Actually, you're the dipshit. Money has no intrinsic value. If the average American made $0.70 an hour, yet remained as productive as they are now, the value of a doller would be much higher, so the *real* cost of living would be the same.
And yes, we had a depression in the 1930s. Such cycles are a natural part of any capitalist economy. What's your point?
Read the article. The Munich transition troubles have little to do with Linux, and a lot to do with the fact that they are having trouble dealing with the rest of the Windows-using world. They've got software that they need to port, and subcontractors that they have to be compatible with.
And where exactly are you getting this load of bull?
Apple's SIGGRAPH presentation on Quartz extreme, page 10. Notice how the "Quartz 2D" line is yellow, which, as the legend indicates, means software rendering. Longhorn will actually use Direct3D for the drawing *inside* the windows, so you can have very rich vector graphics in applications.
And Longhorn is an example of MS copying Apple copying NeXT:)
The Mac doesn't have it. Longhorn has a fully 3D accelerated desktop. As in drawing as well. The Mac only uses 3D acceleration for the specific case of compositing windows together.
I didn't call your evidence shaky, though it is. You initially made the claim that your drivers were "awesome" but you haven't posted any benchmarks to show that the OSS process has actually resulted in fast, conformant drivers. I'll post a benchmark showing the opposite, though: Look at the Radeon 7x00 scores.
In any case, being faster than ATI's binary drivers is not big win. ATI's Linux drivers suck as bad as their earlier Radeon Windows drivers did. And every benchmark I've seen shows that the ATI DRI drivers are even *slower* than the ATI binary drivers!
I'll say it again: there are no high-quality open-source 3D drivers, where "high-quality" means being comparable in speed to the Windows driver for the card, and exposing all the OpenGL features supported by the card.
Compared to the XiG Summit drivers, the DRI drivers for the 7x00 are much much slower as well. And I cannot find anything that says the r200 drivers are any slower than the r100 drivers.
What? The DRI drivers share much of the same codebase, so should display similar relative performance for different generations of cards. And I tried to dig up the latest DRI benchmarks I could find because I remembered that they've gotten better since they came out. The earlier ones were even worse.
Just because the drivers meet your needs doesn't mean that they are high quality. A slow driver might still be useful, but it is *not* high quality. Hence, my original statement stands --- there are no high quality open 3D drivers.
And I never said that they were always 1/3 as fast. On some tests they are 1/3 as fast. On the benchmarks I linked to, in nearly every single test, the DRI drivers come in dead last. I think once they managed to come in third.
It is well known that the ATI DRI drivers are slower than even the proprietory ATI Linux ones (which suck) and much slower than the ATI Windows or XIG drivers. In high-detail OpenGL scenes, the DRI drivers can be less than half as fast. Good benchmark here.
It is well-known that the DRI drivers for the Radeon cards are much slower than the Windows or XIG drivers for them. In many cases, especially on high-polygon-count scenes like those used in CAD apps, the DRI drivers are a half to a third as fast as the XIG ones!
I hated Zelda and the original Final Fantasy. The Final Fantasy games on the SNES were great, though, but they did have very good graphics for their time.
How about a completely different toolkit? And are you seriously claiming that every single GTK2 theme in existance is ugly? Several of them are done by professional artists (no 14 year old!)
The P4 seems to handle indirect accesses extremely well. They did a benchmark of bcc awhile ago. Bcc is a version of GCC that does bounds checking. Now, bounds-checking in C sucks because you have arbitrary pointer arithmatic. So a pointer balloons from a 4-byte word that fits in a register, to a 12-byte structure that must be accessed indirectly. On a P3 and an Itanium, the penalty was huge, reaching 117% for the P3. However, the penalty on the P4 was only 34%.
You really have to ask yourself why a venture/vulture outfit like Canopy has invested in TrollTech -- it's not just for giggles. ----------- Dude. SCO (formerly Caldera) wasn't always evil. There was a little thing called "Caldera Linux" that showed quite a bit of promise. And guess what --- it was based on Qt/KDE!!! Now you can see why Caldera chose to invest in Trolltech.
And Trolltech's pricing is in the range of that of other development tools, UNIX or not. Visual Studio is a "PC" dev tool, but the pro version will cost you $1000+ and the Enterprise version will cost you $2300+. In any case, Trolltech's pricing was set long before Caldera turned evil!
1) You're not just limited to GPL with GPL'ed Qt. You can use pretty much any open source license that's GPL-compatible or QPL-compatible.
2) Sun never chose not to use Qt because of the GPL. Sun chose GNOME as a platform because:
1) It was C, rather than C++. Their engineers were more familiar with C, and at the time, KDE wouldn't compile with their Sun Forte C++ development tools.
2) Since GNOME was not as far along, they got to have a huge influence on the HIG work. They could never have had that kind of influence on KDE.
3) They liked the GNOME approach of going with existing technologies. In particular, they liked the use of industry-standard CORBA for Bonobo, rather than a custom solution.
(1) and (3) were official statements from Sun, while (2) can be inferred from the huge work Sun has done on GNOME's UI.
And Ximian chose not to use Qt because it was fricking started by Miguel de Icaza! You know, they GNOME founder?!!
Qt isn't that expensive. A single-developer copy of Qt costs less than Visual Studio Enterprise, much less than a copy of Rational Rose, and less than the cost of a Clearcase license.
Slashdot Mirror
Actually, you're the dipshit. Money has no intrinsic value. If the average American made $0.70 an hour, yet remained as productive as they are now, the value of a doller would be much higher, so the *real* cost of living would be the same.
And yes, we had a depression in the 1930s. Such cycles are a natural part of any capitalist economy. What's your point?
Read the article. The Munich transition troubles have little to do with Linux, and a lot to do with the fact that they are having trouble dealing with the rest of the Windows-using world. They've got software that they need to port, and subcontractors that they have to be compatible with.
Apple never bought NeXT. NeXT bought Apple :)
And where exactly are you getting this load of bull?
:)
Apple's SIGGRAPH presentation on Quartz extreme, page 10. Notice how the "Quartz 2D" line is yellow, which, as the legend indicates, means software rendering. Longhorn will actually use Direct3D for the drawing *inside* the windows, so you can have very rich vector graphics in applications.
And Longhorn is an example of MS copying Apple copying NeXT
Linus couldn't do that. He doesn't own the complete copyrights to the code in the kernel.
The Mac doesn't have it. Longhorn has a fully 3D accelerated desktop. As in drawing as well. The Mac only uses 3D acceleration for the specific case of compositing windows together.
I didn't call your evidence shaky, though it is. You initially made the claim that your drivers were "awesome" but you haven't posted any benchmarks to show that the OSS process has actually resulted in fast, conformant drivers. I'll post a benchmark showing the opposite, though: Look at the Radeon 7x00 scores.
In any case, being faster than ATI's binary drivers is not big win. ATI's Linux drivers suck as bad as their earlier Radeon Windows drivers did. And every benchmark I've seen shows that the ATI DRI drivers are even *slower* than the ATI binary drivers!
I'll say it again: there are no high-quality open-source 3D drivers, where "high-quality" means being comparable in speed to the Windows driver for the card, and exposing all the OpenGL features supported by the card.
My ID is lower than both of yours :)
Compared to the XiG Summit drivers, the DRI drivers for the 7x00 are much much slower as well. And I cannot find anything that says the r200 drivers are any slower than the r100 drivers.
What? The DRI drivers share much of the same codebase, so should display similar relative performance for different generations of cards. And I tried to dig up the latest DRI benchmarks I could find because I remembered that they've gotten better since they came out. The earlier ones were even worse.
Just because the drivers meet your needs doesn't mean that they are high quality. A slow driver might still be useful, but it is *not* high quality. Hence, my original statement stands --- there are no high quality open 3D drivers.
And I never said that they were always 1/3 as fast. On some tests they are 1/3 as fast. On the benchmarks I linked to, in nearly every single test, the DRI drivers come in dead last. I think once they managed to come in third.
It is well known that the ATI DRI drivers are slower than even the proprietory ATI Linux ones (which suck) and much slower than the ATI Windows or XIG drivers. In high-detail OpenGL scenes, the DRI drivers can be less than half as fast. Good benchmark here.
It is well-known that the DRI drivers for the Radeon cards are much slower than the Windows or XIG drivers for them. In many cases, especially on high-polygon-count scenes like those used in CAD apps, the DRI drivers are a half to a third as fast as the XIG ones!
I hated Zelda and the original Final Fantasy. The Final Fantasy games on the SNES were great, though, but they did have very good graphics for their time.
Actually, I find those games tremendously boring. Because of that, I mostly play RPGs, which are a lot more immersive when the have rich graphics.
Its actually more powerful than a PSX. It uses (for obvious reasons!) a kind-of minidisc media, rather than a CD.
Please fucking use it before commenting. The UI is *very* different. Its actually usable without special training now...
How about a completely different toolkit? And are you seriously claiming that every single GTK2 theme in existance is ugly? Several of them are done by professional artists (no 14 year old!)
Dipshit.
If it is genuine, then great! Go Kerry for having the guts to protest a stupid war.
However, Kerry, you're still an ass for support the most recent stupid war...
There are no quality free drivers. The ATI DRI drivers have terrible performance.
The P4 seems to handle indirect accesses extremely well. They did a benchmark of bcc awhile ago. Bcc is a version of GCC that does bounds checking. Now, bounds-checking in C sucks because you have arbitrary pointer arithmatic. So a pointer balloons from a 4-byte word that fits in a register, to a 12-byte structure that must be accessed indirectly. On a P3 and an Itanium, the penalty was huge, reaching 117% for the P3. However, the penalty on the P4 was only 34%.
PA-RISC and Itanium are completely different.
You really have to ask yourself why a venture/vulture outfit like Canopy has invested in TrollTech -- it's not just for giggles.
-----------
Dude. SCO (formerly Caldera) wasn't always evil. There was a little thing called "Caldera Linux" that showed quite a bit of promise. And guess what --- it was based on Qt/KDE!!! Now you can see why Caldera chose to invest in Trolltech.
And Trolltech's pricing is in the range of that of other development tools, UNIX or not. Visual Studio is a "PC" dev tool, but the pro version will cost you $1000+ and the Enterprise version will cost you $2300+. In any case, Trolltech's pricing was set long before Caldera turned evil!
Couple of points:
1) You're not just limited to GPL with GPL'ed Qt. You can use pretty much any open source license that's GPL-compatible or QPL-compatible.
2) Sun never chose not to use Qt because of the GPL. Sun chose GNOME as a platform because:
1) It was C, rather than C++. Their engineers were more familiar with C, and at the time, KDE wouldn't compile with their Sun Forte C++ development tools.
2) Since GNOME was not as far along, they got to have a huge influence on the HIG work. They could never have had that kind of influence on KDE.
3) They liked the GNOME approach of going with existing technologies. In particular, they liked the use of industry-standard CORBA for Bonobo, rather than a custom solution.
(1) and (3) were official statements from Sun, while (2) can be inferred from the huge work Sun has done on GNOME's UI.
And Ximian chose not to use Qt because it was fricking started by Miguel de Icaza! You know, they GNOME founder?!!
Qt isn't that expensive. A single-developer copy of Qt costs less than Visual Studio Enterprise, much less than a copy of Rational Rose, and less than the cost of a Clearcase license.