Actually, I would say that bandwidth savings would still be a biggie.. Not everyone has so much bandwidth that the difference in size would be trivial. Also, "the [processor] savings inn't that big a deal in these days of standard-issue "Gigahertz processors". If you are getting 40-CB drives, doubtless you also have a decently fast processor, right? On my 400-MHz, the extract time between bz2 and gz is negligible, only compression seems to be that much slower in bz2...
So, reinventing the wheel with different bugs shows intelligence? I'm sorry, but not using existing libraries that work well is not intelligent, simply arrogant, the "Not Invented Here" syndrome. That aside, they do a significant amount themselves, they have pretty good (perhaps the best) avi and asf parsers. Also, their input/output routines are quite flexible. That said, the gui is a piece of crap, and the compile can be a bit fickle, and the developers do occasionally exhibit degrees of arrogance. I like PythonTheater, but I'm partial:) If you think mplayer uses external code too much, you could *really* criticize PythonTheater, using avifile, smpeg, and SDL for all decoding and input output, as well as using ROX's configuration system. That is really the strength of open software, being able to use components that have been developed more and not starting from scratch.
At least open source in the Linux realm. During the economic boon, many businesses had so much money and resources, they could afford to effectively throw money away on open source, in the hopes that eventually it would provide opportunities to combat MS. But now the companies that are left are more wary of expenditures. As much as I hate to say it, commercial contributions contributed a great deal to open source, and now that is mostly gone.
Also, some companies that gave employees a lot of free paid time have gone under, giving a lot of people a lot less time to work on their hobby projects, since they had to find a job at a more demanding, efficient place (my personal experience).
Direct commercial support is withdrawing, and inadvertant support by companies that were slack is dwindling. Fortunately, there is still momentum and Linux is thankfully more well-known now, so things won't stop, but they won't go nearly so quickly as they have the past couple of years.
All this research seems to want to throw away the desktop paradigm because it is just old, and computers can do so much more now. It seems less attention is given to how people use computers than what the computers can do. The current desktop paradigm is good, with virtual and multiple desktops, you can group information as you see fit. You don't need to use a clunky, harder to navigate 3D interface to do this same thing. The 3D paradigm has great benefit to games/simulation where you actually *want* to take the extra time to explore the environment. When you are just trying to get work done, 2D is much easier to see and wrap our heads around. Just like in real life, when we play, we do so in 3D, when working, we sit down at a desk and lay everything out in 2D. There is a reason why, for example, more people paint than work in plaster. As far as these other paradigms, they seem to be all about deciding for the user what the user wants to see and how he wants what he sees to be organized. This is a very bad mode of operation, people are intelligent enough to know what they do and do not want to see, and how to organize it. No matter what algorithm you use to guess what the user wants, it will never be 100% accurate, so just provide convenient access to everything, and be flexible enough to let the user modify it as he sees fit. Just because a concept is old, does not mean that it should be completely scrapped. This is a bad tendency that computer research people have, and they need to come to reality sometimes...
Gee, an interface based on plain text, sounds like a typewriter to me. I'm sorry but I don't think the concept of files is at all flawed on a system level. A notebook is all find and dandy for scribbles, but for more rapid organization and easier manipulation of data, the file paradigm is very much relevant. You may take notes in a notebook, but important stuff you stick in a file cabinet so you know *exactly* where you can find it, and can find similar related information. Turning computers into big paperless typewriters seems like a huge leap backward to me.
Well, actually, while ATI does not actively support Linux themselves, they do release more information about their chips than nVidia does. Because of this, third party driver developers can develop open-source, cross-platform drivers, which with nVidia, you are stuck with Linux on x86. So while nVidia invests more development time, ATI gives more information and therefore has more well-developed drivers on more platforms. Their linux driver implementation may not be as good as nVidias, but will be around even if ATI went out. If nVidia goes, and all we have are the binary drivers compiled for Kernel 2.4.x and XFree 4.1.x, then nVidia owners are in a bit of a fix....
I know that Linux hardware support is at least more comprehensive, which is why I use it on the desktop. I am preparing to configure a server, and was wondering if I should do FreeBSD. I have used it in the past, and am reasonably self-assured in many areas, but I have lingering questions about these aspects:
Software RAID-5: I see vinum, is that as good as or better than linux equiv? Are there more alternatives?
lvm: seems to be integrated with vinum, is it relatively easy to shrink and grow fses and make more fses in a vinum managed group?
nat/firewalling: I've heard very little about ipf and ipnat, how good are they at what they do? Do they do stateful firewalling? How intuitive are they to configure blocking/forwarding rules vs. iptables (note I consider iptables to be extremely intuitive)?
ipsec: I see that there is support for ipsec, does it interoperate with FreeS/WAN? (Must connect to a site and tunnel network traffic with a linux FreeS/WAN box at other end.)
I have a small linux box performing the firewalling/ipsec right now. I plan to upgrade and have volume management over a raid array, as well as apache, nfs, nis, samba (file serving and PDC), and want to maintain configurability while insuring stability. 2.4.x series of kernels have seemed to be a little too flaky in my usage for a high-availability solution, and FreeBSD seemed rock-solid when I used it.
I started in high school. There were two distinct networks running, the DECnet one and a student-run network of Sun4 systems that the professionals didn't understand enough to run on their own. Schools are typically in need of fairly technical people and are most willing to give people without any experience a chance. The pay crap, but it puts that work experence on your resume. A company took one look at my experience and hired me up. I happen to be a CS person, but academics seemed to matter less than experience, though I keep both strong. If you are willing to do Unix admin, and have the work experience and references to back you up, in my experience you can get those jobs. Make sure you get your hands on as a diverse set of Unix variants as you can. Companies love to see a long list of Unix variants in professionaly work history. And old hardware as well as new. Old hardware both gives the impression of a longer history, and lets them know you have delt with systems that aren't as mature as todays.
First just a comment saying that C and OO approaches are not mutually exclusive. You can have an OO approach in C (as gtk does). It is ugly as hell, and really doesn't make things that much easier to maintain than traditional C code, but it is possible. Not really defending this, just saying OO can be implemented in practically any language, just some can do it better than others..
As to why it is still in C++, I'll guess to make it consistent with the rst of Gnome (obvious) Why was Gnome done in C? Probably partially out of language bigotry. But some somehwat more valid reasons:
1) Give programmers maximum choice. It is easier to call C libraries from C++ apps than vice-version. If it had been based in C++, the C wrappers would be needed for any functionality, while C++ can call native C code without problems (usually)
2) A belief that C++ cannot be as fast as C. There is a little bit of overhead in C++, somewhat blown out of proportion by anti-C++ people, and therefore people think C++ is inefficient. Not really enough of a performance problem to justify this, but it is an explanation.
3) To this day g++ has been wishy-washy with how C++ code should be compiled. With gcc-3, hopefully we are coming to the end of those days. libstdc++ has changed so many times in terms of ABI, that programs compiled for one distro have little hope of making it on another. For maximum binary and source portability, C code was, especially at the beginning of gnome, the only choice.
There may be others, but these occur to me right off..
Nautilus is the biggest, most bloated piece of crap file manager I have ever seen. It is unusably slow on a 400MHz machine, even with all the "enhanced graphics" turned off. They have made some optimizations, but not nearly fast enough yet. I have to laugh that they have fam support. Not saying that is a bad thing, it would normally be extremely cool. Nautilus is so slow however, fam becomes pointless. Konqueror is better, but still sluggish. GMC is the only acceptably fast filemanager explicitly for Gnome or KDE, it has speed on par with Windows Explorer (that's not really saying much though...) The filemanager that I use that really shows what is possible is ROX (http://rox.sourceforge.net/). It makes gmc/windows explorer look dreadfully slow, and still remains easy to use for new people, and powerfully configurable for advanced users. It doesn't have support for crap like embedding web browsing or viewing components, but simply does what a file manager should do, manage files, leaving content intepretation to programs better suited for each type.
I too am slated to get my degree in May. The comp sci curriculum in school is ok and all, but just too damn boring really. The real world isn't like that. I had a job when things went down, and I've found two other places to work since then, and in the one that shutdown, and the one now (but not the crap in between) there is a lot of 'real' computer work to do. School is, for the most part, rehashing problems and techniques solved long ago with tons of documented, step-by-step solutions, which just take the time of following through those instructions, but very little in the way of actual original thought. At work, it is a matter of, we are at point A, we need to be at point B, we have no idea how to get there, and no one else knows, else we woudln't be bothering. You actually get to apply your skills rather than simply prove you can follow instructions. School doesn't present problems without a known solution, because that would be unfair, especially if it is not possible. Companies, on the other hand, are all about the unknown and potentially unfeasible problems. Much more rewarding than schoolwork.
Cool, now I have something to play movs:) So now I can play mpg/avi/asf/wmv with PyhtonTheater, DVD with Ogle (more likely my PS2), realmedia with the realone preview release (XVideo support:). What this plugin lacks that all my other players has is XVideo support... Of course it is undertsandable, but I won't be kicking my movs into fullscreen any time soon..
I like the conventional paradigm for seperate applications, but within applications I like support for both windowed and tabbed interfaces, but not managed from within the same app window unless it is tabbed. That way similar content is grouped together, and information can be layed out next to each other as needed. I would advocated either tiling windows or simply using maximized windows for tabbing, but then web pages and terminals and other application windows are all mixed up with no logical grouping whatsoever. Gnome panel and XPs taskbar at least, after a time, turns them into menus, but that isn't too much better. Tabbed web browsers (i.e. galeon) and multi-tabbed terminals are fantastic way to go:)
You do realize that Hitler had next to nothing to do with WWI and the great depression? He was just a normal German soldier in WWI, nothing special.
Re:does anyone find it weird that...
on
MS DOS: A Eulogy
·
· Score: 2
But the Windows 9x Command.com is a far cry from a *good* CLI, like many unix shells provide. I'd be screaming about the CLI in linux if I was stuck with a shell without file completion and decent command history support (not just the up button, but searchable history, DOSKEY not good enough) NT cmd is better, but not good enough...
I especially think PyGTK is an example of somewhat relatively well-done work. In my experience with Perl and Python, Python tends to lend itself better than Perl for maintenance. However, the way in which data members of a class can be declared *anywhere* could create problems on larger projects. Also, having to have self as an argument to your class functions is a bit strange, and having to reference all class methods and members explicitly through self is a bit strange too, strange way of distinguishing between local and class methods and variables. And the way private vs. public is a bit strange. Other than this, I really like Python, a really impressive interpreted language, with a really excellent way to interface with C code, more impressive to me than any other non-C language, espeically makes JNI of Java look like complete crap.
Oh, they are all for standrads, *their* standards. Notice they carefully say the standards *we* follow. They are meaning not w3c standards, but MS IE "standards" We see the price we paid for "Free" IE, Internet "standards" owned by microsoft..
I love how in the CNet article how they say the experience is degraded by mozilla and such because "simply because they don't support the standards we support closely". IE makes up its own standards, and then declares them to be more important than the official w3c standards. It's getting worse all the time with them... Well, it was bad all along, but they are growing increasingly confident and brave and, unfortunately, successful in their conquests.
I think it is a pretty good OS, far better than 9x based systems. The UI is interesting, though I'm not sure it's worth the performance hit. At least they offer the old interface. No point of upgrading from W2k unless you either really like the new UI, or need multimedia support... As for the UI, http://www.stardock.com/ has some interesting ways of refreshing that old win98 look, with Windowblinds and DesktopX.
What really puts me off is the licensing. I do relatively frequent changes to my hardware. I don't feel like frequently requesting activation code and therefore be flagged as suspicious. If it would truly be a one time thing to me, I wouldn't care so much, but asking Microsoft before I can actually use any upgrades I buy would suck.
For now, I'll stick with W2k for when I *absolutely* need Windows, and Linux for everyday purposes.
(BTW, my evaluation of Windows XP is based on a nice official beta I installed for a while before zapping it with W2k again)
As to the comment about allowing memory architecture without CPU change, it's highly unlikely. Even with the northbridge on the CPU, different memory architectures are rarely ever pin-compatible. So, in essence, both CPU and motherboard absolutely must match in terms of memory, whereas before you could theoretically have either AMD or Intel with, say RDRAM or SDRAM (though RDRAM is really stupid..)
Actually, I think it would have not been as likely to be noticed. Instead, the chipset would be blamed as having poor quality, but not necessarily intention. Seeing that the same code produces blatantly different output shows that the drivers are intentionally deceptive..
And they blatantly lie... I would have no problems with optimizations, but what it is doing is discarding user preferences for high-quality textures. There is a visual difference between the "quack3" and "quake3" runs with ATI at the *same* settings. The drivers cheat for framerate because review sites don't ever actually look closely at the game itself, only the numbers that get spit out. Makes you wonder about those 3DMark scores..
Slashdot Mirror
Actually, I would say that bandwidth savings would still be a biggie.. Not everyone has so much bandwidth that the difference in size would be trivial. Also, "the [processor] savings inn't that big a deal in these days of standard-issue "Gigahertz processors". If you are getting 40-CB drives, doubtless you also have a decently fast processor, right? On my 400-MHz, the extract time between bz2 and gz is negligible, only compression seems to be that much slower in bz2...
So, reinventing the wheel with different bugs shows intelligence? I'm sorry, but not using existing libraries that work well is not intelligent, simply arrogant, the "Not Invented Here" syndrome. That aside, they do a significant amount themselves, they have pretty good (perhaps the best) avi and asf parsers. Also, their input/output routines are quite flexible. That said, the gui is a piece of crap, and the compile can be a bit fickle, and the developers do occasionally exhibit degrees of arrogance. I like PythonTheater, but I'm partial :) If you think mplayer uses external code too much, you could *really* criticize PythonTheater, using avifile, smpeg, and SDL for all decoding and input output, as well as using ROX's configuration system. That is really the strength of open software, being able to use components that have been developed more and not starting from scratch.
At least open source in the Linux realm. During the economic boon, many businesses had so much money and resources, they could afford to effectively throw money away on open source, in the hopes that eventually it would provide opportunities to combat MS. But now the companies that are left are more wary of expenditures. As much as I hate to say it, commercial contributions contributed a great deal to open source, and now that is mostly gone.
Also, some companies that gave employees a lot of free paid time have gone under, giving a lot of people a lot less time to work on their hobby projects, since they had to find a job at a more demanding, efficient place (my personal experience).
Direct commercial support is withdrawing, and inadvertant support by companies that were slack is dwindling. Fortunately, there is still momentum and Linux is thankfully more well-known now, so things won't stop, but they won't go nearly so quickly as they have the past couple of years.
All this research seems to want to throw away the desktop paradigm because it is just old, and computers can do so much more now. It seems less attention is given to how people use computers than what the computers can do. The current desktop paradigm is good, with virtual and multiple desktops, you can group information as you see fit. You don't need to use a clunky, harder to navigate 3D interface to do this same thing. The 3D paradigm has great benefit to games/simulation where you actually *want* to take the extra time to explore the environment. When you are just trying to get work done, 2D is much easier to see and wrap our heads around. Just like in real life, when we play, we do so in 3D, when working, we sit down at a desk and lay everything out in 2D. There is a reason why, for example, more people paint than work in plaster. As far as these other paradigms, they seem to be all about deciding for the user what the user wants to see and how he wants what he sees to be organized. This is a very bad mode of operation, people are intelligent enough to know what they do and do not want to see, and how to organize it. No matter what algorithm you use to guess what the user wants, it will never be 100% accurate, so just provide convenient access to everything, and be flexible enough to let the user modify it as he sees fit. Just because a concept is old, does not mean that it should be completely scrapped. This is a bad tendency that computer research people have, and they need to come to reality sometimes...
Gee, an interface based on plain text, sounds like a typewriter to me. I'm sorry but I don't think the concept of files is at all flawed on a system level. A notebook is all find and dandy for scribbles, but for more rapid organization and easier manipulation of data, the file paradigm is very much relevant. You may take notes in a notebook, but important stuff you stick in a file cabinet so you know *exactly* where you can find it, and can find similar related information. Turning computers into big paperless typewriters seems like a huge leap backward to me.
Well, actually, while ATI does not actively support Linux themselves, they do release more information about their chips than nVidia does. Because of this, third party driver developers can develop open-source, cross-platform drivers, which with nVidia, you are stuck with Linux on x86. So while nVidia invests more development time, ATI gives more information and therefore has more well-developed drivers on more platforms. Their linux driver implementation may not be as good as nVidias, but will be around even if ATI went out. If nVidia goes, and all we have are the binary drivers compiled for Kernel 2.4.x and XFree 4.1.x, then nVidia owners are in a bit of a fix....
I know that Linux hardware support is at least more comprehensive, which is why I use it on the desktop. I am preparing to configure a server, and was wondering if I should do FreeBSD. I have used it in the past, and am reasonably self-assured in many areas, but I have lingering questions about these aspects:
Software RAID-5: I see vinum, is that as good as or better than linux equiv? Are there more alternatives?
lvm: seems to be integrated with vinum, is it relatively easy to shrink and grow fses and make more fses in a vinum managed group?
nat/firewalling: I've heard very little about ipf and ipnat, how good are they at what they do? Do they do stateful firewalling? How intuitive are they to configure blocking/forwarding rules vs. iptables (note I consider iptables to be extremely intuitive)?
ipsec: I see that there is support for ipsec, does it interoperate with FreeS/WAN? (Must connect to a site and tunnel network traffic with a linux FreeS/WAN box at other end.)
I have a small linux box performing the firewalling/ipsec right now. I plan to upgrade and have volume management over a raid array, as well as apache, nfs, nis, samba (file serving and PDC), and want to maintain configurability while insuring stability. 2.4.x series of kernels have seemed to be a little too flaky in my usage for a high-availability solution, and FreeBSD seemed rock-solid when I used it.
I started in high school. There were two distinct networks running, the DECnet one and a student-run network of Sun4 systems that the professionals didn't understand enough to run on their own. Schools are typically in need of fairly technical people and are most willing to give people without any experience a chance. The pay crap, but it puts that work experence on your resume. A company took one look at my experience and hired me up. I happen to be a CS person, but academics seemed to matter less than experience, though I keep both strong. If you are willing to do Unix admin, and have the work experience and references to back you up, in my experience you can get those jobs. Make sure you get your hands on as a diverse set of Unix variants as you can. Companies love to see a long list of Unix variants in professionaly work history. And old hardware as well as new. Old hardware both gives the impression of a longer history, and lets them know you have delt with systems that aren't as mature as todays.
First just a comment saying that C and OO approaches are not mutually exclusive. You can have an OO approach in C (as gtk does). It is ugly as hell, and really doesn't make things that much easier to maintain than traditional C code, but it is possible. Not really defending this, just saying OO can be implemented in practically any language, just some can do it better than others..
As to why it is still in C++, I'll guess to make it consistent with the rst of Gnome (obvious) Why was Gnome done in C? Probably partially out of language bigotry. But some somehwat more valid reasons:
1) Give programmers maximum choice. It is easier to call C libraries from C++ apps than vice-version. If it had been based in C++, the C wrappers would be needed for any functionality, while C++ can call native C code without problems (usually)
2) A belief that C++ cannot be as fast as C. There is a little bit of overhead in C++, somewhat blown out of proportion by anti-C++ people, and therefore people think C++ is inefficient. Not really enough of a performance problem to justify this, but it is an explanation.
3) To this day g++ has been wishy-washy with how C++ code should be compiled. With gcc-3, hopefully we are coming to the end of those days. libstdc++ has changed so many times in terms of ABI, that programs compiled for one distro have little hope of making it on another. For maximum binary and source portability, C code was, especially at the beginning of gnome, the only choice.
There may be others, but these occur to me right off..
Nautilus is the biggest, most bloated piece of crap file manager I have ever seen. It is unusably slow on a 400MHz machine, even with all the "enhanced graphics" turned off. They have made some optimizations, but not nearly fast enough yet. I have to laugh that they have fam support. Not saying that is a bad thing, it would normally be extremely cool. Nautilus is so slow however, fam becomes pointless. Konqueror is better, but still sluggish. GMC is the only acceptably fast filemanager explicitly for Gnome or KDE, it has speed on par with Windows Explorer (that's not really saying much though...) The filemanager that I use that really shows what is possible is ROX (http://rox.sourceforge.net/). It makes gmc/windows explorer look dreadfully slow, and still remains easy to use for new people, and powerfully configurable for advanced users. It doesn't have support for crap like embedding web browsing or viewing components, but simply does what a file manager should do, manage files, leaving content intepretation to programs better suited for each type.
I too am slated to get my degree in May. The comp sci curriculum in school is ok and all, but just too damn boring really. The real world isn't like that. I had a job when things went down, and I've found two other places to work since then, and in the one that shutdown, and the one now (but not the crap in between) there is a lot of 'real' computer work to do. School is, for the most part, rehashing problems and techniques solved long ago with tons of documented, step-by-step solutions, which just take the time of following through those instructions, but very little in the way of actual original thought. At work, it is a matter of, we are at point A, we need to be at point B, we have no idea how to get there, and no one else knows, else we woudln't be bothering. You actually get to apply your skills rather than simply prove you can follow instructions. School doesn't present problems without a known solution, because that would be unfair, especially if it is not possible. Companies, on the other hand, are all about the unknown and potentially unfeasible problems. Much more rewarding than schoolwork.
PythonTheater plays wmv, http://xtheater.sourceforge.net/, as well as mpg, asf, and avi.
Cool, now I have something to play movs :) So now I can play mpg/avi/asf/wmv with PyhtonTheater, DVD with Ogle (more likely my PS2), realmedia with the realone preview release (XVideo support :). What this plugin lacks that all my other players has is XVideo support... Of course it is undertsandable, but I won't be kicking my movs into fullscreen any time soon..
I like the conventional paradigm for seperate applications, but within applications I like support for both windowed and tabbed interfaces, but not managed from within the same app window unless it is tabbed. That way similar content is grouped together, and information can be layed out next to each other as needed. I would advocated either tiling windows or simply using maximized windows for tabbing, but then web pages and terminals and other application windows are all mixed up with no logical grouping whatsoever. Gnome panel and XPs taskbar at least, after a time, turns them into menus, but that isn't too much better. Tabbed web browsers (i.e. galeon) and multi-tabbed terminals are fantastic way to go :)
You do realize that Hitler had next to nothing to do with WWI and the great depression? He was just a normal German soldier in WWI, nothing special.
But the Windows 9x Command.com is a far cry from a *good* CLI, like many unix shells provide. I'd be screaming about the CLI in linux if I was stuck with a shell without file completion and decent command history support (not just the up button, but searchable history, DOSKEY not good enough) NT cmd is better, but not good enough...
I especially think PyGTK is an example of somewhat relatively well-done work. In my experience with Perl and Python, Python tends to lend itself better than Perl for maintenance. However, the way in which data members of a class can be declared *anywhere* could create problems on larger projects. Also, having to have self as an argument to your class functions is a bit strange, and having to reference all class methods and members explicitly through self is a bit strange too, strange way of distinguishing between local and class methods and variables. And the way private vs. public is a bit strange. Other than this, I really like Python, a really impressive interpreted language, with a really excellent way to interface with C code, more impressive to me than any other non-C language, espeically makes JNI of Java look like complete crap.
They'll stop sending activation codes and tell you to buy their latest software, of course... We *all* love MS..
Oh, they are all for standrads, *their* standards. Notice they carefully say the standards *we* follow. They are meaning not w3c standards, but MS IE "standards" We see the price we paid for "Free" IE, Internet "standards" owned by microsoft..
I love how in the CNet article how they say the experience is degraded by mozilla and such because "simply because they don't support the standards we support closely". IE makes up its own standards, and then declares them to be more important than the official w3c standards. It's getting worse all the time with them... Well, it was bad all along, but they are growing increasingly confident and brave and, unfortunately, successful in their conquests.
Works fine with Netscape 6.1 here, without any tricks here..
I think it is a pretty good OS, far better than 9x based systems. The UI is interesting, though I'm not sure it's worth the performance hit. At least they offer the old interface. No point of upgrading from W2k unless you either really like the new UI, or need multimedia support... As for the UI, http://www.stardock.com/ has some interesting ways of refreshing that old win98 look, with Windowblinds and DesktopX.
What really puts me off is the licensing. I do relatively frequent changes to my hardware. I don't feel like frequently requesting activation code and therefore be flagged as suspicious. If it would truly be a one time thing to me, I wouldn't care so much, but asking Microsoft before I can actually use any upgrades I buy would suck.
For now, I'll stick with W2k for when I *absolutely* need Windows, and Linux for everyday purposes.
(BTW, my evaluation of Windows XP is based on a nice official beta I installed for a while before zapping it with W2k again)
As to the comment about allowing memory architecture without CPU change, it's highly unlikely. Even with the northbridge on the CPU, different memory architectures are rarely ever pin-compatible. So, in essence, both CPU and motherboard absolutely must match in terms of memory, whereas before you could theoretically have either AMD or Intel with, say RDRAM or SDRAM (though RDRAM is really stupid..)
Actually, I think it would have not been as likely to be noticed. Instead, the chipset would be blamed as having poor quality, but not necessarily intention. Seeing that the same code produces blatantly different output shows that the drivers are intentionally deceptive..
And they blatantly lie... I would have no problems with optimizations, but what it is doing is discarding user preferences for high-quality textures. There is a visual difference between the "quack3" and "quake3" runs with ATI at the *same* settings. The drivers cheat for framerate because review sites don't ever actually look closely at the game itself, only the numbers that get spit out. Makes you wonder about those 3DMark scores..