Of course shadows aren't perfect in current 3D systems. I never said or implied otherwise. The previous poster seemed to think the point of raycasting was to get better shadows, when that's really one of the most obvious downsides to raycasting.
Personally, I'd have people pay more attention to textures and make things more efficient than concentrate on a new shiny method to improve shadows.
Shadows are actually the glaring weakness in ray tracing setups. Ray tracing tends to lead to shadows going instantly from completely solid to non-existant. You don't get the soft edges.
Ray tracing tends to be great for reflections though.
Why does this compile? Because unless you put the word explicit before the constructor taking an int as an argument C++ automatically converts any int to a ThreadPool whenever given the choice. The C++ community generally blames these sort of errors on the user for not being an expert in the language. This is stupid. In a sane language the language should do the most obvious and least damaging thing by default and require explicit permission to do anything else. In this case the language does the most bizarre and stupid thing unless you explicitly tell it not to.
That's basic C++, not an expert topic by any means. The language would be a pain to use if it didn't do things like that. For an example in your integer conversion, if you're doing any sort of custom math class (say a Big Number class or a Fixed Point class), you really the compiler to automatically convert ints into the proper class. If you're working with strings, you want the compiler to promote a string literal into an std::string when necessary. You came up with a contrived example to explain your point. If you've got a case like that where the normal behavior doesn't make sense, then go ahead and use the explicit keyword, that's why it's there.
This code will call A's version of the doSomething method and the fail to call B's destructor. Why? because virtual functions have a (very) slight cost in memory and performance so C++ does the most efficient thing by default rather than the least stupid. If you want things to work correctly you the coder have to explicitly tell the compiler to make the destructor and method virtual with the virtual keyword.
I resisted C++ for a long time, as I thought not having virtual functions by default defeated the point of using classes. I'm still not thrilled with it, but I've learned to take it as it is. I think C++ is a terrible language for designing core libraries, but it is rather useful when designing applications. At that point you know if it makes sense or not for something to be virtual.
If everything was virtual, I'm not sure it would necessarily mean C++ was a better language, I think it would just change what projects people use it for. Right now it's a great middle ground for embedded and low end systems - you get to pick and choose what parts of OOP benefit your project. Again, going back to one of my examples before, if you wrote a Fixed Point math class and had everything virtual, that would significantly hurt your performance.
If you don't believe me take a look at mozilla.org's Portability Guide. I can sum it up for you if you'd like: if you have to write C++ please use as little of it as possible. Oh and another list of all the stupid reminders you have to continually provide to compiler to minimally reduce unexpected behavior. I actually work on embedded software at NASA and they have pretty much the same thing.
To be fair, a lot of that is due to Mozilla.org's desire to run on old Unix platforms. The big Unix developers stopped putting much effort into their compilers in the early to mid nineties, before a lot of the fancier stuff in C++ existed, let alone was standardized.
And that's a problem with the entire intellectual property battle. Batting an old woman over the head and taking her purse is clearly, to anyone with a conscience, illegal and wrong. Clicking a link on South Park Zone doesn't look or feel or act any different than clicking a link on Comedy Central Zone, so how is the average person supposed to come to the decision that it's "wrong"?
If you look at their terms & disclaimers pages, it's rather sketchy. Not how they make a big point out of how they don't actually host video and only link to other sites.
Comedy Central could probably go after them for using the South Park trademarks & copyrighted images at the very least if they chose to.
It's like the old question about how someone is supposed to instinctively know copying a Music CD is illegal when the same company that makes the Music CD also makes and sells blank CDs for copying?
Every major label CD you buy in a store has warnings on it about it being illegal to copy it. Some CDs have it on the back, others have it on the liner notes. Of course, the warnings go too far and claim things to be illegal that aren't, but that's a different issue.
Uh no. It's not the same thing at all. The author of an academic paper does not get paid for publishing, and often has to pay for the service. (Though usually the university the author works with pays this on his behalf)
I was referring to books, where I've seen distribution issues come up. I never saw any issues with papers while I was in college.
If what you say is true, then yeah, that is really messed up.
Wow -- it's all almost as illogical as academia, where professors have to beg permission from publishers to distribute their own works to students. Almost.
It's the same exact thing. The author gives the publisher some or all rights to their work in exchange for money and/or royalties. If they have a problem with it, they should have negotiated for more rights.
When you sit down you easily see the position of the keyboard
Only if you look at the keyboard when you go to use it. I don't. It's usually on a draw under the desk. I sit down then pull it out and place my hands without looking. I also frequently take my hands off the keyboard while at the computer, using the bumps to place my hands back on without looking.
Re:In the case of the DS, I get the 'flip-ability'
on
Meet the Laptop of 2015
·
· Score: 2, Informative
there is usually no inherent advantage to having multiple physically seperated screens, when you can just logically partition a single screen as necessary.
The dual screens reduces processing power needs. The 2D hardware on the DS requires far, far less power than the 3D hardware, and is also much cheaper to make. The DS design has 2 2D engines and 1 3D engine. Doing one screen would've required bumping up the power of the 3D engine substantially, and probably would require more RAM as well.
Really? So I guess we'll see FPS players migrating to Wii now, huh?
Of course not. The kind of FPS player that makes such a big deal over the controls are usually graphics whores, so they wouldn't consider the Wii. Not to mention online play on the Wii sucks so far.
Seriously, what kind of mouse do you use if you think the wiimote is on par with a mouse? The And what keyboard setup do you use if you think those buttons compare to a keyboard?
Try playing Metroid Prime 3 or Medal of Honor Heroes 2. Earlier FPS games hadn't figured out how to properly use the Wiimote.
And it's buttons that's exactly the issue with a keyboard - using an analog stick gives you far better control over movement. Sure, they're not great for aiming, but nothing tops them for moving.
I have XM, and it came built in to my car, so I can't move the antenna. It's mounted to the roof.
Mine's on the roof too, directly over the rearview mirror. Rarely have an issue. I know somewho who borrowed an XM radio from someone, tossed the antenna by the rear window (inside the car), and drove cross country like that without any problems worth mentioning. I guess you must have a lousy antenna.
I'd say about half the time bridges cause a gap (8-9 seconds after I've passed the bridge. My stereo must buffer.)
XM's two satellites broadcast with one 6 seconds (I think) behind the other, so if you can pick up both satellites, you get a buffer.
Interesting how Murdoch can take over the WSJ with all the holdings in news media (print/airwaves) with nary a hitch in the proceedings, but something outside the gov't (and lobbyists') regulatory control takes more than a year.
It's not outside the government's control. The FCC licenses for the spectrum used by satellite radio was granted on the condition that XM and Sirius never merge. They could have merged easily if one company was willing to give up their spectrum, but that would've made the company worthless.
I'm hoping that if the merger goes through and digital FM takes off, they will merge the duplicate satellite channels into CD quality channels.
I'm afraid of them attempting to do that. There aren't a lot of duplicate channels. There's some duplicates in the talk stations, but those are already rather low bitrate, so you wouldn't gain a lot from dropping them. You could merge the top 40 type stations as well, but that's only a few.
Past that, XM and Sirius have rather different philosophies on programming. Sirius tends to stick mostly to the hits, making them a commercial free version of normal FM. XM mixes in album tracks and occasionally rather obscure stuff. There's a LOT more variety to XM playlists. I know I wouldn't consider paying for Sirius style stations, but I gladly do for XM.
I hate commercials too... But I hate losing signal when I go under a bridge, or when there's a big tree next to the road, or when it's cloudy. I hate it even more when I paid $120/year for that signal...
Try moving your antenna. I have to be driving through rather heavy tree coverage to lose the signal. Bridges very rarely cause a dropout.
If you have Sirius, that might be the problem. XM's satellites are in geosynchronous orbit, so if you're in the continental US, you'll always have full visibility of them. Sirius has 3 satellites in a weird orbit. You'll usually have 2 satellites visible, but sometimes only one.
If you're looking for continuous, commercial free, specialized radio channels with national coverage.... is there really any competition left after this merger?
Well, there doesn't seem to be THAT much of a market for that service, or we wouldn't be talking about this merger. They did it because having to compete with each other sucked away both companies ability to make money.
That said, as an XM subscriber, I'm afraid of the merger. Sirius stations basically follow the FM model, but with a slightly narrower focus. While XM has a few top 40 type stations, most of their stations have far more variety in their playlists and will gladly do things like play non-single tracks off an album or the occasional b-side. That's the stuff I care about.
I was talking specifically about home consoles not handhelds, which is why I did not include the PlayStation protable, the DS or GameBoy. Point was that Sony is still leading the home console sales, even if it's not the PS3 specifically.
Ah, it's been a while since I paid attention to the PS2 numbers. Normally PS2 + PS3 combined aren't that high, but the PS3 sales have gone up in the past few months, so it does make sense.
That's interesting that Nintedo has been able to sell 1.5 million units in one month and less than half a million in another. Does Nintendo produce more units in certain months than others? Normally supply constraints go up as time goes on not down. But then again that's only if they are physical constraints.
Overall Wii production is at a steady 1.5 million per month worldwide. Obviously you can see some fluctuation in monthly sales based on factors like how many were shipped on the 31st of a month vs how many went out the following day. More significantly, they constantly change the allocation of systems between Japan, US, Europe, and Australia. In November and December, most of the production output went to the US to handle US Christmas shopping. In January and February, Japan got more due to the launch of Smash Bros Brawl. March's shipments will probably be either more evenly distributed or in favor of the US due to Smash Bros coming out here.
While most of these points are available on any platform the PC still has an edge on interface. Keyboards, mice and flightsticks all offer a more advance UI than thumb levers.
If you stick to FPS and RTS games, sure. Try playing anything that requires precision running and jumping. Or a fighter.
If you play an FPS with a Wiimote + Nunchuck, you've got aiming on par with what a mouse offers but far better control of your character's movement.
So does that mean every XBox 360 now has just a plain DVD drive? Okayyy..... I can understand Nintendo doing that because they were never aiming to have some super high-powered games machine that can do everything but make tea, but Microsoft?
Yeah... HD-DVD drives weren't going to be ready until a year after the 360 was intended to launch, and would've added ~$200 to the system cost.
The PS3 was delayed, cost a lot more than any other system, and was sold at a large loss due to the inclusion of Blu-Ray. Blu-Ray wasn't there to make a better game system, it was put it at the expensive of the game system to ensure it would win the movie format war.
I just find it funny when someone talks about the dominance of a console in the market and how Microsoft or Nintendo has taken down the Playstation giant, with out actually realizing the the Playstation has continual out sold both competitors month in and month out since the brand was first released.
That's because Sony uses the Playstation brand on all their gaming products. Try looking at the Nintendo brand. The GameBoy and DS have been selling extremely well going back about 20 years now.
Also, there's a limit to how much you can say about Wii sales. Wiis sales are limited by supply, and are expected to be for at least another 6 months (according to GameStop's recent financial reports). PlayStation sales aren't.
Both the top selling current game units, the Wii and NDS are both pirate magnates and trivial to mod accordingly. Nintendo left the Wii open to get the masses buying the boxes knowing most people will get the modded and download "backup" games,
Very, very few people mod their consoles. Don't forget to look at the Wii market - a large portion of it is people who would never consider any other system and just want Wii Sports and the like. They're not even thinking about pirating games.
the NDS requires nothing more than a plug in cartridge.
Yeah, shocking on a system that takes games off a cartridge. The games have encryption built in, which took about 2 years to get completely hacked.
Binary drivers not being provided by either manufacturer for FreeBSD presently I'm inclined to go with the chips more likely to have reliable drivers in the future.
FYI, NVidia has been offering FreeBSD drivers for years, although I think they're limited to 32 bit x86. Likewise for Solaris drivers.
Linux and Windows have both 32 and 64 bit options.
I haven't used Red Hat, but every time I update Fedora, package dependencies get messed up. Maybe RHEL is better, but then again, last time I checked (admittedly a while ago), they recommended reinstalls rather than upgrades, so I'm not holding out hope there.
I gave up on Mandrake right before it became Mandriva due to each upgrade messing up the hardware config. Not only did I have to reconfigure my SoundBlaster Live and my TV card with every point release, but it decided to create new icons for the devices on the desktop with different names, making it more confusing. I wasn't able to find a way to permanently remove the old ones, so it built up quite a collection of icons after a few upgrades.
And what would these instructions be? It doesn't make any sense for there to be instructions leaving it undefined.
I've never seen any reference in the Intel manuals to anything that can change it other than STD (Set Direction) and CLD (Clear Direction). (And for the pedantic ones out there, popf, the context switch related instructions, etc... but it's still defined behavior.)
On the other hand: the instructions affected by this aren't used very much, so if you want optimizations, a good candidate would be to not clear the flag unless it is needed.
That's how the old code worked - it cleared the flag before using instructions impacted by it. However, you can't say the affected instructions aren't used much.
All the string & bulk data instructions care about the direction flag. Unless your compiler generates really bad code, I'd expect memcpy, memmove, memcmp, strcpy, strlen, strcat, etc all to be impacted. realloc most likely would either call memmove or would use similar code as well if it has to move your memory to enlarge the allocation.
Also, things like assigning one struct to another is essentially an inline memcpy, hence that could be affected as well.
With the compiler no longer clearing the direction flag, any of those functions would misbehave if called with the direction flag set.
and future code could both conform to the new ABI *and* avoid the overhead of unnecessary instructions to clear the flag when it is not being used.
The flag only ever gets set when it's needed. Needing the flag set is rather rare, whereas most code needs it clear. Hence why the idea is that the code that sets the flag should be responsible for clearing it, rather than having to constantly clear the flag every time it comes into play.
Slashdot Mirror
Of course shadows aren't perfect in current 3D systems. I never said or implied otherwise. The previous poster seemed to think the point of raycasting was to get better shadows, when that's really one of the most obvious downsides to raycasting.
Personally, I'd have people pay more attention to textures and make things more efficient than concentrate on a new shiny method to improve shadows.
Shadows are actually the glaring weakness in ray tracing setups. Ray tracing tends to lead to shadows going instantly from completely solid to non-existant. You don't get the soft edges.
Ray tracing tends to be great for reflections though.
Why does this compile? Because unless you put the word explicit before the constructor taking an int as an argument C++ automatically converts any int to a ThreadPool whenever given the choice. The C++ community generally blames these sort of errors on the user for not being an expert in the language. This is stupid. In a sane language the language should do the most obvious and least damaging thing by default and require explicit permission to do anything else. In this case the language does the most bizarre and stupid thing unless you explicitly tell it not to.
That's basic C++, not an expert topic by any means. The language would be a pain to use if it didn't do things like that. For an example in your integer conversion, if you're doing any sort of custom math class (say a Big Number class or a Fixed Point class), you really the compiler to automatically convert ints into the proper class. If you're working with strings, you want the compiler to promote a string literal into an std::string when necessary. You came up with a contrived example to explain your point. If you've got a case like that where the normal behavior doesn't make sense, then go ahead and use the explicit keyword, that's why it's there.
This code will call A's version of the doSomething method and the fail to call B's destructor. Why? because virtual functions have a (very) slight cost in memory and performance so C++ does the most efficient thing by default rather than the least stupid. If you want things to work correctly you the coder have to explicitly tell the compiler to make the destructor and method virtual with the virtual keyword.
I resisted C++ for a long time, as I thought not having virtual functions by default defeated the point of using classes. I'm still not thrilled with it, but I've learned to take it as it is. I think C++ is a terrible language for designing core libraries, but it is rather useful when designing applications. At that point you know if it makes sense or not for something to be virtual.
If everything was virtual, I'm not sure it would necessarily mean C++ was a better language, I think it would just change what projects people use it for. Right now it's a great middle ground for embedded and low end systems - you get to pick and choose what parts of OOP benefit your project. Again, going back to one of my examples before, if you wrote a Fixed Point math class and had everything virtual, that would significantly hurt your performance.
If you don't believe me take a look at mozilla.org's Portability Guide. I can sum it up for you if you'd like: if you have to write C++ please use as little of it as possible. Oh and another list of all the stupid reminders you have to continually provide to compiler to minimally reduce unexpected behavior. I actually work on embedded software at NASA and they have pretty much the same thing.
To be fair, a lot of that is due to Mozilla.org's desire to run on old Unix platforms. The big Unix developers stopped putting much effort into their compilers in the early to mid nineties, before a lot of the fancier stuff in C++ existed, let alone was standardized.
And that's a problem with the entire intellectual property battle. Batting an old woman over the head and taking her purse is clearly, to anyone with a conscience, illegal and wrong. Clicking a link on South Park Zone doesn't look or feel or act any different than clicking a link on Comedy Central Zone, so how is the average person supposed to come to the decision that it's "wrong"?
If you look at their terms & disclaimers pages, it's rather sketchy. Not how they make a big point out of how they don't actually host video and only link to other sites.
Comedy Central could probably go after them for using the South Park trademarks & copyrighted images at the very least if they chose to.
It's like the old question about how someone is supposed to instinctively know copying a Music CD is illegal when the same company that makes the Music CD also makes and sells blank CDs for copying?
Every major label CD you buy in a store has warnings on it about it being illegal to copy it. Some CDs have it on the back, others have it on the liner notes. Of course, the warnings go too far and claim things to be illegal that aren't, but that's a different issue.
Uh no. It's not the same thing at all. The author of an academic paper does not get paid for publishing, and often has to pay for the service. (Though usually the university the author works with pays this on his behalf)
I was referring to books, where I've seen distribution issues come up. I never saw any issues with papers while I was in college.
If what you say is true, then yeah, that is really messed up.
This is legal, South Park Zone isn't.
Wow -- it's all almost as illogical as academia, where professors have to beg permission from publishers to distribute their own works to students. Almost.
It's the same exact thing. The author gives the publisher some or all rights to their work in exchange for money and/or royalties. If they have a problem with it, they should have negotiated for more rights.
When you sit down you easily see the position of the keyboard
Only if you look at the keyboard when you go to use it. I don't. It's usually on a draw under the desk. I sit down then pull it out and place my hands without looking. I also frequently take my hands off the keyboard while at the computer, using the bumps to place my hands back on without looking.
there is usually no inherent advantage to having multiple physically seperated screens, when you can just logically partition a single screen as necessary.
The dual screens reduces processing power needs. The 2D hardware on the DS requires far, far less power than the 3D hardware, and is also much cheaper to make. The DS design has 2 2D engines and 1 3D engine. Doing one screen would've required bumping up the power of the 3D engine substantially, and probably would require more RAM as well.
Really? So I guess we'll see FPS players migrating to Wii now, huh?
Of course not. The kind of FPS player that makes such a big deal over the controls are usually graphics whores, so they wouldn't consider the Wii. Not to mention online play on the Wii sucks so far.
Seriously, what kind of mouse do you use if you think the wiimote is on par with a mouse? The And what keyboard setup do you use if you think those buttons compare to a keyboard?
Try playing Metroid Prime 3 or Medal of Honor Heroes 2. Earlier FPS games hadn't figured out how to properly use the Wiimote.
And it's buttons that's exactly the issue with a keyboard - using an analog stick gives you far better control over movement. Sure, they're not great for aiming, but nothing tops them for moving.
I have XM, and it came built in to my car, so I can't move the antenna. It's mounted to the roof.
Mine's on the roof too, directly over the rearview mirror. Rarely have an issue. I know somewho who borrowed an XM radio from someone, tossed the antenna by the rear window (inside the car), and drove cross country like that without any problems worth mentioning. I guess you must have a lousy antenna.
I'd say about half the time bridges cause a gap (8-9 seconds after I've passed the bridge. My stereo must buffer.)
XM's two satellites broadcast with one 6 seconds (I think) behind the other, so if you can pick up both satellites, you get a buffer.
Interesting how Murdoch can take over the WSJ with all the holdings in news media (print/airwaves) with nary a hitch in the proceedings, but something outside the gov't (and lobbyists') regulatory control takes more than a year.
It's not outside the government's control. The FCC licenses for the spectrum used by satellite radio was granted on the condition that XM and Sirius never merge. They could have merged easily if one company was willing to give up their spectrum, but that would've made the company worthless.
I'm hoping that if the merger goes through and digital FM takes off, they will merge the duplicate satellite channels into CD quality channels.
I'm afraid of them attempting to do that. There aren't a lot of duplicate channels. There's some duplicates in the talk stations, but those are already rather low bitrate, so you wouldn't gain a lot from dropping them. You could merge the top 40 type stations as well, but that's only a few.
Past that, XM and Sirius have rather different philosophies on programming. Sirius tends to stick mostly to the hits, making them a commercial free version of normal FM. XM mixes in album tracks and occasionally rather obscure stuff. There's a LOT more variety to XM playlists. I know I wouldn't consider paying for Sirius style stations, but I gladly do for XM.
I hate commercials too... But I hate losing signal when I go under a bridge, or when there's a big tree next to the road, or when it's cloudy. I hate it even more when I paid $120/year for that signal...
Try moving your antenna. I have to be driving through rather heavy tree coverage to lose the signal. Bridges very rarely cause a dropout.
If you have Sirius, that might be the problem. XM's satellites are in geosynchronous orbit, so if you're in the continental US, you'll always have full visibility of them. Sirius has 3 satellites in a weird orbit. You'll usually have 2 satellites visible, but sometimes only one.
If you're looking for continuous, commercial free, specialized radio channels with national coverage .... is there really any competition left after this merger?
Well, there doesn't seem to be THAT much of a market for that service, or we wouldn't be talking about this merger. They did it because having to compete with each other sucked away both companies ability to make money.
That said, as an XM subscriber, I'm afraid of the merger. Sirius stations basically follow the FM model, but with a slightly narrower focus. While XM has a few top 40 type stations, most of their stations have far more variety in their playlists and will gladly do things like play non-single tracks off an album or the occasional b-side. That's the stuff I care about.
I was talking specifically about home consoles not handhelds, which is why I did not include the PlayStation protable, the DS or GameBoy. Point was that Sony is still leading the home console sales, even if it's not the PS3 specifically.
Ah, it's been a while since I paid attention to the PS2 numbers. Normally PS2 + PS3 combined aren't that high, but the PS3 sales have gone up in the past few months, so it does make sense.
That's interesting that Nintedo has been able to sell 1.5 million units in one month and less than half a million in another. Does Nintendo produce more units in certain months than others? Normally supply constraints go up as time goes on not down. But then again that's only if they are physical constraints.
Overall Wii production is at a steady 1.5 million per month worldwide. Obviously you can see some fluctuation in monthly sales based on factors like how many were shipped on the 31st of a month vs how many went out the following day. More significantly, they constantly change the allocation of systems between Japan, US, Europe, and Australia. In November and December, most of the production output went to the US to handle US Christmas shopping. In January and February, Japan got more due to the launch of Smash Bros Brawl. March's shipments will probably be either more evenly distributed or in favor of the US due to Smash Bros coming out here.
While most of these points are available on any platform the PC still has an edge on interface. Keyboards, mice and flightsticks all offer a more advance UI than thumb levers.
If you stick to FPS and RTS games, sure. Try playing anything that requires precision running and jumping. Or a fighter.
If you play an FPS with a Wiimote + Nunchuck, you've got aiming on par with what a mouse offers but far better control of your character's movement.
So does that mean every XBox 360 now has just a plain DVD drive? Okayyy..... I can understand Nintendo doing that because they were never aiming to have some super high-powered games machine that can do everything but make tea, but Microsoft?
Yeah... HD-DVD drives weren't going to be ready until a year after the 360 was intended to launch, and would've added ~$200 to the system cost.
The PS3 was delayed, cost a lot more than any other system, and was sold at a large loss due to the inclusion of Blu-Ray. Blu-Ray wasn't there to make a better game system, it was put it at the expensive of the game system to ensure it would win the movie format war.
I just find it funny when someone talks about the dominance of a console in the market and how Microsoft or Nintendo has taken down the Playstation giant, with out actually realizing the the Playstation has continual out sold both competitors month in and month out since the brand was first released.
That's because Sony uses the Playstation brand on all their gaming products. Try looking at the Nintendo brand. The GameBoy and DS have been selling extremely well going back about 20 years now.
Also, there's a limit to how much you can say about Wii sales. Wiis sales are limited by supply, and are expected to be for at least another 6 months (according to GameStop's recent financial reports). PlayStation sales aren't.
Both the top selling current game units, the Wii and NDS are both pirate magnates and trivial to mod accordingly. Nintendo left the Wii open to get the masses buying the boxes knowing most people will get the modded and download "backup" games,
Very, very few people mod their consoles. Don't forget to look at the Wii market - a large portion of it is people who would never consider any other system and just want Wii Sports and the like. They're not even thinking about pirating games.
the NDS requires nothing more than a plug in cartridge.
Yeah, shocking on a system that takes games off a cartridge. The games have encryption built in, which took about 2 years to get completely hacked.
Binary drivers not being provided by either manufacturer for FreeBSD presently I'm inclined to go with the chips more likely to have reliable drivers in the future.
FYI, NVidia has been offering FreeBSD drivers for years, although I think they're limited to 32 bit x86. Likewise for Solaris drivers.
Linux and Windows have both 32 and 64 bit options.
I'd love to get paid to produce documentation that doesn't actually provide any help. Where do I sign up?
I haven't used Red Hat, but every time I update Fedora, package dependencies get messed up. Maybe RHEL is better, but then again, last time I checked (admittedly a while ago), they recommended reinstalls rather than upgrades, so I'm not holding out hope there.
I gave up on Mandrake right before it became Mandriva due to each upgrade messing up the hardware config. Not only did I have to reconfigure my SoundBlaster Live and my TV card with every point release, but it decided to create new icons for the devices on the desktop with different names, making it more confusing. I wasn't able to find a way to permanently remove the old ones, so it built up quite a collection of icons after a few upgrades.
And what would these instructions be? It doesn't make any sense for there to be instructions leaving it undefined.
I've never seen any reference in the Intel manuals to anything that can change it other than STD (Set Direction) and CLD (Clear Direction). (And for the pedantic ones out there, popf, the context switch related instructions, etc... but it's still defined behavior.)
On the other hand: the instructions affected by this aren't used very much, so if you want optimizations, a good candidate would be to not clear the flag unless it is needed.
That's how the old code worked - it cleared the flag before using instructions impacted by it. However, you can't say the affected instructions aren't used much.
All the string & bulk data instructions care about the direction flag. Unless your compiler generates really bad code, I'd expect memcpy, memmove, memcmp, strcpy, strlen, strcat, etc all to be impacted. realloc most likely would either call memmove or would use similar code as well if it has to move your memory to enlarge the allocation.
Also, things like assigning one struct to another is essentially an inline memcpy, hence that could be affected as well.
With the compiler no longer clearing the direction flag, any of those functions would misbehave if called with the direction flag set.
and future code could both conform to the new ABI *and* avoid the overhead of unnecessary instructions to clear the flag when it is not being used.
The flag only ever gets set when it's needed. Needing the flag set is rather rare, whereas most code needs it clear. Hence why the idea is that the code that sets the flag should be responsible for clearing it, rather than having to constantly clear the flag every time it comes into play.