Also, don't forget that hydrogen and biofuel can be used in a fuel cell, which in turn does NOT need to be contain expensive metals (this requirement is a thing of the past, fortunately). The generated electric current goes to an electric engine. This is more efficient, since the energy efficiency ratio can be higher, but more importantly, the engine does NOT need to run when the car is idle! Lots of fuel are wasted in traffic jams, or just by waiting for the traffic light to turn green. With electric engines, no fuel is wasted since you can start moving at 0 rpm, thereby reducing overall fuel consumption.
Batches are necessary, they are right about that. Without batching, you can never use the graphics hardware optimally. Many games are CPU-bound because they issue too many API commands, for example, if there are 5000 visible trees, then you have to send 5000 drawcalls. It gets worse if one mesh has multiple materials, for example a tree with a material for the trunk, another for the leaves etc. In this case, you can only group the geometry with the same material together. Instancing helps reducing the overhead for rendering geometry with the *same* material, but if your game level has 47 materials, all of them visible, you have to render all of them separately. DX10 helps by introducing texture arrays and constant buffers, which means that you can stuff all your textures into one array adressable without issuing commands, same for constants (like, color or specular exponent). In the end, you just issue ONE drawcall, and the mesh gets drawn, with its multiple materials. Mind you, display lists could be an OpenGL equivalent, but usually aren't (performance-wise).
Oh Linux IS ready for me. There are binary drivers with full 3D support after all. And god I HATE this "Linux-is-not-for-you" attitude. This is one of the reasons why Linux has its problems with the mainstreams: wannable-l33tness which shows itself by arrogant exclusion of people. "Linux is not for you, WE decide who is l33t enough to use Linux. Go away, l4m3r, use Windows!" Fact is: A Linux distro HAS TO BE FOR GAMERS TOO if it shall be able to compete with Windows and OSX as a mainstream desktop OS.
And for the record: I develop 3D applications, games, and visualization stuff. So I NEED 3D support. Maybe you Linux zealots despise graphics programming, since only ultra-l33t kernel hacking and network programming is worthy enough for you Homo Superiors. This is one good reason why virtually no 3d programming happens in Linux, and most 3d engines are purely Windows-oriented. Well done scaring off devs!
If a manufacturer can provide a graphics card where the hardware interface is open and which supports all the things you need these days for games like Doom III, Unreal Tournament and Neverwinter Nights (like pixel and vertex shaders), I for one am prepared to put my money where my mouth is and support them.
Only problem being that a lot of the functionality these games require is actually in the driver, which is very very likely to be closed-source forever. Its as if you want to recreate a full personal computer with a state-of-the-art OS on it, but only get the schematics of the hardware, with NO software whatsoever. This will take a while, especially since the closed-source stuff is filled with tons of functionality and is being extended at an enormous pace.
And you would sacrifice all that for slightly faster 3D graphics? I can't comprehend your state of mind. Your priorities are completely foreign to me.
I both make a living and write hobby stuff related with 3D programming. I NEED good 3D support, and want to use not only Windows, but Linux as well. My choice? Non-existant: closed source graphics drivers are the only way.
Why do you need a gang of coders to product a good product? 2 or 3 good coders could produce an excellent driver in a good amount of time. Yes there is more bad and mediocre coders than good ones but the point is there is enough good ones to do this.
Again: these drivers are very complex and hard to write. The requirements are extreme, the feature list very long, and lack of direct communication with the HW department really shows. Also, if they need more than one year to come to today's standard, the whole thing is pointless, people will still choose the closed source ones (because they are better/faster/etc.)
Also, those poor sods would start from scratch. No previous experience with earlier driver development for the cards, no codebase...
I believe, sir, that this is the very definition of short-sightedness.
Of course, if you put things out of context, they can be rearranged to your benefit. I also said that the chance of the drivers being open-sourced is nil.
Who do you suppose writes the drivers at nVidia and ATI? Do you think they have superhuman robots doing all their coding for them? If they are in fact real people, what makes them better than OSS coders?
Lets see... Skill? Experience? Vast, tried-and-true codebases? DIRECT, live communication with the HW department? Very few developers (not only in the OSS field) are actually highly-skilled. nVidia has one thing to attract them that OSS doesn't: money.
Um. Did you need a calculator to figure that out? How many people are we talking here? I think you may have just made up a number here. I'm really getting tired of typing the phrase "no basis in fact".
They have LOTS of devs and entire compilerfarms for their drivers. Also, do you honestly believe that newcomers to the field will magically develop an OS driver actually capable of competing in less than one year? Come on. Maybe if you assemble an ultra-motivated group of very highly skilled cracks.
I am just saying there is plenty of very capable developers that could handle the drivers.
I know. And this is wrong. First, there are not PLENTY of very capable devs. There are plenty of MEDIOCRE ones. This makes it rather unlikely that one gang of uber-coders would form and present us a state-of-the-art driver supporting the latest geForce in less than one year. Also, the nVidia devs have the advantage of being very, very experienced.
Perhaps but the point is some of their customers have made a request, which makes his statement plain wrong.
You are splitting hairs here. These "dozens of calls" are virtually non-existant. They don't see opensource fanatics as the average customer. The average customer just wants good flashy fast 3D.
Firstly that is a very arrogant approach, some of the best developers in the world work on open source stuff, saying it is to hard is just stupid.
Wrong conclusion. FEW open-source-developers belong to the top devs. Most OS devs are mediocre. As always. OS is not a magical title that automatically summons Michael Abrash clones.
As for customers not asking for open-source drivers, all I can say is huh? There have been dozens of calls over the years for drivers to be open sourced!
FSF cannot forbid anything. Linus can.
Also, if the GPL zealots win, then the companies would simply fork the kernel. Possibly create a totally closed-source one. Would that be better?
This is a self-fulfilling prophecy of sorts. The reason that open-source drivers for nVidia and ATI cards don't have usable 3D is precisely because you and others continue to use the proprietary drivers. It short-circuits the demand for free drivers just enough to satiate people into not coding them. You may think you're being pragmatic, but you're really just being short-sighted.
Fact: I need good 3D support. NOW. Not in 2000 years. Fact: The opensource drivers don't offer this. And the CS ones will never ever be OS. Conclusion: use the closed source ones. THIS, sir, is not short-sighted. On the contrary, it is just wishful thinking that those two corps will EVER put their drivers under an opensource license. With this in mind, I load the propietary nvidia module.
You say it's not likely, but you give no reason. Fine, let's assume it's unlikely. They wouldn't have to open their drivers. They'd just have to open the hardware spec enough to allow open driver coding. It could easily work business-wise for them, since they could offload a lot of coding work onto the free software community.
With the slight problem that writing such a driver is not easy. In fact, among all drivers, these are one of the hardest to write. A GPU has a very complex architecture, it is an entire mini-computer, the driver doesn't just pass on commands, the driver includes an entire 3D engine! And no, firmware is NO solution, firmware is cumbersome to update, if the upload is interrupted your card does not work, and the CPU is just plain faster than some microcontroller (what, you want to include a CPU on that graphics card?). And don't tell me that someone will write an OS driver that can compete with the CS ones, this is just plain wrong. Today's features would be avaiable in 5-7 years, if people would work 24/7 on them. 5-7 years are a LONG time in this business.
Again, you're just saying this, but giving no reason why. There is absolutely no valid reason why any proprietary element is unavoidable! That's just bullshit defeatist thinking with no basis in fact other than the status quo: "Things have been this way for as long as I care to remember, so how could they ever be different?"
Oh yeah? Its called business. They OS their stuff, and suddenly there are $50 chinese cards replicating the $400 originals -> nVidia loses a *lot* of money. Also, there are the usual patent issues and corporate secrets. I'm sure ATI would love to see the nVidia code, both for stealing (since they wouldn' OS their stuff, nVidia cannot steal back) and for sueing (again, nVidia cannot sue back because ATI's stuff isn't open).
Large userbase? Compatibility with post-1995 graphics hardware? -> Allow closed source drivers. Be a purist? Allow open source drivers only.
Also keep in mind that your "signal to the IHVs" is ridiculous. They couldn't care less about those open-source fanatics since they have NO REAL POWER. If a purist distro had ~40-50% marketshare, THEN they would write open source drivers. But now? They don't lose anything. So their choice is not "OS or CS", its "shall we write a Linux driver or not", with the CS part included implicitely.
That said, I don't mind if the kernel stays puristic. There are leading distros already (SuSe, Fedora, Mandriva, Ubuntu), one solution would be some guidelines for a stable driver ABI allowing closed-source modules which can be attached to a vanilla kernel. Kind of like a LSB for CS driver kernel support. Some distros implement this, some don't. End of story.
But ATI & nVidia are the ones who REALLY count. You want games? 3D programming? Visualization? GPU-based calculation of non-graphics software? Use their hardware. Word processor? Photoshop? Terminal-only? Fileserver? Stick with your on-board chipset.
The problem is that userspace is SLOW. Even with DRI, you need context switches. Also, the kernel modules perform not only the connection to the main module, they also multiplex the commands sent to the card.
Mod me offtopic, but... why all this GNU/Linux madness? Trying to be a GNU nazi? Its like the persons who always ensure that everything is gender-neutral. "His/her package will be delivered to his/her bank account." (In other languages like German or Spanish its even worse.) We are used to say "Ubuntu", or "Ubuntu Linux", thank you very much. In fact, if you want to split hairs, you should call it Ubuntu GNU/GNOME/Linux, shouldn't you? Or Ubuntu GNU/X11/GNOME/Linux?
Adept is bad, yes. I use Synaptic. But, KDE is MUCH more than a WM. You might have noticed the TONS of kdelibs-based programs, the kioslaves, the kparts, etc. etc. This integration is the difference between a desktop environment and a window manager. In fact, a window manager IS PART of a DE (KDE's window manager is called KDM).
It must be your fault. I installed dapper, apt-getted build-essentials, gcc & g++ 4, make, and can build both sources from the net and my projects without any problems. No ld.so.conf problems arise.... ever.
Actually, it does. Many people live a life where they wait God to come and save them. The result is that the governments of the countries where these people live in take advantage of those people.
Wait - you forget something. What you are talking about is the classical dogmatic religion with high priests, rules, and non-thinking. You have to differentiate here. Belief is not the same as religion.
One should find out what to believe by oneself, NOT by hearing some old priest reading from a thick book. Everybody asks "why am I here", "what is the purpose of life?" etc. these questions, while being non-addressable by science, are valid. The point is that one has to find these answers by actively thinking about it, searching for them etc. As a result, you start looking at yourself, asking who you really are, what you really want in life, questions which are VERY important because they shape a person. This involves a LOT of critical thinking (especially about yourself - one of the hardest things to do), and has the side benefit that you become a critical and very self-aware person who is hard to manipulate. Thus, IMO belief is much closer to philosophy than religion.
But what is religion? Organized, preprocessed belief (which is actually an oxymoron). This is what makes religions so dangerous: high priest X already thought about things, or got wisdom from God/Allah/whatever, you are supposed to follow this - and DO NOT think about it or even question it! (This is why I don't see Buddhism as religion; its far too personal and does not think for you.)
As a consequence, most people are actually doing things the wrong way. And the only reaction is the complete opposite: Atheism. Its ok if you get to become an atheist after thinking about things, and finding your way, but its NOT ok if one chooses atheism simply because religions are so evil.
First: Science does not deal with truths, only with models.
Second: Science cannot be applied everywhere. There are questions that cannot be answered by science, because no answer fulfills the requirements. (Like, "what is outside of the universe", or "why are we here".) There comes a point where the only thing you can do is - believe. In something. Some believe that there are no higher entities (science cannot disprove them, but because of this they are filtered out by Occam's Razor, just like all non-disprovable things). Some believe that life is guided by some god, some believe in a living an conscious Mother Nature etc. Claiming that atheism is "The Only Way" (tm) is just plain wrong because it does not have any advantages over other beliefs.
Well yes, I know that such homemade paradigm shifts are extremely unlikely. But I don't like the tendency towards exaggerated conservative thinking. People are effectively banned because of doing research in "unserious" fields. I don't even mean stuff like perpetuum mobiles - take electrogravity as an example. Anyone theorizing with electricity/gravity connections and usage of this connection is viewed as a crackpot and "ruined his reputation". Mind you, this is JUST for toying with such theories, not even remotely considering attempts at getting money for experiment.
This is something I really dont like; scientists with ruined reputation equal banned heretics. Of course there are tons of crackpots, but IMO no one should be punished for theorizing in "unserious" areas (even the definition of "unserious" and "serious" is dangerous IMO). One should be banned when the great free-energy device X will be ready in two years and you can preorder it now, or because the entire work was a hoax (like that Hwang guy did), not because one said that "it might be possible that ABC is possible, I'm quite skeptical though, I'm looking into this". Of course, once somebody can successfully demonstrate working electrogravity that can be replicated in at least one independent lab, reputation is back. But what if by banning everyone in this field no one has the chance of reaching this stage?
A isn't true, B might be (we don't know yet). Also, in order to expand "information space", you need to expand in physical space. And by killing off dreams about the last real frontier, you aren't doing any good. Just like the farmer boy who always dreams about moving out and becoming something greater, but who is forced by his parents to remain in the farm.
There is a big difference between condemning free thought through religious mania and debunking a hare-brained idea that a college freshman can easily prove to be false (the pursuit of which wastes tax dollars that can be used to feed hungry people).
FTL is not bunk because gawd/allah/odin/yahweh/ram said so. FTL is bunk because it ameaningless state in a classical timelike metric. I won't burn you at the stake for trying to work on FTL. However, I will write a sternly worded letter to the NSF recommending that they don't give you any money for it.
The problem is: with this thinking you kill off many breakthroughs. Remember that theories are just models. Now if by any chance one model is false, and a guy thinks he can prove it AND fix it, he won't get any support because the established model doesn't predict his claims. To prove his claims, he might need some pretty expensive equipment, with the NSF has, for example. But, if YOU prevent this from being tested, you may be killing off one breakthrough. You NEVER know if something works or not in advance for sure. Thats why scientists perform experiments. Of course there are many crackpots, but if science remains in its established, comfortable theories, then nothing will advance.
Slashdot Mirror
Also, don't forget that hydrogen and biofuel can be used in a fuel cell, which in turn does NOT need to be contain expensive metals (this requirement is a thing of the past, fortunately). The generated electric current goes to an electric engine. This is more efficient, since the energy efficiency ratio can be higher, but more importantly, the engine does NOT need to run when the car is idle! Lots of fuel are wasted in traffic jams, or just by waiting for the traffic light to turn green. With electric engines, no fuel is wasted since you can start moving at 0 rpm, thereby reducing overall fuel consumption.
Batches are necessary, they are right about that. Without batching, you can never use the graphics hardware optimally. Many games are CPU-bound because they issue too many API commands, for example, if there are 5000 visible trees, then you have to send 5000 drawcalls. It gets worse if one mesh has multiple materials, for example a tree with a material for the trunk, another for the leaves etc. In this case, you can only group the geometry with the same material together. Instancing helps reducing the overhead for rendering geometry with the *same* material, but if your game level has 47 materials, all of them visible, you have to render all of them separately. DX10 helps by introducing texture arrays and constant buffers, which means that you can stuff all your textures into one array adressable without issuing commands, same for constants (like, color or specular exponent). In the end, you just issue ONE drawcall, and the mesh gets drawn, with its multiple materials.
Mind you, display lists could be an OpenGL equivalent, but usually aren't (performance-wise).
Oh Linux IS ready for me. There are binary drivers with full 3D support after all.
And god I HATE this "Linux-is-not-for-you" attitude. This is one of the reasons why Linux has its problems with the mainstreams: wannable-l33tness which shows itself by arrogant exclusion of people. "Linux is not for you, WE decide who is l33t enough to use Linux. Go away, l4m3r, use Windows!"
Fact is: A Linux distro HAS TO BE FOR GAMERS TOO if it shall be able to compete with Windows and OSX as a mainstream desktop OS.
And for the record: I develop 3D applications, games, and visualization stuff. So I NEED 3D support. Maybe you Linux zealots despise graphics programming, since only ultra-l33t kernel hacking and network programming is worthy enough for you Homo Superiors. This is one good reason why virtually no 3d programming happens in Linux, and most 3d engines are purely Windows-oriented. Well done scaring off devs!
This is naive thinking. 2D framebuffers are so far below the complexity of full 3D support you can't even compare them.
If a manufacturer can provide a graphics card where the hardware interface is open and which supports all the things you need these days for games like Doom III, Unreal Tournament and Neverwinter Nights (like pixel and vertex shaders), I for one am prepared to put my money where my mouth is and support them.
Only problem being that a lot of the functionality these games require is actually in the driver, which is very very likely to be closed-source forever. Its as if you want to recreate a full personal computer with a state-of-the-art OS on it, but only get the schematics of the hardware, with NO software whatsoever. This will take a while, especially since the closed-source stuff is filled with tons of functionality and is being extended at an enormous pace.
And you would sacrifice all that for slightly faster 3D graphics? I can't comprehend your state of mind. Your priorities are completely foreign to me.
I both make a living and write hobby stuff related with 3D programming.
I NEED good 3D support, and want to use not only Windows, but Linux as well.
My choice? Non-existant: closed source graphics drivers are the only way.
They only have 2D experience. The nv driver cannot be fairly compared with the nvidia one.
Why do you need a gang of coders to product a good product? 2 or 3 good coders could produce an excellent driver in a good amount of time. Yes there is more bad and mediocre coders than good ones but the point is there is enough good ones to do this.
Again: these drivers are very complex and hard to write. The requirements are extreme, the feature list very long, and lack of direct communication with the HW department really shows. Also, if they need more than one year to come to today's standard, the whole thing is pointless, people will still choose the closed source ones (because they are better/faster/etc.)
Also, those poor sods would start from scratch. No previous experience with earlier driver development for the cards, no codebase...
I believe, sir, that this is the very definition of short-sightedness.
Of course, if you put things out of context, they can be rearranged to your benefit. I also said that the chance of the drivers being open-sourced is nil.
Who do you suppose writes the drivers at nVidia and ATI? Do you think they have superhuman robots doing all their coding for them? If they are in fact real people, what makes them better than OSS coders?
Lets see... Skill? Experience? Vast, tried-and-true codebases? DIRECT, live communication with the HW department? Very few developers (not only in the OSS field) are actually highly-skilled. nVidia has one thing to attract them that OSS doesn't: money.
Um. Did you need a calculator to figure that out? How many people are we talking here? I think you may have just made up a number here. I'm really getting tired of typing the phrase "no basis in fact".
They have LOTS of devs and entire compilerfarms for their drivers. Also, do you honestly believe that newcomers to the field will magically develop an OS driver actually capable of competing in less than one year? Come on. Maybe if you assemble an ultra-motivated group of very highly skilled cracks.
I am just saying there is plenty of very capable developers that could handle the drivers.
I know. And this is wrong.
First, there are not PLENTY of very capable devs. There are plenty of MEDIOCRE ones. This makes it rather unlikely that one gang of uber-coders would form and present us a state-of-the-art driver supporting the latest geForce in less than one year. Also, the nVidia devs have the advantage of being very, very experienced.
Perhaps but the point is some of their customers have made a request, which makes his statement plain wrong.
You are splitting hairs here. These "dozens of calls" are virtually non-existant. They don't see opensource fanatics as the average customer. The average customer just wants good flashy fast 3D.
Firstly that is a very arrogant approach, some of the best developers in the world work on open source stuff, saying it is to hard is just stupid.
Wrong conclusion. FEW open-source-developers belong to the top devs. Most OS devs are mediocre. As always. OS is not a magical title that automatically summons Michael Abrash clones.
As for customers not asking for open-source drivers, all I can say is huh? There have been dozens of calls over the years for drivers to be open sourced!
Which make up, uh, 1% of all their customers?
FSF cannot forbid anything. Linus can. Also, if the GPL zealots win, then the companies would simply fork the kernel. Possibly create a totally closed-source one. Would that be better?
This is a self-fulfilling prophecy of sorts. The reason that open-source drivers for nVidia and ATI cards don't have usable 3D is precisely because you and others continue to use the proprietary drivers. It short-circuits the demand for free drivers just enough to satiate people into not coding them. You may think you're being pragmatic, but you're really just being short-sighted.
Fact: I need good 3D support. NOW. Not in 2000 years.
Fact: The opensource drivers don't offer this. And the CS ones will never ever be OS.
Conclusion: use the closed source ones.
THIS, sir, is not short-sighted. On the contrary, it is just wishful thinking that those two corps will EVER put their drivers under an opensource license. With this in mind, I load the propietary nvidia module.
You say it's not likely, but you give no reason. Fine, let's assume it's unlikely. They wouldn't have to open their drivers. They'd just have to open the hardware spec enough to allow open driver coding. It could easily work business-wise for them, since they could offload a lot of coding work onto the free software community.
With the slight problem that writing such a driver is not easy. In fact, among all drivers, these are one of the hardest to write. A GPU has a very complex architecture, it is an entire mini-computer, the driver doesn't just pass on commands, the driver includes an entire 3D engine! And no, firmware is NO solution, firmware is cumbersome to update, if the upload is interrupted your card does not work, and the CPU is just plain faster than some microcontroller (what, you want to include a CPU on that graphics card?). And don't tell me that someone will write an OS driver that can compete with the CS ones, this is just plain wrong. Today's features would be avaiable in 5-7 years, if people would work 24/7 on them. 5-7 years are a LONG time in this business.
Again, you're just saying this, but giving no reason why. There is absolutely no valid reason why any proprietary element is unavoidable! That's just bullshit defeatist thinking with no basis in fact other than the status quo: "Things have been this way for as long as I care to remember, so how could they ever be different?"
Oh yeah? Its called business. They OS their stuff, and suddenly there are $50 chinese cards replicating the $400 originals -> nVidia loses a *lot* of money. Also, there are the usual patent issues and corporate secrets. I'm sure ATI would love to see the nVidia code, both for stealing (since they wouldn' OS their stuff, nVidia cannot steal back) and for sueing (again, nVidia cannot sue back because ATI's stuff isn't open).
it depends on what a distro wants.
Large userbase? Compatibility with post-1995 graphics hardware? -> Allow closed source drivers.
Be a purist? Allow open source drivers only.
Also keep in mind that your "signal to the IHVs" is ridiculous. They couldn't care less about those open-source fanatics since they have NO REAL POWER. If a purist distro had ~40-50% marketshare, THEN they would write open source drivers. But now? They don't lose anything. So their choice is not "OS or CS", its "shall we write a Linux driver or not", with the CS part included implicitely.
That said, I don't mind if the kernel stays puristic. There are leading distros already (SuSe, Fedora, Mandriva, Ubuntu), one solution would be some guidelines for a stable driver ABI allowing closed-source modules which can be attached to a vanilla kernel. Kind of like a LSB for CS driver kernel support. Some distros implement this, some don't. End of story.
But ATI & nVidia are the ones who REALLY count. You want games? 3D programming? Visualization? GPU-based calculation of non-graphics software? Use their hardware. Word processor? Photoshop? Terminal-only? Fileserver? Stick with your on-board chipset.
The problem is that userspace is SLOW. Even with DRI, you need context switches. Also, the kernel modules perform not only the connection to the main module, they also multiplex the commands sent to the card.
Mod me offtopic, but...
why all this GNU/Linux madness? Trying to be a GNU nazi? Its like the persons who always ensure that everything is gender-neutral. "His/her package will be delivered to his/her bank account." (In other languages like German or Spanish its even worse.) We are used to say "Ubuntu", or "Ubuntu Linux", thank you very much. In fact, if you want to split hairs, you should call it Ubuntu GNU/GNOME/Linux, shouldn't you? Or Ubuntu GNU/X11/GNOME/Linux?
Adept is bad, yes. I use Synaptic.
But, KDE is MUCH more than a WM. You might have noticed the TONS of kdelibs-based programs, the kioslaves, the kparts, etc. etc. This integration is the difference between a desktop environment and a window manager. In fact, a window manager IS PART of a DE (KDE's window manager is called KDM).
It must be your fault. I installed dapper, apt-getted build-essentials, gcc & g++ 4, make, and can build both sources from the net and my projects without any problems. No ld.so.conf problems arise.... ever.
If this is a troll: nice try.
If this is serious: read Popper.
Actually, it does. Many people live a life where they wait God to come and save them. The result is that the governments of the countries where these people live in take advantage of those people.
Wait - you forget something. What you are talking about is the classical dogmatic religion with high priests, rules, and non-thinking. You have to differentiate here. Belief is not the same as religion.
One should find out what to believe by oneself, NOT by hearing some old priest reading from a thick book. Everybody asks "why am I here", "what is the purpose of life?" etc. these questions, while being non-addressable by science, are valid. The point is that one has to find these answers by actively thinking about it, searching for them etc. As a result, you start looking at yourself, asking who you really are, what you really want in life, questions which are VERY important because they shape a person. This involves a LOT of critical thinking (especially about yourself - one of the hardest things to do), and has the side benefit that you become a critical and very self-aware person who is hard to manipulate. Thus, IMO belief is much closer to philosophy than religion.
But what is religion? Organized, preprocessed belief (which is actually an oxymoron). This is what makes religions so dangerous: high priest X already thought about things, or got wisdom from God/Allah/whatever, you are supposed to follow this - and DO NOT think about it or even question it! (This is why I don't see Buddhism as religion; its far too personal and does not think for you.)
As a consequence, most people are actually doing things the wrong way. And the only reaction is the complete opposite: Atheism. Its ok if you get to become an atheist after thinking about things, and finding your way, but its NOT ok if one chooses atheism simply because religions are so evil.
First: Science does not deal with truths, only with models.
Second: Science cannot be applied everywhere. There are questions that cannot be answered by science, because no answer fulfills the requirements. (Like, "what is outside of the universe", or "why are we here".) There comes a point where the only thing you can do is - believe. In something. Some believe that there are no higher entities (science cannot disprove them, but because of this they are filtered out by Occam's Razor, just like all non-disprovable things). Some believe that life is guided by some god, some believe in a living an conscious Mother Nature etc. Claiming that atheism is "The Only Way" (tm) is just plain wrong because it does not have any advantages over other beliefs.
Well yes, I know that such homemade paradigm shifts are extremely unlikely. But I don't like the tendency towards exaggerated conservative thinking. People are effectively banned because of doing research in "unserious" fields. I don't even mean stuff like perpetuum mobiles - take electrogravity as an example. Anyone theorizing with electricity/gravity connections and usage of this connection is viewed as a crackpot and "ruined his reputation". Mind you, this is JUST for toying with such theories, not even remotely considering attempts at getting money for experiment.
This is something I really dont like; scientists with ruined reputation equal banned heretics. Of course there are tons of crackpots, but IMO no one should be punished for theorizing in "unserious" areas (even the definition of "unserious" and "serious" is dangerous IMO). One should be banned when the great free-energy device X will be ready in two years and you can preorder it now, or because the entire work was a hoax (like that Hwang guy did), not because one said that "it might be possible that ABC is possible, I'm quite skeptical though, I'm looking into this". Of course, once somebody can successfully demonstrate working electrogravity that can be replicated in at least one independent lab, reputation is back. But what if by banning everyone in this field no one has the chance of reaching this stage?
A isn't true, B might be (we don't know yet). Also, in order to expand "information space", you need to expand in physical space. And by killing off dreams about the last real frontier, you aren't doing any good. Just like the farmer boy who always dreams about moving out and becoming something greater, but who is forced by his parents to remain in the farm.
There is a big difference between condemning free thought through religious mania and debunking a hare-brained idea that a college freshman can easily prove to be false (the pursuit of which wastes tax dollars that can be used to feed hungry people).
FTL is not bunk because gawd/allah/odin/yahweh/ram said so. FTL is bunk because it ameaningless state in a classical timelike metric. I won't burn you at the stake for trying to work on FTL. However, I will write a sternly worded letter to the NSF recommending that they don't give you any money for it.
The problem is: with this thinking you kill off many breakthroughs.
Remember that theories are just models. Now if by any chance one model is false, and a guy thinks he can prove it AND fix it, he won't get any support because the established model doesn't predict his claims. To prove his claims, he might need some pretty expensive equipment, with the NSF has, for example. But, if YOU prevent this from being tested, you may be killing off one breakthrough. You NEVER know if something works or not in advance for sure. Thats why scientists perform experiments. Of course there are many crackpots, but if science remains in its established, comfortable theories, then nothing will advance.