attached to getting your mods accepted by the main project, and that effectively forking the project in order to add your one new feature is generally the worst thing you could possibly do in terms of long-term reliability. You might possibly be able to fix the bug you've noticed, but making sure you haven't introduced three more bugs in the process is a little more tricky.
That's software. The main thing stopping most of us from hacking our hardware isn't DRM, it's not having the technical competence and the extremely expensive kit to go with it. Heck, it's virtually impossible to do routine maintenance on most cars nowadays without the vendor's test rig.
There are advantages for everyone in open standards, and maybe in open hardware, but the advantages don't come from Grannie Brown being able to turn her PC into a video recorder on a wet Tuesday afternoon. You still need the backing of a fairly large institution (university or company) to do hardware hacking that achieves more than a mention on Slashdot hardware hacks.
Why don't you need one? In the photo on the last/. article the kite isn't dead ahead of the ship, and you'd expect that to be the case most of the time. If that tends to pull the ship off course, don't you end up using the rudder like a rather poor keel, effectively dragging the ship through the sea sideways, and thus wasting a load of energy, not to mention the stress on various parts of the ship?
# Open source technologies rigorously tested and matured through the Red Hat sponsored Fedora project
# With each major version, stable application interfaces and 7 years of product support
I know what binary compatible means. Do you know what 'importance of brand names to non-technical people controlling chequebooks' means? And Fedora isn't Gentoo, but it still smells of hobbyist to customers used to paying $$$ for server software.
Not only does Redhat 'tolerate' CentOS (see above), it also puts money into encouraging people to use Fedora, which is not only free but generally significantly more advanced than RHEL. For people who want free software and enjoy recompiling their kernel, Fedora is a much more obvious choice than a clone of CentOS.
There was never any money in selling distros to dorm-room techies, and RHEL was never a good distribution for that market, because it's so conservative. I run Ubuntu on my desktop machines, because it's free, and it works, and it has all the multimedia stuff that RedHat don't ship as standard. On my company's production servers it's RHEL every time, because it's stable, because it will still be supported in 5 years' time if necessary, and because RHEL is a de facto standard in hosting terms. If a client's code doesn't work with RHEL, we can tell them to fix their code. If we were running some wacky, customised version of Gentoo they'd tell us to fix our server (whether or not anything was broken).
Running CentOS would give us the conservatism of RHEL without any of the respectability. I can't see how that would be useful to us.
Great, so now I get the legendary security of IISS coupled with the legendary security of PHP? To be fair, it shouldn't be much worse than PHP/Apache, since the most common PHP setup effectively disables user-based access control and lets any PHP script anywhere on the server access any data created by any other PHP script anywhere on the server.
I do hope they've kept that thing where you install a new PHP application and then try to run the admin.php script via the web server as nobody before anyone else on the Internet does. That's a race condition implemented with real vision if ever I saw one. The more I think about it, the more it seems to me that Windows and PHP deserve each other.
Because supporting your own embedded version of Linux that no-one outside one small room in the basement of your offices is going to modify, on your own hardware, the spec of which isn't going to change, is relatively easy once you've got the thing working - in fact it's probably easier than supporting a proprietary embedded system. On the other hand, supporting any of a dozen major linux distros running on a thousand different hardware setups, using different sets of drivers for each and every peripheral, with the choice of at least two desktops and millions of permutations of modules, before the user started customising and recompiling, and no standard way to distribute your software to all distros apart from a tarball'd set of source files, isn't easier than supporting Windows or Mac end users. Especially given that at least some linux users are going to be more interested in proving they are smarter than the helpdesk team than in getting the product to work, and that a lot of linux fans will use a OSX or Windows when they have to.
And, as others have said, why would you expect one to follow the other anyway? If my company was making money from using an embedded OSS system, I might be inclined to put $$$ or developer hours into helping the OSS development community, but I really cannot see why I would be under any moral obligation to help the distributors of non-embedded distros I don't use or the desktop users who are consumers just like me.
the biases of slashdot editors don't count for much in terms of reliable data
I think I acknowledged that above. But while slashdot editors are a small sample, slashdot users are a much larger group, and I don't see a huge and growing grassroots movement against/. editorial bias along the lines of "Please can we recognise that limiting individual rights is sometimes a good idea!", which suggests to me that the editors are delivering more or less what the user base wants. And because of the/. moderation system we know this isn't because the evil editors are deleting all the critical comments while we sleep.
When individual rights are lost, you can't really word that as a gain for society. It's a loss for society.
Sometimes you can, sometimes you can't. Emprisonment is about loss of individual rights, but most of us think that society is justified in emprisoning at least some people, and that society as a whole benefits from having some people out of circulation.
The way "Your rights online" is one of the busiest/. categories, the way half the stories have little or nothing to do with IT, and the way articles are almost always spun in terms of "What individual rights will be lost?" rather than "What might society as a whole gain?", for example?
Having said that, I haven't seen any survey data, and I suspect that the population of nerds is closer to the centre of the bell curve than those who make the most noise.
Slashdot Mirror
I can understand that, but, if you followed that logic consistently, you'd never buy a computer at all. I bought a 701, which I liked apart from the screen. Then I bought a Windows 900, on which I've installed Kubuntu, and I'm quite happy with it. It's a bit irritating that the next model is out already, but I'll be using mine on a series of train trips next week. If I had done things your way, I'd be reading magazine reviews instead of doing any work...
I think 'revolutionize' is a bit of an overstatement. Most people who are likely to fancy a bit of Linux-based hardware diagnosis can already do that with their favourite boot-from-CD distro, but it's hardly the sort of thing your average Word power user is going to try on the bus on the way to work.
Including an OS on the motherboard makes sense for Asus - at least it is then possible to do basic hardware diagnostics independently of, say, Windows diagnostics.
But, in terms of Linux adoption, it's only exciting if people keep linux once they've finished building the computer, and the precedents here are hardly promising.
And, even if you like Linux (which I do), would you want to keep the version supplied with your m/b? On my first EeePC, I tried to get to like Xandros, I really did, but in the end I wiped it and installed Kubuntu. My Dark Side Brother played with Xandros until he broke it, and then installed XP. And it's going to happen even more with the EeePC 900, since the Linux version has a larger SSD than the Windows version (at least in the UK), so you buy the Linux version in order to install Windows.
Which beam would that be? I own several linux boxes, one Mac and no Windows machines... My point about Vista is that no-one is saying "This is the most wonderful OS ever".
The thing the article and most of the comments miss is that Apple don't sell your a product, they sell you a self-image. Buying a PC isn't a bad choice or bad value for money, it makes you a corporate drone. If you own a Mac, you are a free-thinking, creative, pretty young thing - whatever your age, IQ or silouette. You don't pay Steve Jobs to ship you a computer. You pay him to make you a better person.
So when people criticise a Mac, it hits Mac owners at the same level as comments about their personal appearance. When you say that the AirBook is overpriced, it has the same emotional impact as telling the AirBook's owner to consider plastic surgery.
Actually, price criticism isn't too serious, because it invites the "I care about myself enough to buy the best" defence. The real killer is pointing out that, actually, Apple does far more lock-ins than Microsoft nowadays. Those who bought their self-esteem from Apple read that as "You thought you had found the One True Church, but in reality you joined a sect led by an LA snake-oil salesman."
On that basis, expect Mac Rage to turn thermonuclear if/when the Linux desktop ever gets any sort of market penetration, because even Apple can't make Linux look like an evil empire.
And, yes, Linux geeks do the same sort of thing, except that they do it in a nerdier and cheaper way. TBH, Windows users are probably the least religious group out there right now, especially post-Vista.
Toshiba was found guilty of illegally selling the Soviet Union
Wow, respect! How much did they get for it? With such incredible marketing skills, it seems all the stranger that they couldn't sell the HD-DVD format. I mean, that was legal...
I've lost count of how many times I've seen this rather basic mistake. Perl is compiled before it executes. So there's a compiler overhead with each execution, and that can amount to several seconds if you use loads of libraries, which is why it's best to run non-trivial perl websites using mod_perl. But once it's compiled, it's as compiled as C and rather more compiled than java.
For many text-based tasks, Perl is blindingly fast. The canonical example (from the Camel Book) is implementing grep in perl, which runs faster than some native OS implementations. That's partly because it's optimised for that sort of task, but it's also that it has a ruthlessly efficient implementation of hashes. Sure, you can do better in C if you really try, but if you don't really try a perl nested hash structure will blow C linear search out of the water.
The Camel book acknowledges that perl isn't the fastest option for tight loops, and ray tracing would probably fall into that category. But, more generally, the other thing about perl is that it has an extremely large collection of APIs to C libraries. If I want to parse XML, LibXML/LibXSLT are among the fastest C libraries out there, and those libraries run about as fast when called from perl as they do called from C.
Maybe it's just me but I don't equate an increase in search engine queries regarding programming in python as indicating an increase in the popularity of python.
Indeed. It probably tells you more about the randomness of the documentation. I almost never search for anything connected to perl using a generic search engine, since pretty much everything is on CPAN. OTOH, if your preferred language has a 'right-click' approach to modularisation (which seems to me to be the case with PHP), you spend all your time googling for the right bit of code to right-click.
Ok, I'll bite. Since my earlier example concerned emacs, how much does Richard Stallman charge for an emacs warranty?
I'm sorry you are astonished and angry. But your post went
In future, if you must be patronising, try to be patronising and right. I've explained why I think your entire explanation heads off in the wrong direction, and some of what you say is downright wrong. For example, your claim that open sourcing of code somehow ensures ongoing support of the code is patent nonsense, as a brief look at sourceforge will confirm. There are lots of open source projects out there with an active user base and no developers.
Now, on that point, I'm inclined to agree that open sourcing the code provides better guarantees of ongoing support in most circumstances than closed source code. But posting a gross generalisation and then signing off with "There, was that so difficult" is a way to paint a large target on your butt.
Incidentally, I see from your website that you do write code, so I take back my suggestion to the contrary a few posts up.
A choice you can't afford to make isn't a choice. I understand the benefits of the code being available to be modified. I'm simply pointing out that, in practice, most people have no option but to use open source software as if it couldn't be modified, because they do not have the resources (technical or financial) to modify it. This means the theoretical possibility of them modifying the code isn't any kind of selling point for them.
Also, I think you'll find that a lot of closed source companies are willing to modify or fork their code, for a price. I've seen more than one small software vendor make small fixes, for free, in response to my feedback. OTOH, XFree86 forked after complaints that the project owners weren't as responsive as my small closed-source vendors. The picture is a lot more complex than "Open Good, Closed Bad".
Ok, you win. Open source clearly isn't perfect. Therefore, we should all throw in the towel and go back to closed-source software. After all, since open source software isn't perfect, closed source software is clearly better.
No, we should just stop over-selling open source as if the mere fact that the code is available makes all things possible and solves every problem from bugs to world poverty and acne. Selling points for the general public are more along the lines of long-term availability of the software, a better record on fixing bugs and a culture that encourages interoperability. "You can fix it yourself" isn't a selling point for most people, even if it were true. "This car comes with no warranty, there are no dealers, but you are free to cast your own engine parts when it breaks down, and even to distribute those engine parts to third parties" isn't a sales pitch you are going to see on TV any time soon.
If there is any bug, or desireable feature that is missing, or really any kind of improvement to be made, it can be made by anyone. This includes you, but you don't have to do it yourself - chances are there is somebody who wants the same improvement and will make it and share it with the world.
There ought to be a Slashdot autoresponder for this suggestion. It is not and never has been true of software, and is even less true of hardware. If you really think you can personally add whatever feature you fancy to any and all open-source software in anything like a reasonable timescale, I suspect you've never tried. I'd really like emacs to display all regexes properly (# is a particular problem in both perl and tcl modes). The bug has been around for years, and I'm sure tens of thousands of technical users have noticed. Can you fix that for me by next Wednesday?
If you think it makes sense to even try fixing other people's code in most cases, you are probably still in full-time education. I suspect I could work out a fix for the above problem if I had nothing else to do for a few months, but the reason I use emacs in the first place is because I have other things to do. That's true of the vast majority of software users. The options are to live with the bugs, hold out for a bug fix and find another product that does the same job better.
Cathedral and Bazaar describes (probably, I suspect, in a rather rose-tinted way) the development of fetchmail as a collaborative open source project, but, for heaven's sake, it's part of server-side email infrastructure. The contributors were all übertechies. What proportion of end users of systems that use fetchmail do you think have even heard of fetchmail, let alone being able to improve the source code, let alone persuading whoever handles their mail to install the improved version?
The reality is that modern software is generally extremely complex to maintain, that there's a huge learning curve in unpicking even well-documented projects, that there are politics attached to getting your mods accepted by the main project, and that effectively forking the project in order to add your one new feature is generally the worst thing you could possibly do in terms of long-term reliability. You might possibly be able to fix the bug you've noticed, but making sure you haven't introduced three more bugs in the process is a little more tricky.
That's software. The main thing stopping most of us from hacking our hardware isn't DRM, it's not having the technical competence and the extremely expensive kit to go with it. Heck, it's virtually impossible to do routine maintenance on most cars nowadays without the vendor's test rig.
There are advantages for everyone in open standards, and maybe in open hardware, but the advantages don't come from Grannie Brown being able to turn her PC into a video recorder on a wet Tuesday afternoon. You still need the backing of a fairly large institution (university or company) to do hardware hacking that achieves more than a mention on Slashdot hardware hacks.
Why don't you need one? In the photo on the last /. article the kite isn't dead ahead of the ship, and you'd expect that to be the case most of the time. If that tends to pull the ship off course, don't you end up using the rudder like a rather poor keel, effectively dragging the ship through the sea sideways, and thus wasting a load of energy, not to mention the stress on various parts of the ship?
Makes about as much sense as most statements by the FSF.
I know what binary compatible means. Do you know what 'importance of brand names to non-technical people controlling chequebooks' means? And Fedora isn't Gentoo, but it still smells of hobbyist to customers used to paying $$$ for server software.
Not only does Redhat 'tolerate' CentOS (see above), it also puts money into encouraging people to use Fedora, which is not only free but generally significantly more advanced than RHEL. For people who want free software and enjoy recompiling their kernel, Fedora is a much more obvious choice than a clone of CentOS.
There was never any money in selling distros to dorm-room techies, and RHEL was never a good distribution for that market, because it's so conservative. I run Ubuntu on my desktop machines, because it's free, and it works, and it has all the multimedia stuff that RedHat don't ship as standard. On my company's production servers it's RHEL every time, because it's stable, because it will still be supported in 5 years' time if necessary, and because RHEL is a de facto standard in hosting terms. If a client's code doesn't work with RHEL, we can tell them to fix their code. If we were running some wacky, customised version of Gentoo they'd tell us to fix our server (whether or not anything was broken).
Running CentOS would give us the conservatism of RHEL without any of the respectability. I can't see how that would be useful to us.
you have to think in OOXML.
Great, so now I get the legendary security of IISS coupled with the legendary security of PHP? To be fair, it shouldn't be much worse than PHP/Apache, since the most common PHP setup effectively disables user-based access control and lets any PHP script anywhere on the server access any data created by any other PHP script anywhere on the server.
I do hope they've kept that thing where you install a new PHP application and then try to run the admin.php script via the web server as nobody before anyone else on the Internet does. That's a race condition implemented with real vision if ever I saw one. The more I think about it, the more it seems to me that Windows and PHP deserve each other.
Shouldn't that be "resolution of less than 2 nanometers", or maybe "accuracy better than 2 nanometers"?
Because supporting your own embedded version of Linux that no-one outside one small room in the basement of your offices is going to modify, on your own hardware, the spec of which isn't going to change, is relatively easy once you've got the thing working - in fact it's probably easier than supporting a proprietary embedded system. On the other hand, supporting any of a dozen major linux distros running on a thousand different hardware setups, using different sets of drivers for each and every peripheral, with the choice of at least two desktops and millions of permutations of modules, before the user started customising and recompiling, and no standard way to distribute your software to all distros apart from a tarball'd set of source files, isn't easier than supporting Windows or Mac end users. Especially given that at least some linux users are going to be more interested in proving they are smarter than the helpdesk team than in getting the product to work, and that a lot of linux fans will use a OSX or Windows when they have to.
And, as others have said, why would you expect one to follow the other anyway? If my company was making money from using an embedded OSS system, I might be inclined to put $$$ or developer hours into helping the OSS development community, but I really cannot see why I would be under any moral obligation to help the distributors of non-embedded distros I don't use or the desktop users who are consumers just like me.
I think I acknowledged that above. But while slashdot editors are a small sample, slashdot users are a much larger group, and I don't see a huge and growing grassroots movement against /. editorial bias along the lines of "Please can we recognise that limiting individual rights is sometimes a good idea!", which suggests to me that the editors are delivering more or less what the user base wants. And because of the /. moderation system we know this isn't because the evil editors are deleting all the critical comments while we sleep.
Sometimes you can, sometimes you can't. Emprisonment is about loss of individual rights, but most of us think that society is justified in emprisoning at least some people, and that society as a whole benefits from having some people out of circulation.
The way "Your rights online" is one of the busiest /. categories, the way half the stories have little or nothing to do with IT, and the way articles are almost always spun in terms of "What individual rights will be lost?" rather than "What might society as a whole gain?", for example?
Having said that, I haven't seen any survey data, and I suspect that the population of nerds is closer to the centre of the bell curve than those who make the most noise.