The trade off to have 32 registers was probably not worth the die space and extra complexity. Having 16 probably gave most of the benefit, and having 32 provided diminishing returns.
At least with the general purpose registers, AMD wanted to go to 32, but couldn't do it without changing the instruction set. I'd assume the same thing applies to the SSE registers.
The spider ball boss is easy once you figure out the trick. There are 4 bomb slots - 2 on each side of the way up. The more damage you do, the faster the boss attacks. The trick is to hit the outer bomb slots first, as they're much harder to get to than the inner ones. If you do that, you can usually rush up to the inner ones without much difficulty.
That's probably the first reasonable response I've ever heard in respect to the general right interpretation of the second amendment.
Anyway, if that is the intended interpretation, it would mean that women and people with disabilities don't have the right to have a gun. Which goes back to my original statement that I don't think they intended the amendment to be a blanket right for everyone to have whatever guns they want.
You're so dead set on your opinion that you're not even paying attention to what anyone else says. I'm not sure why I'm bothering, as you completely ignore or misinterpret anything that doesn't fit your belief.
First off, you insisted that "The ACLU doesn't think the second amendment means the same thing you do" wasn't a valid reason for them to not defend your interpretation of it. If you're not going to consider that valid, then nothing is valid. It really was a waste of time to bother saying anything, but I did hope to at least get a somewhat intelligent response back. I was just hoping to get some insight into what you would consider valid reasoning.
I did read a few pages of your link, and found it to be just like all the others on the topic. They conclude "well regulated militia" means "everyone". Which goes back to what I said before, why bother to include something that could possibly be misunderstood if it adds no extra meaning? You could chop out the first half of the amendment without losing any meaning, and you'd make the intention a lot more clear.
I also never said that I believed the second amendment was written to provide state militias. I simply said that I believed it wasn't intended to be a blanket guarantee that everyone can have guns without restriction. I think the first half of the amendment is worded very awkwardly and is completely out of place with the rest.
As to your comment about bills relating to the states being at the end, that's just molding things to fit your view. Putting aside any questions about the second amendment, the only one referencing the states is the 10th amendment, which is a catchall. It just makes the most sense to put that at the end. I really wouldn't infer anything else about that placement.
The point of a gun is to cause harm. The ACLU doesn't defend things that cause harm to others. (Let's just not get into the whole abortion thing for now...) They'll defend freedom of speech, but they won't fight for you to be able to yell fire in a crowded building. Maybe you'd have a case if they ignored an attempt to completely ban guns, but generally, the gun control issues come down to banning high end weapons you really don't want some random guy on the street to have, or laws to avoid impulse buying.
As to the actual ammendment, here's the text:
A well regulated Militia, being necessary to the security of a free State, the right of the people to keep and bear Arms, shall not be infringed.
If that was intended as a general right for everyone then why didn't they simplify it?
The right of the people to keep and bear Arms shall not be infringed.
That would've done the job fine and avoided all confusion on the issue. Look at the rest of the Bill of Rights - there's isn't any unnecessary language anywhere. The Constitution in general isn't overly wordy. I find it very hard to believe that they chose to add that much extra text for no reason.
This changed in Vista for one reason: DRM. Microsoft made it so that certain processes, namely wmplayer.exe and halo2.exe, cannot be a target of the debugger API calls for obvious reasons.
wmplayer.exe is obvious why they'd care, but I can't see a reason for halo2.exe. I guess it would make it harder to cheat in an online game, but I can't see changing such a core part of the OS for an FPS game. Maybe if you were talking something like WoW you could make more of a case, but still... or did you have something else in mind for it?
Why would they do this? Just make the default be the deletion of the symlink instead of the target. And why would you need this permission if you already have permission to modify the directory (to create the link) and the permission to read the target? Instead they go and take the ability of normal users (meaning most people) to use symlinks by default.
I used to work on GEOS years ago. The OS had long supported symlinks but never shipped with any user accessible way to create them. Eventually someone decided to enable the UI to create them. We caught some weird issues. Users liked to do things like create links to the root of their C drive and place them on the desktop, or other similar things that resulted in symlinks pointing to directories higher up the directory path. The Find File utility wasn't coded with links in mind, so any attempt to search your C: drive took forever. It would just keep recursing over the same paths until it reached the path length limits.
So, the short story is symlinks let you create cycles in the directory tree. If you unleash this onto applications not designed to handle it, you will run into issues that will not at all be obvious to the users.
If you mean brick and mortar, I don't have the money to pay somebody to wait in line outside for hours on a Saturday night.
Yes, brick and mortar is what I was referring to. It's not that hard to get a Wii if you try. Shipments come in pretty often. Talk to the people at a store near you and find out the trends.
Anyway, if you feel the way to get something is to pay someone else to go to the store for you, that has very little to do with the price of the system. You might as well complain that Wiis are expensive because some random guy on eBay wants $1000 for a Wii.
I'll show you as soon as you show me where in the Constitution it authorizes HUD, Medicare, Medicaid, Social Security and everything else our gov't does that is not specifically spelled out in the Constitution.
I believe it would be this:
Section 8. The Congress shall have power to lay and collect taxes, duties, imposts and excises, to pay the debts and provide for the common defense and general welfare of the United States; but all duties, imposts and excises shall be uniform throughout the United States;
Specifically, the part about providing for the general welfare of the United States.
I wasn't talking about Wii Sports. I was talking about not being able to mail-order a Wii without having to buy five $50 games of the retailer's choice. What if I only want Zelda and WarioWare?
Try a retail store then. Your problem is with your choice of stores, not with the pricing of the Wii.
In the US, taxes vary completely from state to state. Some states break it down further, and have different taxes per county. New York City adds its own taxes onto the state taxes. New Jersey cuts the sales tax in crappy areas, in an effort to encourage people to spend their money there and bring more money into the local economy. Other states have other oddities.
Also, what's taxed varies from state to state. In New Jersey, prepared foods are taxed, but unprepared foods are not. This gets some weird corner cases - if you buy a Snapple bottle from a refrigerator you pay tax on it, but not if you buy one off a shelf.
In short, US sales taxes are even more complicated than income taxes. It's not as well known because most people don't have to deal with more than a couple different tax zones with any regularity.
I don't think we are trying to avoid realism in games, I think we are trying to experience things we can't in real life. Like being in a war, crashing a car or speeding lightning fast, or flying a spaceship.
I'm with you so far...
All these things I want to be as realistic as possible, because I really do not want or can not do them in real life.
And here's where we diverge. I don't care if the world is realistic or not. I care that everything feels like it belongs together.
In most modern games, we get great looking backgrounds (as long as we avoid things like trees), but people with plastic skin. It just looks wrong.
Once you start breaking from reality, you can design your world around what you can do. If you do a brightly colored game, you can usually get away with simpler skin tones that render easily. It's also easier to get away with strange scenarios, as long as they are fun. Or you can toss everyone in space suits and design "realistic" aliens with designs that render well.
It's like when you watch a movie and it's really obvious that a certain character or effect was a really bad special effects job. We generally do a pretty good job of that stuff now, but look at the computer power it takes to pull it off. We're not getting that in gaming machines anytime soon.
I think the issue is more one of market demand. Who cares if you can put 8 GB on a stick if 99% of the potential customers are running an OS that can barely handle 4 GB ?
Sure, there's the high end Unix crowd that would go crazy over that stuff, but trying asking SGI or the Itanium department how profitable it is to cater to that market nowadays.
Also, don't forget that Windows hasn't had a major upgrade since 2001. Windows upgrades are a large factor in how much RAM people need.
I hadn't checked the most recent financial statements. I knew both companies were expecting to be profitable right around now, give or take a quarter. Last I heard Sirius was expecting it to come early this year, with XM expecting it the end of last year.
Sirius's numbers have been growing faster, but it's not all due to Howard Stern. Sirius's subscriber totals count customers in free trial periods, whereas XM counts paying subscribers. Cars that come with Sirius preinstalled get a free subscription, sometimes up to 3 years. Not only does Sirius count those members, but they also count the cars in the lot before they are sold. Sirius has made more deals with car manufacturers over the past year, which ends up increasing their subscriber counts.
The real problem is if they are trying to merge it means neither is profitable, no shock there, so both will likely go under. For my two cents I say let them merge if you do one thing. Separate the pricing so you have your choice. I have no interest in supporting shock jocks, Oprah neither. I say offer package deals where you can get talk or music or a bundle of both. If you cant have competition at least have choice.
Sirius wants to merge because their market cap is significantly higher than XM's, but XM has more subscribers and is closer to being profitable. They feel they have a chance of buying XM below value, and improving their position in the process.
XM would only agree to it if Sirius gave them an offer well above what you'd expect by comparing market caps.
The quick launch pref has been buggy since the Mozilla days. Generally, if SeaMonkey/Mozilla crashes with quick launch on, it won't go into quick launch mode the next time you start it. If you go into preferences, uncheck quick launch and then check it again, quick launch will turn on. Quick launch works fine though when starting up at login or after a clean shutdown.
I'm guessing a flag is stored on disk somewhere when quicklaunch starts which doesn't get cleared on a crash.
VCRs automatically do some cleanup on the incoming signal. Macrovision is a filter applied to the output of a device such as a DVD player which results in no visible change in the picture, but confuses the cleanup processing of the VCR, resulting in a signal with colors that gradually cycle from light to dark and back repeatedly.
Blu-Ray and HD-DVD use the same video codecs. Both support 1080p. Any difference in quality is due to the settings chosen by the person creating the disc, and has nothing to do with the format. Obviously if you crank the bit rate up really high, the HD disc will run out of space before the Blu-Ray disc, but you could split the movie across multiple discs at that point.
For the authoring system, that's entirely a software issue. Both sides were willing to comprise and use the other side's software to avoid a format war, but neither would budge on the physical disc format. From what I've heard though, neither system is clearly better than the other. Each has its own favorable aspects.
Players for both formats play normal DVD discs. If there are any compatibility issues, it's player specific. That stuff always happens with early hardware. Remember when The Matrix first came out on DVD and half the players available had issues with it?
The only significant differences between the formats is the storage. Sony designed the format with the express goal of making the best possible format. This ended up requiring completely new manufacturing plants to make the discs. Some movie studios didn't like the idea, and tried to design a format that required minimal tweaks to their existing manufacturing lines. Hence the storage space sacrifices.
There isn't any significant difference in the cost of manufacturing either type of disc. Blu-Ray costs more to make *now* because the studios have to make up the cost of building new manufacturing plants. If Blu-Ray becomes popular, those initial costs will be made up, making either type of disc equal in price.
From what I've heard, Nintendo has two offices in Redmond. One is the corporate office, which is similar to what you're talking about. Marketing, sales, localization, etc. The other is Nintendo Software Technology. That office actually makes games. They opened in time for GameCube development. They did WaveRace: Blue Storm, 1080 Snowboarding: Avalanche, and Metroid Prime: Hunters. Probably more, but that's all I'm certain of.
That's the theory, but things never work out that way in practice. I've been bitten several times by these issues. If you assign a person to refactor large amounts of code they don't have the means to test (honestly, who has every network card the kernel supports?), they will make mistakes.
Don't just think entire drivers - think drivers that support several similar but not quite identical devices. Those are the ones that really suffer from the constant changes. The less popular models end up getting lost in the shuffle.
- faster, a lot faster fixes than for closed-source drivers;
The biggest cause for this is testing. Most companies will not release an updated driver until it has gone through quality assurance testing. Open source developers do not have dedicated QA teams. Yes, companies like RedHat do, but that's for their own releases, not for the mainstream ones.
Sure, some companies don't put any priority on bug fixing. But most hobbiest programmers prefer to focus their time on new featues rather than fixing obscure bugs. That's an issue with the people, not the process.
- fixes at all. a lot of vendors just don't fix some problems;
And most open source projects get abandoned. Look around at SourceForge some time. Don't say the code is available, fix it yourself. On a large code base, that's just not a realistic option.
- no need to toss out hardware because vendor has decided to stop supporting it (which happens pretty fast in most cases);
That's much less of an issue if the driver API is stable.
- faster driver availability for new kernel versions (no need to wait for nvidia, for example, to release new version);
That's only an issue because the driver API isn't stable. If there was a stable binary interface, you wouldn't need a new driver from nvidia every time the kernel changed.
- availability of new functionality (like power management and probably a bunch of other things);
Valid point.
- i suppose also changes that improve security, stability & performance are easier to make.
Yes, incompatible changes are much easier to make. But it's only necessary because of the current development model. Kernel developers aren't concerned about getting things right the first time because there is very little penalty to changing things later. Moving towards a stable API implies doing much more planning early on. It's quite likely that this would lead to less work being done overall, as problems are cheaper to fix the earlier they are found.
- driver availability on other architectures (maybe less important for average user, but there were some problematic drivers on amd64 - and i suppose, having usb hw work on some more exotic hw would significantly ease the quest to build home complected low-heat and silent media computer, for example)
This is the biggest reason to care about open source drivers, assuming that you don't care about the politics.
- much better support (vendors tend to be less interested how their product works than most kernel driver devs) and bigger chances to diagnose the problem - solving problems with proprietary kernel modules is no fun...
This is only true due to the small percentage of computer users using Linux. The current Linux userbase also tends to be rather tech savy, making support easier. If Linux was to start becoming mainstream, the support picture would be completely different.
Yes, and that someone most likely does not have the device your driver is for, making it pure luck whether they do the fix properly or not. Your odds of the change being done properly are much better if your hardware is popular, but still not guaranteed.
The 6 million polygons number everyone throws about wasn't a "real world scenario". It was a simple, unoptimized tech demo with music, physics, and AI. Nintendo meant it as a performance minimum - you can get 6 million polygons per second without even trying hard.
MS and Sony's numbers were the highest peak performance you could obtain. Something like something triangles with no lighting, shading, or texturing. MS's numbers might've been ever so slightly more realistic. I know that used to flash a demo with a scene full of butterflies (a whopping 2 polygons each... maybe 4).
Slashdot Mirror
The trade off to have 32 registers was probably not worth the die space and extra complexity. Having 16 probably gave most of the benefit, and having 32 provided diminishing returns.
At least with the general purpose registers, AMD wanted to go to 32, but couldn't do it without changing the instruction set. I'd assume the same thing applies to the SSE registers.
The spider ball boss is easy once you figure out the trick. There are 4 bomb slots - 2 on each side of the way up. The more damage you do, the faster the boss attacks. The trick is to hit the outer bomb slots first, as they're much harder to get to than the inner ones. If you do that, you can usually rush up to the inner ones without much difficulty.
That's probably the first reasonable response I've ever heard in respect to the general right interpretation of the second amendment.
Anyway, if that is the intended interpretation, it would mean that women and people with disabilities don't have the right to have a gun. Which goes back to my original statement that I don't think they intended the amendment to be a blanket right for everyone to have whatever guns they want.
You're so dead set on your opinion that you're not even paying attention to what anyone else says. I'm not sure why I'm bothering, as you completely ignore or misinterpret anything that doesn't fit your belief.
First off, you insisted that "The ACLU doesn't think the second amendment means the same thing you do" wasn't a valid reason for them to not defend your interpretation of it. If you're not going to consider that valid, then nothing is valid. It really was a waste of time to bother saying anything, but I did hope to at least get a somewhat intelligent response back. I was just hoping to get some insight into what you would consider valid reasoning.
I did read a few pages of your link, and found it to be just like all the others on the topic. They conclude "well regulated militia" means "everyone". Which goes back to what I said before, why bother to include something that could possibly be misunderstood if it adds no extra meaning? You could chop out the first half of the amendment without losing any meaning, and you'd make the intention a lot more clear.
I also never said that I believed the second amendment was written to provide state militias. I simply said that I believed it wasn't intended to be a blanket guarantee that everyone can have guns without restriction. I think the first half of the amendment is worded very awkwardly and is completely out of place with the rest.
As to your comment about bills relating to the states being at the end, that's just molding things to fit your view. Putting aside any questions about the second amendment, the only one referencing the states is the 10th amendment, which is a catchall. It just makes the most sense to put that at the end. I really wouldn't infer anything else about that placement.
The point of a gun is to cause harm. The ACLU doesn't defend things that cause harm to others. (Let's just not get into the whole abortion thing for now...) They'll defend freedom of speech, but they won't fight for you to be able to yell fire in a crowded building. Maybe you'd have a case if they ignored an attempt to completely ban guns, but generally, the gun control issues come down to banning high end weapons you really don't want some random guy on the street to have, or laws to avoid impulse buying.
As to the actual ammendment, here's the text:
A well regulated Militia, being necessary to the security of a free State, the right of the people to keep and bear Arms, shall not be infringed.
If that was intended as a general right for everyone then why didn't they simplify it?
The right of the people to keep and bear Arms shall not be infringed.
That would've done the job fine and avoided all confusion on the issue. Look at the rest of the Bill of Rights - there's isn't any unnecessary language anywhere. The Constitution in general isn't overly wordy. I find it very hard to believe that they chose to add that much extra text for no reason.
This changed in Vista for one reason: DRM. Microsoft made it so that certain processes, namely wmplayer.exe and halo2.exe, cannot be a target of the debugger API calls for obvious reasons.
wmplayer.exe is obvious why they'd care, but I can't see a reason for halo2.exe. I guess it would make it harder to cheat in an online game, but I can't see changing such a core part of the OS for an FPS game. Maybe if you were talking something like WoW you could make more of a case, but still... or did you have something else in mind for it?
Why would they do this? Just make the default be the deletion of the symlink instead of the target. And why would you need this permission if you already have permission to modify the directory (to create the link) and the permission to read the target? Instead they go and take the ability of normal users (meaning most people) to use symlinks by default.
I used to work on GEOS years ago. The OS had long supported symlinks but never shipped with any user accessible way to create them. Eventually someone decided to enable the UI to create them. We caught some weird issues. Users liked to do things like create links to the root of their C drive and place them on the desktop, or other similar things that resulted in symlinks pointing to directories higher up the directory path. The Find File utility wasn't coded with links in mind, so any attempt to search your C: drive took forever. It would just keep recursing over the same paths until it reached the path length limits.
So, the short story is symlinks let you create cycles in the directory tree. If you unleash this onto applications not designed to handle it, you will run into issues that will not at all be obvious to the users.
If you mean brick and mortar, I don't have the money to pay somebody to wait in line outside for hours on a Saturday night.
Yes, brick and mortar is what I was referring to. It's not that hard to get a Wii if you try. Shipments come in pretty often. Talk to the people at a store near you and find out the trends.
Anyway, if you feel the way to get something is to pay someone else to go to the store for you, that has very little to do with the price of the system. You might as well complain that Wiis are expensive because some random guy on eBay wants $1000 for a Wii.
I'll show you as soon as you show me where in the Constitution it authorizes HUD, Medicare, Medicaid, Social Security and everything else our gov't does that is not specifically spelled out in the Constitution.
I believe it would be this:
Section 8. The Congress shall have power to lay and collect taxes, duties, imposts and excises, to pay the debts and provide for the common defense and general welfare of the United States; but all duties, imposts and excises shall be uniform throughout the United States;
Specifically, the part about providing for the general welfare of the United States.
I wasn't talking about Wii Sports. I was talking about not being able to mail-order a Wii without having to buy five $50 games of the retailer's choice. What if I only want Zelda and WarioWare?
Try a retail store then. Your problem is with your choice of stores, not with the pricing of the Wii.
In the US, taxes vary completely from state to state. Some states break it down further, and have different taxes per county. New York City adds its own taxes onto the state taxes. New Jersey cuts the sales tax in crappy areas, in an effort to encourage people to spend their money there and bring more money into the local economy. Other states have other oddities.
Also, what's taxed varies from state to state. In New Jersey, prepared foods are taxed, but unprepared foods are not. This gets some weird corner cases - if you buy a Snapple bottle from a refrigerator you pay tax on it, but not if you buy one off a shelf.
In short, US sales taxes are even more complicated than income taxes. It's not as well known because most people don't have to deal with more than a couple different tax zones with any regularity.
I don't think we are trying to avoid realism in games, I think we are trying to experience things we can't in real life. Like being in a war, crashing a car or speeding lightning fast, or flying a spaceship.
I'm with you so far...
All these things I want to be as realistic as possible, because I really do not want or can not do them in real life.
And here's where we diverge. I don't care if the world is realistic or not. I care that everything feels like it belongs together.
In most modern games, we get great looking backgrounds (as long as we avoid things like trees), but people with plastic skin. It just looks wrong.
Once you start breaking from reality, you can design your world around what you can do. If you do a brightly colored game, you can usually get away with simpler skin tones that render easily. It's also easier to get away with strange scenarios, as long as they are fun. Or you can toss everyone in space suits and design "realistic" aliens with designs that render well.
It's like when you watch a movie and it's really obvious that a certain character or effect was a really bad special effects job. We generally do a pretty good job of that stuff now, but look at the computer power it takes to pull it off. We're not getting that in gaming machines anytime soon.
I think the issue is more one of market demand. Who cares if you can put 8 GB on a stick if 99% of the potential customers are running an OS that can barely handle 4 GB ?
Sure, there's the high end Unix crowd that would go crazy over that stuff, but trying asking SGI or the Itanium department how profitable it is to cater to that market nowadays.
Also, don't forget that Windows hasn't had a major upgrade since 2001. Windows upgrades are a large factor in how much RAM people need.
I hadn't checked the most recent financial statements. I knew both companies were expecting to be profitable right around now, give or take a quarter. Last I heard Sirius was expecting it to come early this year, with XM expecting it the end of last year.
Sirius's numbers have been growing faster, but it's not all due to Howard Stern. Sirius's subscriber totals count customers in free trial periods, whereas XM counts paying subscribers. Cars that come with Sirius preinstalled get a free subscription, sometimes up to 3 years. Not only does Sirius count those members, but they also count the cars in the lot before they are sold. Sirius has made more deals with car manufacturers over the past year, which ends up increasing their subscriber counts.
The real problem is if they are trying to merge it means neither is profitable, no shock there, so both will likely go under. For my two cents I say let them merge if you do one thing. Separate the pricing so you have your choice. I have no interest in supporting shock jocks, Oprah neither. I say offer package deals where you can get talk or music or a bundle of both. If you cant have competition at least have choice.
Sirius wants to merge because their market cap is significantly higher than XM's, but XM has more subscribers and is closer to being profitable. They feel they have a chance of buying XM below value, and improving their position in the process.
XM would only agree to it if Sirius gave them an offer well above what you'd expect by comparing market caps.
And then some clever entrepeneur [sic] makes a cheap receiver that receives both, but for free.
And he'll go to jail. Do you have the slightest idea how these things work?
Yeah, just like how the guy who cracked DVD encryption went to jail, and everyone stopped copying DVD movies.
The quick launch pref has been buggy since the Mozilla days. Generally, if SeaMonkey/Mozilla crashes with quick launch on, it won't go into quick launch mode the next time you start it. If you go into preferences, uncheck quick launch and then check it again, quick launch will turn on. Quick launch works fine though when starting up at login or after a clean shutdown.
I'm guessing a flag is stored on disk somewhere when quicklaunch starts which doesn't get cleared on a crash.
VCRs automatically do some cleanup on the incoming signal. Macrovision is a filter applied to the output of a device such as a DVD player which results in no visible change in the picture, but confuses the cleanup processing of the VCR, resulting in a signal with colors that gradually cycle from light to dark and back repeatedly.
But 1280x1024 makes the pixels look short and fat. 1280x960 helps keep them in shape.
Blu-Ray and HD-DVD use the same video codecs. Both support 1080p. Any difference in quality is due to the settings chosen by the person creating the disc, and has nothing to do with the format. Obviously if you crank the bit rate up really high, the HD disc will run out of space before the Blu-Ray disc, but you could split the movie across multiple discs at that point.
For the authoring system, that's entirely a software issue. Both sides were willing to comprise and use the other side's software to avoid a format war, but neither would budge on the physical disc format. From what I've heard though, neither system is clearly better than the other. Each has its own favorable aspects.
Players for both formats play normal DVD discs. If there are any compatibility issues, it's player specific. That stuff always happens with early hardware. Remember when The Matrix first came out on DVD and half the players available had issues with it?
The only significant differences between the formats is the storage. Sony designed the format with the express goal of making the best possible format. This ended up requiring completely new manufacturing plants to make the discs. Some movie studios didn't like the idea, and tried to design a format that required minimal tweaks to their existing manufacturing lines. Hence the storage space sacrifices.
There isn't any significant difference in the cost of manufacturing either type of disc. Blu-Ray costs more to make *now* because the studios have to make up the cost of building new manufacturing plants. If Blu-Ray becomes popular, those initial costs will be made up, making either type of disc equal in price.
From what I've heard, Nintendo has two offices in Redmond. One is the corporate office, which is similar to what you're talking about. Marketing, sales, localization, etc. The other is Nintendo Software Technology. That office actually makes games. They opened in time for GameCube development. They did WaveRace: Blue Storm, 1080 Snowboarding: Avalanche, and Metroid Prime: Hunters. Probably more, but that's all I'm certain of.
That's the theory, but things never work out that way in practice. I've been bitten several times by these issues. If you assign a person to refactor large amounts of code they don't have the means to test (honestly, who has every network card the kernel supports?), they will make mistakes.
Don't just think entire drivers - think drivers that support several similar but not quite identical devices. Those are the ones that really suffer from the constant changes. The less popular models end up getting lost in the shuffle.
- faster, a lot faster fixes than for closed-source drivers;
The biggest cause for this is testing. Most companies will not release an updated driver until it has gone through quality assurance testing. Open source developers do not have dedicated QA teams. Yes, companies like RedHat do, but that's for their own releases, not for the mainstream ones.
Sure, some companies don't put any priority on bug fixing. But most hobbiest programmers prefer to focus their time on new featues rather than fixing obscure bugs. That's an issue with the people, not the process.
- fixes at all. a lot of vendors just don't fix some problems;
And most open source projects get abandoned. Look around at SourceForge some time. Don't say the code is available, fix it yourself. On a large code base, that's just not a realistic option.
- no need to toss out hardware because vendor has decided to stop supporting it (which happens pretty fast in most cases);
That's much less of an issue if the driver API is stable.
- faster driver availability for new kernel versions (no need to wait for nvidia, for example, to release new version);
That's only an issue because the driver API isn't stable. If there was a stable binary interface, you wouldn't need a new driver from nvidia every time the kernel changed.
- availability of new functionality (like power management and probably a bunch of other things);
Valid point.
- i suppose also changes that improve security, stability & performance are easier to make.
Yes, incompatible changes are much easier to make. But it's only necessary because of the current development model. Kernel developers aren't concerned about getting things right the first time because there is very little penalty to changing things later. Moving towards a stable API implies doing much more planning early on. It's quite likely that this would lead to less work being done overall, as problems are cheaper to fix the earlier they are found.
- driver availability on other architectures (maybe less important for average user, but there were some problematic drivers on amd64 - and i suppose, having usb hw work on some more exotic hw would significantly ease the quest to build home complected low-heat and silent media computer, for example)
This is the biggest reason to care about open source drivers, assuming that you don't care about the politics.
- much better support (vendors tend to be less interested how their product works than most kernel driver devs) and bigger chances to diagnose the problem - solving problems with proprietary kernel modules is no fun...
This is only true due to the small percentage of computer users using Linux. The current Linux userbase also tends to be rather tech savy, making support easier. If Linux was to start becoming mainstream, the support picture would be completely different.
Yes, and that someone most likely does not have the device your driver is for, making it pure luck whether they do the fix properly or not. Your odds of the change being done properly are much better if your hardware is popular, but still not guaranteed.
The 6 million polygons number everyone throws about wasn't a "real world scenario". It was a simple, unoptimized tech demo with music, physics, and AI. Nintendo meant it as a performance minimum - you can get 6 million polygons per second without even trying hard.
MS and Sony's numbers were the highest peak performance you could obtain. Something like something triangles with no lighting, shading, or texturing. MS's numbers might've been ever so slightly more realistic. I know that used to flash a demo with a scene full of butterflies (a whopping 2 polygons each... maybe 4).