Its probably that they can't make the V5 sell well enough. The cips are only 166Mhz, it should be a cinch for any decent manufacturer to pull them off.
Really? I bet I can teach a totaly newbie to install updated EAX drivers faster than you can teach them to install ALSA! I bet I teach that same person to install the Windows Detonator3 drivers much faster than you can teach them to dl the RPMS, bull with the.so files, and edit XF86Config to get them working. I can install my scanner faster in Windows (just put in the disc when it asks) Of course they're both about as easy to USE, but that's because KDE is so similar to Windows. The problem comes when you don't have an IT staff behind you to help you do things like add hardware or reconfigure something. It is pretty easy to walk someone through an install of DSL and NAT on Windows. Would you really want to do the same on Linux? As far as I can tell, going from dialup to DSL requires a little bit of CLI, and NAT absolutely requires CLI.
NT may take longer to boot (not appriably though) but it runs faster. DirectX blits faster than SDL, windows scroll smoother in NT than in KDE2 or GNOME, menus pop up faster, apps take less time to load (than in KDE2 and GNOME at least, regular X apps are pretty fast) and 3D runs faster.
I use it because it's just good enough is the excuse people use to run Windows 95. Why bother with an OS that doesn't take full advantage of your hardware? Why pay $200 for a graphics card and be limited to the $170 that your OS uses? SDL isn't comparable to DirectX (I've used both, trust me)(SVGAlib isn't even in the same league given that it doesn't offer hardware acceleration) OpenGL is falling behind Direct3D, OpenAL isn't even ready yet, and nobody uses it yet. Overall, multimedia on Linux is pretty pathetic. I mean you can say, "ok, it's good enough for me," but I judge the quality of something benchmarked against the best product available in that catagory. Right now, NT is the best product available in the multimedia OS catagory. Sure you may be able to run MAME in Linux, but the point is that you'll be able to run it FASTER in NT.
As far as I can tell, most RPMs are compiled for i386. I COULD compile myself, but I have better things to do. Right now in Slackware, I compile glibc, X, and my window manager by myself. However, compiling stuff is really a chore I'd rather not do, and I tend to use many different programs, so it's not a compile once and forget it issue.
First, one question: do you think BeOS has great multimedia support? Heck, it doesn't even play
DVD's!;)
>>>>
I never said it did. I don't have a DVD drive, so what do I care? When I go trumpeting BeOS as the be-all end-all multimedia OS, then take me to task for it. However, In Be's defense, I have to say the MediaKit blows away anything I've ever seen in the multi-media arena.
Windows and Linux and MacOS all have SDL.
>>>>>
SDL is just a wrapper.
It supports DirectX on Windows, and DGA on X,
which is the equivalent.
>>>>
DGA on X is equivalent to DirectX? Where's the 3D Audio API in DGA? Or maybe you mean DGA is comparable to DirectDraw? (Which still isn't totally true, show me the DGA API that allows me to change the screen-res.) DirectDraw is just a small part of DirectX.
The tests I saw for 3D acceleration didn't differ by much; it's way fast for me, even for Q3A. My speakers can't even take advantage of 3D sound decently, and I think I saw patches for all that stuff, but I really don't care yet. However, Linux *does* have great 3D
acceleration, and the audio stuff is in hardware! Maybe our definitions of great differ, but the actual
performance I'm seeing is pretty impressive. And I'm not saddened that I don't have 'force-feedback' on my Gravis Gamepad, I don't know of any games that support it, and I thought it was a dumb idea on the Playstation!
>>>>
Your arguement boils down to "I see no need for it so It's not important."
Fact> Linux 3D is slower than WindowsNT 3D. The tests showed it close (but significantly slower at high res) but they were skewed due to driver differences (I'm talking about the NVIDIA drivers.)
Fact> My speakers, and many other people's DO support 3D sound well. For these people, Linux multi-media is sub-par. Fire up Half-Life with an A3D card. It totally kicks ass and the 3D sound adds a lot to the game.
Fact> Half the hardware on my sound card goes wasted when I'm in Linux.
Fact> If you're impressed with Linux audio, take a look at DirectSound. You'd faint.
Fact> Force feedback is a dumb idea. The mere fact that you're comparing it to PS (which is just a vibrator) shows you've never used it. For certain games (racing games) it's awesome. And a LOT of game support force-feedback.
Yeah, Windows has more hardware support; that's because Microsoft doesn't have to write *ALL*
the drivers. This situation is changing of course, and I like being able to poke around with the
source code, but it'll take a while for this one to change--the corporate culture can be pretty
entrenched about these things.
>>>>>
Corporate culture aside, MS still does have more hardware support. Excuses are excuses, and drivers are actual stuff that you can base an arguement on.
There are many different limited versions of Win2K. I don't care if it supports 2 processors out of
the box, or what the configuration-of-the-week is; the bottom line is, Microsoft will always sell you
the same product for much more by just making a minor registry tweak so you get the "new
features", and I'm fundamentally opposed to that, because it's stupid; just as stupid as CPUs and
overclocking, nowadays.
>>>>>>>>>>.
Overlocking is not stupid if the consumer is the one doing it. Getting 100MHz for free by doing a little jumper manipulation? Adjusting voltages to get that perfect stability plateu? Hacking at it's finest!
Well, I'd like to see the results of the tests before I draw my conclusions; you may be right. But if I
did the tests, and one platform consistently crashed under certain conditions, I'd note that and put
it in my review; that's NOT a feature.
>>>>>>>>>
Windows is unstable compared to Linux. True. However, WindowsNT easily has a week or two of uptime, and for most gamers or multimedia people that's enough. They reboot their machines every night (Unless they're doing a rendering or something, and WindowsNT can usually handle that.) They don't install wierd software. (Hell RealPlayer flaky on BeOS and Linux too, it's not a Windows-specific problem.) For the average gamer, crashing is not an issue. A game being playable at one res setting higher on one OS (generally, you can play Quake on Windows at one res higher than you can in Linux) IS an issue.
Another benefit of Linux is the multitude of configuration options. Given the time and resources,
I'd love to just benchmark Linux against itself! That is to say, configure one box with a standard
kernel, OSS, ext2, XFree86 3.3.6, and a couple of IDE drives, and then configure another box with
an optimized kernel, ALSA, reiserfs, XFree86 4.0.1, software RAID... well, in any case benchmark
all the components against each other, and find out what the fastest, most stable Linux
configuration is for a given hardware configuration. That isn't as straightforward in NT, because
there aren't that many configuration options, and many of them aren't obvious or readily available.
>>>>>
Huh?
A) DGA is nothing compared to DirectDraw.
B) Given good driver writers (and Matrox ain't one of them) a Windows driver is rock solid. My system has a NVIDIA card and NT 4. I started out with the Detonator drivers months ago and it has yet to crash. In fact, my life in Windows-land has been pretty happy. If your Windows system is all buggy and crashy, get some NVIDIA and Creative-Labs hardware, an Intel, Asus, or Abit mobo, and a K7 or Intel chip and everything should be peachy. The idea is to be quality hardware from quality manufacturers. If you do, you enjoy not only good stability, but the increased performance of drivers that are "closer to the metal."
---
Exactly how much extra performance do those optimizations offer? From a multimedia software performance POV? I've heard that non-Pentium optimized code runs only about 70% as fast as Pentium optimized code on the Pentium. Also, I know that the i686 architecture has several quirks and speed-ups, and I don't know if most Linux distros really take full advantage of my system. Does compiling for i386 exact a peformance hit on Pentiums, and if so, why the hell do it?
Hipocracy means saying one thing and doing the other. Releasing your software under the GPL for Linux (saying that you embrace free software) while keeping it for-pay under Windows (which is essentially NOT embracing free software) is hipocracy by definition. Unless you can prove that using the GPL doesn't mean that you embrace free software, or that charging the developers on one OS and not the other is in tune with the ideals of OSS, then you've got no case.
BTW> People who correct spelling are the most anal-retentive jack-holes you'll ever meet.
Maybe because the only decent OS for 3D is Windows? Maybe because he does Windows development. Maybe he is a DTP guy and needs Photoshop and Quark XPress. (though you'd probably use a Mac in that case.) Maybe he likes having a stable web-browser. Maybe he is an AOL user? Being smart does mean hating Windows. If you're smart, you use what suits you best, not what you need to feel elite. Linux is good, but there are many tasks for which Windows is just better suited, and to think otherwise is just deluding yourself.
TW, Linux has great multimedia support. Okay, okay, let me explain. I have a Matrox G400, and the hardware
acceleration is *sweet*. Also, my SB Live Value is just excellent, I love the hardware mixing, the multiple DSP's... I
installed ALSA and now the MIDI patches ("soundfonts") work too, and they sound good.
>>>>>>
Wrong. Linux doesn't have DirectX, as such, it's multimedia support by definition cannot compare with that of Windows. Not a flame, think about it. Most consumer hardware is designed for DirectX. Aside from idiot companies (ATI!) most manufacturers expose all the hardware's features through DirectX. While alternative APIs can take advantage of these features, rarely-used, or difficult to implement things always get left behind. Right now, Windows still has the fastest 3D hardware acceleration, and on NVIDIA cards (the only ones that offer competitive Linux drivers) Windows is still a good bit faster than Windows. In the tests where Linux was close to Windows, Linux had an unfair advantage. Apparently, the cool new speed-ups in the Detonator3 drivers were already in the Linux/X4 drivers. Thus, Linux 3D is close, but on cigar. As for sound, Linux is a distant second to Windows. ALSA still doesn't offer as many features as DirectSound, and all transistors on your EMU10K1 chip used for 3D sound are totally wasted in Linux. So, Linux lacks great 3D acceleration, great audio acceleration, and lacks 3D sound and force-feedback all together. How can it possibly have "great multimedia support?"
It's all about picking supported hardware, though, which you still have to do for Win2K, as well. My DVD drive isn't supported, because I (a) couldn't find much information about that on the net, and (b) just bought it first, figuring I'd test it out later. I'll probably ask the developers about this, since it works under Windows. I've heard it can be made to work under VMWare too, so all I really need is some debugging info.:)
>>>>>>>>
You have to choose the correct HW on Windows as well, but you've got more to choose from. On Windows, if I want great 3D acceleration, I can pick any number of cards from ATI to Matrox to NVIDIA to 3DFX (ugh!) On Linux, I'm pretty much limited to Matrox and NVIDIA (not 3DFx with X4.0)
Plus support for multi-processing makes it even better. Now, I don't have more than one processor, but I might set up
a dual-proc test box if I can ever find an old board for it. However, I've seen it done on Linux, and it is sweet. No paying extra for a different version that just consists of a stupid registry hack, either. It has decent multi-processing support out of the box. And I'd love to see a comparison to Win2K here, since that's one thing that's supposed to be
better in 2.4.
>>>>>>>>>
Huh? Win2K Pro has SMP out of box. There IS no single proc limited version of Win2K.
How about that, eh, guys? Something based around Linux Kernel 2.4.0 with a bunch of stable stuff, vs. Win2KSP1, or whatever is current and patched by then. Test multi-processing, test well-supported hardware, RAID, whatever. Just test the hell out of it.
>>>>>>
Windows would win a lot of the tests. The benefits of Linux aren't so much in raw benchmarks but in overall quality. Sure Windows may win total TCP/IP throughput scores, but it will probably crash under high load. Sure the FS may be able to transfer more date through the system (though I don't think NTFS is faster than ext2) but will it do that consistantly, or in spurts? Also, anything that taps DirectX will totally blow Linux away, since a DirectX application can for the most part be considered seperate from the underlying OS. (Especially since Win2K allows DirectX direct hardware access.)
Re:More Debian-valid Software -- is this really go
on
Qt Going GPL
·
· Score: 2
Both Mandrake 7.1 and Slackware 7.1 come with both egcs 1.1.2 and gcc-2.95.2. They also come with all the additional components for each compiler (egcs-java and gcc-java, etc).
You're missing several points in your oversimplification of OSs to include only Windows and UNIX.
There was still some DOS functionality as of NT 4.0, but I beleive it was mostly removed for Win2000.
>>>>>>
There is no DOS functionality in NT. All DOS programs are run in a virtual machine. That virtual machine is more or less unchanged in Win2K.
NT is also a microkernel, which means it naturaly has some extra overhead in it that Linux's monolithic
kernel does not. The still-mostly-vapor GNU HURD is also a microkernel. If done properly, the extra
overhead isn't that much. The question is, did Microsoft do it properly? I don't really know.
>>>>>
Yes MS did it properly. The things holding back NT don't tend to be core system related at all. It's all the stuff MS added on top that sucks. NT4 for example is a good bit faster than Linux for most desktop-type operations. However, when you look at Windows2000 with all the crap they added (active desktop and all) you notice it's much slower. NT doesn't suffer so much from core-system bloat and bugs, but stuff-added-on-top bloat and bugs.
In any case, its much easier for hackers to create a stable and speedy core system, while difficult for them to make a good UI.
>>>>>>>
Just plain wrong. BeOS: Fast stable, good UI. QNX: Fast, stable, decent UI. There are a whole bunch of systems out there that are fast and stable, and have good UIs to boot. Even Quartz seems to be pretty fast (in so far as a DPDF system can be.)
On the other hand, its far easier for a corperate project to make a reasonably good UI, but
diffcult to make a stable and speedy core system.
>>>>>
Again, not true. QNX is probably more stable than Linux and a more stable to boot (at least according to those who've used it.) Again, BeOS is managed by a corporation, and stability and speed aren't exactly high on Be users' lists of complaints.
Re:QPL is still an option
on
Qt Going GPL
·
· Score: 2
Isn't that hipocracy? I mean propriatory software developers on Windows are no worse then their Linux counterparts. If Linux developers get it free, then so should Windows developers!
Re:Gtk/QT use for commercial products
on
Qt Going GPL
·
· Score: 2
Uh, I don't really think $2500 is going to be a deciding factor in software development. I mean most projects spend 10x that on pizza over the life of the project! If Qt saves just 5% development time for a dozen programmers in a medium project, $2500 will mean nothing.
Re:More Debian-valid Software -- is this really go
on
Qt Going GPL
·
· Score: 2
If you hadn't realized this, but Linux distros are full of redundant software. That's one of the main reasons I switched to Slackware. In RedHat 6.2, there are 3 versions of awk. Most distros come with two versions of gcc and their addandant language components (OBJC, Java, etc.) Mandrake comes with 3 Java compilers. (GCC-Java, Guavac, and Jikes.) Even Slackware has 3 versions of libc. Then there is all the redundancy that KDE and GNOME bring. Face it, if you were scared of a lot of code living on your computer, you wouldn't use Linux.
WTF is with the screwy "Funny" mods? Don't get me wrong, I don't mind being modded up, but why "Funny." If you think I said something incorrect, please answer instead of modding me UP. I got one the other day too for saying "Java can be compiled." What's up?
Then let the best company win. My point is there is that I have never seen so much crap about the "ethics and morals and conduct" of a company than I do on Slashdot. I use whatever is the fastest, and for a lot of things (particularly anything SSE like Photoshop) Intel is the fastest.
Just because AMD also FUDs, doesn't make it okay for Intel to FUD. AMD *has* to
FUD to survive against competitor Intel's FUD. That doesn't make it alright either. I
know it interferes with the magical "free market", and the invisible hand-job and all,
but it would be nice if *someone* would take these companies to task for false
advertising when they FUD.
>>>>>>>>
There is FUD and market speak. FUD is outright lies. Market speak is streching of the truth. Intel saying the PIII makes the Internet faster is market speak. It DOES make your browser run impreceptibly faster and it helps a little on media rich and 3D rich websites (of which there are about 2.) The only people for whom market speak should matter are those too stupid to see through it. Believe it or not, the US economy thrives on taking advantage of those too stupid to do their own research. It happens with american cars, iMacs, graphics cards (wonder why ATI is on so many computers?) insurance, houses, sneakers, socks, abso-fucking-lutely everything.
Intel pushes IDE because IDE relies on CPU horsepower to initiate transfers. SCSI
relies on the controller. Therefore, if HD access takes CPU cycles, this increases the
demand (in a bogus way) for faster CPUs. Bottom line for Intel; slower and ultimately
costlier machines for consumers. Same principle works with Winmodems, and USB
too, if I'm not mistaken.
>>>>>>>>>>>
That's absoultely untrue. Under UltraDMA, one or two drives have no more CPU usage than on SCSI. Your comment was true back in the days of PIO, but not anymore. The only place SCSI helps these days is if you need the raw I/O rate, need absolute reliability, need access the the generally faster SCSI drives, or have more than two drives in you computer. Winmodems don't only exist because of Intel. In an age of $300 computers and low double digit profit margins, the $15 or so a winmodem saves means a lot more money for the manufacturer. Also, Winmodems were created when the cheapest "real" modems you could get were nearly $100. The whole point, though is moot. Intel doesn't really do anything to push Winmodems. As far as I can tell, 3Com (USRobotics) was the one the entrenched Winmodems.
I do agree that PCI was a good spec. So why did they abandon it for AGP?
>>>>>>
They didn't abandon it for AGP. As far as I can see, AGP is only a graphics standard.
If you read
Intel's press releases (and it sounds like you do, on Sunday morning, in front of an
altar, accompanied by organ music), it's to reduce the need for VRAM on the card,
but the REAL reason was so they could corner the market on graphics chips.
>>>>>>>>>>>
The VRAM issue was quite big. Hindsight is 20/20, but as I remember it, when AGP came out, the best 3D cards had 4MB of RAM (Riva128 and ATI Rage Pro). It was before the big memory price bust, and it seemed that cards could continue to use expensive SGRAM. Who the hell would have thought that in two years graphics cards would come standard with 32MB of 150MHz DDR-SDRAM? Back then (actually 6 months after AGP was introduced and years after they thought to make it) the state of the art was the V2 and it's massive 8MB (out of 12) of 100MHz texture memory. If the pace of 3D was the same from V2->Present as it was from V1->V2, then my RivaTNT 16MB would be the state of the art. However, NVIDIA totally blew apart the 3D market with it's 6month cycle. Nobody expected 3D to come this far this fast. No me, not you, not Intel.
Why
did they fail? Because their chips SUCK. They had decent competition.
>>>>>>
Actually, the i740 was pretty good. Short-lived, but pretty good. However, you're probably mistaken on Intel's intentions to corner the 3D market. If that's what they were after, they wouldn't have quite so quickly. (Do you really think Intel couldn't design a better graphics chip? Especially with Real3D's help?) My guess is that Intel was trying to make inroads into the sub-$1000 PC market with a cheap, system-memory based graphics card. At that time, AMD and Cyrix were making big strides there and it was really exploding.
And
nowadays, I challenge you to even FIND an AGP card with 4mb of VRAM. Or less than 16.
>>>>
Take a look at the advertisements for Microcenter (or CompUSA, or whatever. (A major retailer in Virginia.) All the cheap sub-$700 PCs have i740-based card with "4MB display cache."
And finally, about the AMD statement about using Slot A for higher performance,
that may be the public statement, but in Private, you can bet your loose and hairy
sphincter that the reason was so they didn't have to pay their main competitor
MORE license fees, and take it up the ass when Intel decided to change the spec
again (Slot 2, Socket 370, etc. ad infinutum, ad nauseum, ad analintrusiun), but they
took it up the ass anyway when Intel told the mobo manufacturers; "nice license you
got there, it'd be a *shame* if something were to happen to it!"
>>>>>>>>
According to AMD, they had a "blanket-cross license with Intel for all it's patents." (AMD's Presidents quote) Methinks they used SlotA more to cuddle up with Digital.
You don't really get it. The magic in CORBA and KParts isn't that they allow apps to communicate, hell, you've add that ever since signals. What's cool is it allows you to build applications in terms of other applications. It allows programmers to create apps that are mainly sets of components lashed together with glue code. UNIX modularity at it's finest.
A major gribe I have about RPM (and one of the reasons I moved to Slackware) is that it assumes that you'll do everything through RPM. It doesn't mesh well with non-RPM data. What I'd really like to see added is the ability to merge additional non-RPM-installed files into the RPM database. For example, I prefer to compile X myself, so forever afterwards, RPM keeps complaining about missing X libraries. And no, --nodeps isn't a viable option.
BTW> Maybe it already allows you to do this. However, there is nothing in the RPM docs, nor on the net that shows how.
Well, if Intel had pushed for onboard SCSI instead of going with IDE, maybe we could be using Ultra160 drives at the price of current IDE drives.
>>>
Intel wasn't the first company in the core logic chipset game. Nearly all of the early PC chipsets used IDE. It wasn't until much later that Intel started making chipsets, and at that point, IDE was already entrenched. You can't fault Intel for just doing like everybody else when they made their first chipsets.
I've seen such benchmarks; the difference is minimal until you get up to AGPx4. In fact, for most people there is no detectable difference.
>>>
Those benchmarks are for cards like the Voodoo series (up to V5) that don't use AGP effectivly. I'm talking about REAL AGP cards.
for those people with 4MBers, AGP is a godsend.
>>>>>>>
If you've only got 4MB of local graphics RAM, you're sure as hell going to see the difference between AGP and PCI. And that's not that uncommon. A lot of cheap comps come with only 4MB of memory but AGP 4x.
Not true. Apple was able to do it with G3's without moving to a slot architecture.
Untrue. It was precisely because Intel had the Slot1 specifications locked up with patents that AMD had to select a different bus architecture for the Athlon.
>>>>>>>>
Read MaximumPC (or maybe boot's) interview with AMD's President. He specifically says that they have a cross-license agreement with Intel that would allow them to use Slot1, but used SlotA for the higher performance.
That's because Intel keeps FUCKING with us. They lie, they release crappy products, they make absurd FUD-filled product roadmaps they have
no intention of fulfilling, they release products that do not ship in sufficient quantities,
>>>>>>>
Exactly what "crappy" products has Intel released? As far as I can remember, the fastest chip on the block is still the PIII 1.13 GHz. At least according to SysMark2000 (a script of 8 common applications) and Quake. Check out the reviews of the 1.1GHz Athlon at www.sharkyextreme.com. What FUD filled road-maps? Sure Itanium has been delayed several times, but hell, all chip-makers have these problems. For example, AMD and it's long delayed K5.
they perpetuate old out-dated worthless technology
(x86, ISA, IDE, etc.),
>>>>>>
x86 chips are the price/performance leaders for most things. They tried to kill ISA two or three years ago, but consumers wouldn't let it die. Intel in no way pushes IDE that hard, I think the PC market in general does that. Show me an alternative! Find me a SCSI harddrive that can match my Maxtor DiamondMax 20 at the same price (20GB for $150). Didn't think so.
and the new technology they do promote is almost always totally bass-ackwards, and designed primarily to either grab
more marketshare, or defend marketshare they have to the death, by any means other than what you'd logically think would be the BEST way to
get marketshare: make faster, more stable, more affordable chips than the competition. (USB, Slot1, AGP, etc.)
>>>>>>>>
As for pushing technologies to gain market share, that's what companies do, live with it. However, in recent memory, all of Intel's desicions have been pretty good. USB was a decent technology that freed up those IRQs. PCI was a whole lot better than VLB. USB 2.0 might not be as good as firewire, but do you really want to pay Apple licensing fees? As for AGP, I'd really like to see a performance benchmark of PCI vs. AGP. Not all of us can afford 32MB cards, and for those people with 4MBers, AGP is a godsend. Slot1 was a mistake. However, what was their alternative? They wanted backside cache, and a slot was the only way to do it at the time. AMD had to do it too. (BTW> Slot1 isn't as closed as you think. AMD's license with Intel would have allowed them to use it as well.)
And most infuriatingly, the main reason Intel seems to succeed is because stupid people buy into their "Intel Inside" campaign, that a PIII makes
surfing the internet faster.
>>>>>>>
You can't judge a company on marketing speak. Let's see, the recent Suse Slashdot banner touts Suse 6.4 as the best OS ever. That's why Linux sucks! However, you can judge them by the fact that right now, the PIII is the fastest Quake chip.
Slashdot Mirror
Its probably that they can't make the V5 sell well enough. The cips are only 166Mhz, it should be a cinch for any decent manufacturer to pull them off.
Really? I bet I can teach a totaly newbie to install updated EAX drivers faster than you can teach them to install ALSA! I bet I teach that same person to install the Windows Detonator3 drivers much faster than you can teach them to dl the RPMS, bull with the .so files, and edit XF86Config to get them working. I can install my scanner faster in Windows (just put in the disc when it asks) Of course they're both about as easy to USE, but that's because KDE is so similar to Windows. The problem comes when you don't have an IT staff behind you to help you do things like add hardware or reconfigure something. It is pretty easy to walk someone through an install of DSL and NAT on Windows. Would you really want to do the same on Linux? As far as I can tell, going from dialup to DSL requires a little bit of CLI, and NAT absolutely requires CLI.
NT may take longer to boot (not appriably though) but it runs faster. DirectX blits faster than SDL, windows scroll smoother in NT than in KDE2 or GNOME, menus pop up faster, apps take less time to load (than in KDE2 and GNOME at least, regular X apps are pretty fast) and 3D runs faster.
I use it because it's just good enough is the excuse people use to run Windows 95. Why bother with an OS that doesn't take full advantage of your hardware? Why pay $200 for a graphics card and be limited to the $170 that your OS uses? SDL isn't comparable to DirectX (I've used both, trust me)(SVGAlib isn't even in the same league given that it doesn't offer hardware acceleration) OpenGL is falling behind Direct3D, OpenAL isn't even ready yet, and nobody uses it yet. Overall, multimedia on Linux is pretty pathetic. I mean you can say, "ok, it's good enough for me," but I judge the quality of something benchmarked against the best product available in that catagory. Right now, NT is the best product available in the multimedia OS catagory. Sure you may be able to run MAME in Linux, but the point is that you'll be able to run it FASTER in NT.
As far as I can tell, most RPMs are compiled for i386. I COULD compile myself, but I have better things to do. Right now in Slackware, I compile glibc, X, and my window manager by myself. However, compiling stuff is really a chore I'd rather not do, and I tend to use many different programs, so it's not a compile once and forget it issue.
First, one question: do you think BeOS has great multimedia support? Heck, it doesn't even play ;)
DVD's!
>>>>
I never said it did. I don't have a DVD drive, so what do I care? When I go trumpeting BeOS as the be-all end-all multimedia OS, then take me to task for it. However, In Be's defense, I have to say the MediaKit blows away anything I've ever seen in the multi-media arena.
Windows and Linux and MacOS all have SDL.
>>>>>
SDL is just a wrapper.
It supports DirectX on Windows, and DGA on X,
which is the equivalent.
>>>>
DGA on X is equivalent to DirectX? Where's the 3D Audio API in DGA? Or maybe you mean DGA is comparable to DirectDraw? (Which still isn't totally true, show me the DGA API that allows me to change the screen-res.) DirectDraw is just a small part of DirectX.
The tests I saw for 3D acceleration didn't differ by much; it's way fast for me, even for Q3A. My speakers can't even take advantage of 3D sound decently, and I think I saw patches for all that stuff, but I really don't care yet. However, Linux *does* have great 3D
acceleration, and the audio stuff is in hardware! Maybe our definitions of great differ, but the actual
performance I'm seeing is pretty impressive. And I'm not saddened that I don't have 'force-feedback' on my Gravis Gamepad, I don't know of any games that support it, and I thought it was a dumb idea on the Playstation!
>>>>
Your arguement boils down to "I see no need for it so It's not important."
Fact> Linux 3D is slower than WindowsNT 3D. The tests showed it close (but significantly slower at high res) but they were skewed due to driver differences (I'm talking about the NVIDIA drivers.)
Fact> My speakers, and many other people's DO support 3D sound well. For these people, Linux multi-media is sub-par. Fire up Half-Life with an A3D card. It totally kicks ass and the 3D sound adds a lot to the game.
Fact> Half the hardware on my sound card goes wasted when I'm in Linux.
Fact> If you're impressed with Linux audio, take a look at DirectSound. You'd faint.
Fact> Force feedback is a dumb idea. The mere fact that you're comparing it to PS (which is just a vibrator) shows you've never used it. For certain games (racing games) it's awesome. And a LOT of game support force-feedback.
Yeah, Windows has more hardware support; that's because Microsoft doesn't have to write *ALL*
the drivers. This situation is changing of course, and I like being able to poke around with the
source code, but it'll take a while for this one to change--the corporate culture can be pretty
entrenched about these things.
>>>>>
Corporate culture aside, MS still does have more hardware support. Excuses are excuses, and drivers are actual stuff that you can base an arguement on.
There are many different limited versions of Win2K. I don't care if it supports 2 processors out of
the box, or what the configuration-of-the-week is; the bottom line is, Microsoft will always sell you
the same product for much more by just making a minor registry tweak so you get the "new
features", and I'm fundamentally opposed to that, because it's stupid; just as stupid as CPUs and
overclocking, nowadays.
>>>>>>>>>>.
Overlocking is not stupid if the consumer is the one doing it. Getting 100MHz for free by doing a little jumper manipulation? Adjusting voltages to get that perfect stability plateu? Hacking at it's finest!
Well, I'd like to see the results of the tests before I draw my conclusions; you may be right. But if I
did the tests, and one platform consistently crashed under certain conditions, I'd note that and put
it in my review; that's NOT a feature.
>>>>>>>>>
Windows is unstable compared to Linux. True. However, WindowsNT easily has a week or two of uptime, and for most gamers or multimedia people that's enough. They reboot their machines every night (Unless they're doing a rendering or something, and WindowsNT can usually handle that.) They don't install wierd software. (Hell RealPlayer flaky on BeOS and Linux too, it's not a Windows-specific problem.) For the average gamer, crashing is not an issue. A game being playable at one res setting higher on one OS (generally, you can play Quake on Windows at one res higher than you can in Linux) IS an issue.
Another benefit of Linux is the multitude of configuration options. Given the time and resources,
I'd love to just benchmark Linux against itself! That is to say, configure one box with a standard
kernel, OSS, ext2, XFree86 3.3.6, and a couple of IDE drives, and then configure another box with
an optimized kernel, ALSA, reiserfs, XFree86 4.0.1, software RAID... well, in any case benchmark
all the components against each other, and find out what the fastest, most stable Linux
configuration is for a given hardware configuration. That isn't as straightforward in NT, because
there aren't that many configuration options, and many of them aren't obvious or readily available.
>>>>>
Huh?
A) DGA is nothing compared to DirectDraw.
B) Given good driver writers (and Matrox ain't one of them) a Windows driver is rock solid. My system has a NVIDIA card and NT 4. I started out with the Detonator drivers months ago and it has yet to crash. In fact, my life in Windows-land has been pretty happy. If your Windows system is all buggy and crashy, get some NVIDIA and Creative-Labs hardware, an Intel, Asus, or Abit mobo, and a K7 or Intel chip and everything should be peachy. The idea is to be quality hardware from quality manufacturers. If you do, you enjoy not only good stability, but the increased performance of drivers that are "closer to the metal."
---
Exactly how much extra performance do those optimizations offer? From a multimedia software performance POV? I've heard that non-Pentium optimized code runs only about 70% as fast as Pentium optimized code on the Pentium. Also, I know that the i686 architecture has several quirks and speed-ups, and I don't know if most Linux distros really take full advantage of my system. Does compiling for i386 exact a peformance hit on Pentiums, and if so, why the hell do it?
Hipocracy means saying one thing and doing the other. Releasing your software under the GPL for Linux (saying that you embrace free software) while keeping it for-pay under Windows (which is essentially NOT embracing free software) is hipocracy by definition. Unless you can prove that using the GPL doesn't mean that you embrace free software, or that charging the developers on one OS and not the other is in tune with the ideals of OSS, then you've got no case.
BTW> People who correct spelling are the most anal-retentive jack-holes you'll ever meet.
Maybe because the only decent OS for 3D is Windows? Maybe because he does Windows development. Maybe he is a DTP guy and needs Photoshop and Quark XPress. (though you'd probably use a Mac in that case.) Maybe he likes having a stable web-browser. Maybe he is an AOL user? Being smart does mean hating Windows. If you're smart, you use what suits you best, not what you need to feel elite. Linux is good, but there are many tasks for which Windows is just better suited, and to think otherwise is just deluding yourself.
TW, Linux has great multimedia support. Okay, okay, let me explain. I have a Matrox G400, and the hardware
:)
acceleration is *sweet*. Also, my SB Live Value is just excellent, I love the hardware mixing, the multiple DSP's... I
installed ALSA and now the MIDI patches ("soundfonts") work too, and they sound good.
>>>>>>
Wrong. Linux doesn't have DirectX, as such, it's multimedia support by definition cannot compare with that of Windows. Not a flame, think about it. Most consumer hardware is designed for DirectX. Aside from idiot companies (ATI!) most manufacturers expose all the hardware's features through DirectX. While alternative APIs can take advantage of these features, rarely-used, or difficult to implement things always get left behind. Right now, Windows still has the fastest 3D hardware acceleration, and on NVIDIA cards (the only ones that offer competitive Linux drivers) Windows is still a good bit faster than Windows. In the tests where Linux was close to Windows, Linux had an unfair advantage. Apparently, the cool new speed-ups in the Detonator3 drivers were already in the Linux/X4 drivers. Thus, Linux 3D is close, but on cigar. As for sound, Linux is a distant second to Windows. ALSA still doesn't offer as many features as DirectSound, and all transistors on your EMU10K1 chip used for 3D sound are totally wasted in Linux. So, Linux lacks great 3D acceleration, great audio acceleration, and lacks 3D sound and force-feedback all together. How can it possibly have "great multimedia support?"
It's all about picking supported hardware, though, which you still have to do for Win2K, as well. My DVD drive isn't supported, because I (a) couldn't find much information about that on the net, and (b) just bought it first, figuring I'd test it out later. I'll probably ask the developers about this, since it works under Windows. I've heard it can be made to work under VMWare too, so all I really need is some debugging info.
>>>>>>>>
You have to choose the correct HW on Windows as well, but you've got more to choose from. On Windows, if I want great 3D acceleration, I can pick any number of cards from ATI to Matrox to NVIDIA to 3DFX (ugh!) On Linux, I'm pretty much limited to Matrox and NVIDIA (not 3DFx with X4.0)
Plus support for multi-processing makes it even better. Now, I don't have more than one processor, but I might set up
a dual-proc test box if I can ever find an old board for it. However, I've seen it done on Linux, and it is sweet. No paying extra for a different version that just consists of a stupid registry hack, either. It has decent multi-processing support out of the box. And I'd love to see a comparison to Win2K here, since that's one thing that's supposed to be
better in 2.4.
>>>>>>>>>
Huh? Win2K Pro has SMP out of box. There IS no single proc limited version of Win2K.
How about that, eh, guys? Something based around Linux Kernel 2.4.0 with a bunch of stable stuff, vs. Win2KSP1, or whatever is current and patched by then. Test multi-processing, test well-supported hardware, RAID, whatever. Just test the hell out of it.
>>>>>>
Windows would win a lot of the tests. The benefits of Linux aren't so much in raw benchmarks but in overall quality. Sure Windows may win total TCP/IP throughput scores, but it will probably crash under high load. Sure the FS may be able to transfer more date through the system (though I don't think NTFS is faster than ext2) but will it do that consistantly, or in spurts? Also, anything that taps DirectX will totally blow Linux away, since a DirectX application can for the most part be considered seperate from the underlying OS. (Especially since Win2K allows DirectX direct hardware access.)
Netscape's code is sh*t on every platform!
Both Mandrake 7.1 and Slackware 7.1 come with both egcs 1.1.2 and gcc-2.95.2. They also come with all the additional components for each compiler (egcs-java and gcc-java, etc).
BTW> OBJC=objective C.
You're missing several points in your oversimplification of OSs to include only Windows and UNIX.
There was still some DOS functionality as of NT 4.0, but I beleive it was mostly removed for Win2000.
>>>>>>
There is no DOS functionality in NT. All DOS programs are run in a virtual machine. That virtual machine is more or less unchanged in Win2K.
NT is also a microkernel, which means it naturaly has some extra overhead in it that Linux's monolithic
kernel does not. The still-mostly-vapor GNU HURD is also a microkernel. If done properly, the extra
overhead isn't that much. The question is, did Microsoft do it properly? I don't really know.
>>>>>
Yes MS did it properly. The things holding back NT don't tend to be core system related at all. It's all the stuff MS added on top that sucks. NT4 for example is a good bit faster than Linux for most desktop-type operations. However, when you look at Windows2000 with all the crap they added (active desktop and all) you notice it's much slower. NT doesn't suffer so much from core-system bloat and bugs, but stuff-added-on-top bloat and bugs.
In any case, its much easier for hackers to create a stable and speedy core system, while difficult for them to make a good UI.
>>>>>>>
Just plain wrong. BeOS: Fast stable, good UI. QNX: Fast, stable, decent UI. There are a whole bunch of systems out there that are fast and stable, and have good UIs to boot. Even Quartz seems to be pretty fast (in so far as a DPDF system can be.)
On the other hand, its far easier for a corperate project to make a reasonably good UI, but
diffcult to make a stable and speedy core system.
>>>>>
Again, not true. QNX is probably more stable than Linux and a more stable to boot (at least according to those who've used it.) Again, BeOS is managed by a corporation, and stability and speed aren't exactly high on Be users' lists of complaints.
So, basically, GPL is freedom with provisos?
Isn't that hipocracy? I mean propriatory software developers on Windows are no worse then their Linux counterparts. If Linux developers get it free, then so should Windows developers!
Uh, I don't really think $2500 is going to be a deciding factor in software development. I mean most projects spend 10x that on pizza over the life of the project! If Qt saves just 5% development time for a dozen programmers in a medium project, $2500 will mean nothing.
If you hadn't realized this, but Linux distros are full of redundant software. That's one of the main reasons I switched to Slackware. In RedHat 6.2, there are 3 versions of awk. Most distros come with two versions of gcc and their addandant language components (OBJC, Java, etc.) Mandrake comes with 3 Java compilers. (GCC-Java, Guavac, and Jikes.) Even Slackware has 3 versions of libc. Then there is all the redundancy that KDE and GNOME bring. Face it, if you were scared of a lot of code living on your computer, you wouldn't use Linux.
WTF is with the screwy "Funny" mods? Don't get me wrong, I don't mind being modded up, but why "Funny." If you think I said something incorrect, please answer instead of modding me UP. I got one the other day too for saying "Java can be compiled." What's up?
If you can, please show me one that will do that for ALSA!
Then let the best company win. My point is there is that I have never seen so much crap about the "ethics and morals and conduct" of a company than I do on Slashdot. I use whatever is the fastest, and for a lot of things (particularly anything SSE like Photoshop) Intel is the fastest.
Just because AMD also FUDs, doesn't make it okay for Intel to FUD. AMD *has* to
FUD to survive against competitor Intel's FUD. That doesn't make it alright either. I
know it interferes with the magical "free market", and the invisible hand-job and all,
but it would be nice if *someone* would take these companies to task for false
advertising when they FUD.
>>>>>>>>
There is FUD and market speak. FUD is outright lies. Market speak is streching of the truth. Intel saying the PIII makes the Internet faster is market speak. It DOES make your browser run impreceptibly faster and it helps a little on media rich and 3D rich websites (of which there are about 2.) The only people for whom market speak should matter are those too stupid to see through it. Believe it or not, the US economy thrives on taking advantage of those too stupid to do their own research. It happens with american cars, iMacs, graphics cards (wonder why ATI is on so many computers?) insurance, houses, sneakers, socks, abso-fucking-lutely everything.
Intel pushes IDE because IDE relies on CPU horsepower to initiate transfers. SCSI
relies on the controller. Therefore, if HD access takes CPU cycles, this increases the
demand (in a bogus way) for faster CPUs. Bottom line for Intel; slower and ultimately
costlier machines for consumers. Same principle works with Winmodems, and USB
too, if I'm not mistaken.
>>>>>>>>>>>
That's absoultely untrue. Under UltraDMA, one or two drives have no more CPU usage than on SCSI. Your comment was true back in the days of PIO, but not anymore. The only place SCSI helps these days is if you need the raw I/O rate, need absolute reliability, need access the the generally faster SCSI drives, or have more than two drives in you computer. Winmodems don't only exist because of Intel. In an age of $300 computers and low double digit profit margins, the $15 or so a winmodem saves means a lot more money for the manufacturer. Also, Winmodems were created when the cheapest "real" modems you could get were nearly $100. The whole point, though is moot. Intel doesn't really do anything to push Winmodems. As far as I can tell, 3Com (USRobotics) was the one the entrenched Winmodems.
I do agree that PCI was a good spec. So why did they abandon it for AGP?
>>>>>>
They didn't abandon it for AGP. As far as I can see, AGP is only a graphics standard.
If you read
Intel's press releases (and it sounds like you do, on Sunday morning, in front of an
altar, accompanied by organ music), it's to reduce the need for VRAM on the card,
but the REAL reason was so they could corner the market on graphics chips.
>>>>>>>>>>>
The VRAM issue was quite big. Hindsight is 20/20, but as I remember it, when AGP came out, the best 3D cards had 4MB of RAM (Riva128 and ATI Rage Pro). It was before the big memory price bust, and it seemed that cards could continue to use expensive SGRAM. Who the hell would have thought that in two years graphics cards would come standard with 32MB of 150MHz DDR-SDRAM? Back then (actually 6 months after AGP was introduced and years after they thought to make it) the state of the art was the V2 and it's massive 8MB (out of 12) of 100MHz texture memory. If the pace of 3D was the same from V2->Present as it was from V1->V2, then my RivaTNT 16MB would be the state of the art. However, NVIDIA totally blew apart the 3D market with it's 6month cycle. Nobody expected 3D to come this far this fast. No me, not you, not Intel.
Why
did they fail? Because their chips SUCK. They had decent competition.
>>>>>>
Actually, the i740 was pretty good. Short-lived, but pretty good. However, you're probably mistaken on Intel's intentions to corner the 3D market. If that's what they were after, they wouldn't have quite so quickly. (Do you really think Intel couldn't design a better graphics chip? Especially with Real3D's help?) My guess is that Intel was trying to make inroads into the sub-$1000 PC market with a cheap, system-memory based graphics card. At that time, AMD and Cyrix were making big strides there and it was really exploding.
And
nowadays, I challenge you to even FIND an AGP card with 4mb of VRAM. Or less than 16.
>>>>
Take a look at the advertisements for Microcenter (or CompUSA, or whatever. (A major retailer in Virginia.) All the cheap sub-$700 PCs have i740-based card with "4MB display cache."
And finally, about the AMD statement about using Slot A for higher performance,
that may be the public statement, but in Private, you can bet your loose and hairy
sphincter that the reason was so they didn't have to pay their main competitor
MORE license fees, and take it up the ass when Intel decided to change the spec
again (Slot 2, Socket 370, etc. ad infinutum, ad nauseum, ad analintrusiun), but they
took it up the ass anyway when Intel told the mobo manufacturers; "nice license you
got there, it'd be a *shame* if something were to happen to it!"
>>>>>>>>
According to AMD, they had a "blanket-cross license with Intel for all it's patents." (AMD's Presidents quote) Methinks they used SlotA more to cuddle up with Digital.
You don't really get it. The magic in CORBA and KParts isn't that they allow apps to communicate, hell, you've add that ever since signals. What's cool is it allows you to build applications in terms of other applications. It allows programmers to create apps that are mainly sets of components lashed together with glue code. UNIX modularity at it's finest.
A major gribe I have about RPM (and one of the reasons I moved to Slackware) is that it assumes that you'll do everything through RPM. It doesn't mesh well with non-RPM data. What I'd really like to see added is the ability to merge additional non-RPM-installed files into the RPM database. For example, I prefer to compile X myself, so forever afterwards, RPM keeps complaining about missing X libraries. And no, --nodeps isn't a viable option.
BTW> Maybe it already allows you to do this. However, there is nothing in the RPM docs, nor on the net that shows how.
Well, if Intel had pushed for onboard SCSI instead of going with IDE, maybe we could be using Ultra160 drives at the price of current IDE drives.
>>>
Intel wasn't the first company in the core logic chipset game. Nearly all of the early PC chipsets used IDE. It wasn't until much later that Intel started making chipsets, and at that point, IDE was already entrenched. You can't fault Intel for just doing like everybody else when they made their first chipsets.
I've seen such benchmarks; the difference is minimal until you get up to AGPx4. In fact, for most people there is no detectable difference.
>>>
Those benchmarks are for cards like the Voodoo series (up to V5) that don't use AGP effectivly. I'm talking about REAL AGP cards.
for those people with 4MBers, AGP is a godsend.
>>>>>>>
If you've only got 4MB of local graphics RAM, you're sure as hell going to see the difference between AGP and PCI. And that's not that uncommon. A lot of cheap comps come with only 4MB of memory but AGP 4x.
Not true. Apple was able to do it with G3's without moving to a slot architecture.
Untrue. It was precisely because Intel had the Slot1 specifications locked up with patents that AMD had to select a different bus architecture for the Athlon.
>>>>>>>>
Read MaximumPC (or maybe boot's) interview with AMD's President. He specifically says that they have a cross-license agreement with Intel that would allow them to use Slot1, but used SlotA for the higher performance.
That's because Intel keeps FUCKING with us. They lie, they release crappy products, they make absurd FUD-filled product roadmaps they have
no intention of fulfilling, they release products that do not ship in sufficient quantities,
>>>>>>>
Exactly what "crappy" products has Intel released? As far as I can remember, the fastest chip on the block is still the PIII 1.13 GHz. At least according to SysMark2000 (a script of 8 common applications) and Quake. Check out the reviews of the 1.1GHz Athlon at www.sharkyextreme.com. What FUD filled road-maps? Sure Itanium has been delayed several times, but hell, all chip-makers have these problems. For example, AMD and it's long delayed K5.
they perpetuate old out-dated worthless technology
(x86, ISA, IDE, etc.),
>>>>>>
x86 chips are the price/performance leaders for most things. They tried to kill ISA two or three years ago, but consumers wouldn't let it die. Intel in no way pushes IDE that hard, I think the PC market in general does that. Show me an alternative! Find me a SCSI harddrive that can match my Maxtor DiamondMax 20 at the same price (20GB for $150). Didn't think so.
and the new technology they do promote is almost always totally bass-ackwards, and designed primarily to either grab
more marketshare, or defend marketshare they have to the death, by any means other than what you'd logically think would be the BEST way to
get marketshare: make faster, more stable, more affordable chips than the competition. (USB, Slot1, AGP, etc.)
>>>>>>>>
As for pushing technologies to gain market share, that's what companies do, live with it. However, in recent memory, all of Intel's desicions have been pretty good. USB was a decent technology that freed up those IRQs. PCI was a whole lot better than VLB. USB 2.0 might not be as good as firewire, but do you really want to pay Apple licensing fees? As for AGP, I'd really like to see a performance benchmark of PCI vs. AGP. Not all of us can afford 32MB cards, and for those people with 4MBers, AGP is a godsend. Slot1 was a mistake. However, what was their alternative? They wanted backside cache, and a slot was the only way to do it at the time. AMD had to do it too. (BTW> Slot1 isn't as closed as you think. AMD's license with Intel would have allowed them to use it as well.)
And most infuriatingly, the main reason Intel seems to succeed is because stupid people buy into their "Intel Inside" campaign, that a PIII makes
surfing the internet faster.
>>>>>>>
You can't judge a company on marketing speak. Let's see, the recent Suse Slashdot banner touts Suse 6.4 as the best OS ever. That's why Linux sucks! However, you can judge them by the fact that right now, the PIII is the fastest Quake chip.