If you want to play 3D games on Linux today, you need to use binary drivers.
That is incorrect. There are 100% open source drivers for several 3D video cards. See DRI.
So, essentially, your position is that you should make significant sacrifices for no conceivable gain,
No, my position is that the long-term success of Linux depends on the distribution remaining open source. Your desire to play 3D video games using $600 video cards is relatively irrelevant.
I thought the core value of linux was to have a free, stable, fast, powerfull UNIX-like OS?
A single non-free binary driver means the system as a whole is not free. Haven't you ever wondered why many Linux distros won't bundle the nvidia driver, instead requiring the users to download it as a separate step? It's because the nvidia driver isn't free.
I agree. Unfortunately, the desktop video card market is largely binary (NVIDIA/ATI),
All the nvidia and ati cards have open source drivers. The open source drivers sometimes lack features, or don't perform as well, but they are open source. It is not correct to say that the "video card market is largely binary". That is a choice that the end-user makes.
With Linux, yes it is. If you don't care about open-source then use something else.
Don't be such an idealistic ass.
Idealism is not a dirty word. Idealism means seeing the bigger picture and foregoing fleeting fancies in the pursuit of long-term success. If you're too short-sighted to understand that then perhaps Linux is not for you.
I agree with the grandparent post. Reward nvidia with your money.
Reward nvidia for releasing *binary* drivers?
Anybody with an interest in the long-term success of Linux should be *punishing* nvidia by NOT buying their hardware.
Reward the companies that produce open source drivers, or publish specs, or help the developers of open-source drivers. Don't reward companies who are destroying the core value of Linux.
jwz had a real world problem, he's no mug when it comes to linux, but was defeated by the fact that there is no simple solution to a simple problem.
There are multiple simple solutions. He could have enabled dmix. He could have installed Fedora Core 4. He could have bought a $10 card that supports the feature he wanted. He chose a complex solution; changing all of his applications and operating system and hardware platform to solve an insignificant configuration issue. Classic prima donna behaviour.
In that time I've seen a transition from 68k to PPC and from Mac OS to OS X. The reason? It just works.
You've got the rose-coloured glasses on. If you truly used Macs during the 68k era then they didn't Just Work(tm). They crashed. Extensions were a nightmare. There was more black magic involved in MacOS than there was in DOS and Windows combined. Hardware support was spotty at best. Even when the hardware was meant to work, that was no guarantee it would work. I endlessly cursed SCSI scanners that didn't scan, or Local Talk networks that didn't talk, or soft-modems that dropped connections and ran slowly. Apple had a supported hardware list that makes Linux look enriched by comparison. The only peripherals that ever worked "perfectly" were serial-based laserwriters, even though they were SLOW. Apple's TCP/IP strategy was a joke. Applications were notoriously buggy throughout MacOS 7 and 8 days. And the 68k to PPC transition was NOT smooth, no matter what the hazy memories on Slashdot might say.
Its the lesson that the Linux advocates such as the parent poster have to learn. In the end ideology isn't important, all that matters is does it work.
The thing is, MacOS still doesn't work. I've got hardware here that doesn't work on MacOS X. TV tuners, USB wireless dongles, bluetooth phones. They've all got issues with OS X. Yet we still have fanboys claiming It Just Works. What a load of rot. More rose-coloured glasses.
I haven't learn the lesson? On the contrary, I have learnt the lesson. I've learnt that NO computer Just Works. They've all got issues and complications and brain-damaged behaviour. What I have learnt is that you need specialised knowledge no matter what OS you use. MacOS, OS X, DOS, Windows, Solaris, Linux, they're all roughly the same in terms of complexity. I don't find any of them to be a panacea for ease of use. I don't find any of them to be exceptionally difficult. They're all equally annoying in their own unique and special
ways.
So given that not one of them is a technical marvel, what does matter? Applications, of course. But all of them have roughly the same level of applications. So what else? Price, that's very important, but they're all in roughly the same ballpark for price too. So what's left? Freedom. That's the only distinguishing feature left. With Windows or MacOS X you're simply not free. You call that ideology. I call it pragmatism. If you don't know why freedom is important, then that is YOUR LOSS. You obviously haven't been burnt before. When you do figure it out, Linux will be here waiting for you. I promise I won't even say "I told you so".
So, if, in fact, the tuner type is at an undiscoverable i2c address, yes, I do have another question - the question already asked in another followup, namely "how do the vendor's Windows drivers tell the difference?"
Does the Windows driver, by virtue of (presumably) having been written by the vendor of the card, include code that knows where to look on the i2c bus to get the tuner type, so that the ultimate problem is that the vendor isn't being helpful to developers of drivers for other OSes?
Close. It's not the case that the Windows driver looks at a magic I2C location and is told the tuner type. The Windows driver already knows both the address and the tuner type. It can just blindly use the chip, secure in the knowledge that it will work. The reason the Windows driver works neatly is that the driver only needs to support a limited number of card variants. Every vendor has their own driver with their own vendor-specific tweaks. The Linux drivers are vendor-agnostic and they need to support 100s of cards.
Many of the tuner chips have "signatures", so the Linux drivers can probe the I2C bus to find vendor-specific variations. Unfortunately some of the tuners don't have signatures, or they lockup when you try and probe them. This is why in a few rare cases the user is still asked to add their tuner model as a module parameter. It's not always the case; for example, the driver for my TV card auto-detects the tuner from 3 known possibilities.
Anyway, I think that answers your question in more detail than you needed. You asked why the driver doesn't figure this out automatically. Now you know.
but how is Joe User supposed to answer these same questions? Trial and error?
Joe User is not being asked to answer those questions. Guy Harris seemed to think it was a trivial matter of querying the card. My point is that it doesn't work like that. TV tuner cards are notoriously difficult to write drivers for. They don't use discoverable busses such as PCI or USB to connect the tuner chips; they use busses like I2C.
That's the fucking problem. My company is ready to pay high premium for Linux laptops for instance. Where are they? Last time we spent $1000 of man power to configure half a dozen linux laptop.
Actually, I do have fond memories of writing 6502 assembler for the BBC Micro and 68000 for the Atari ST back in the 1980s. Back then pretty much any program you wanted beyond games you had to write yourself. And it was fun. But that was 20 years ago. There's no fun now writing low level code to get the 10,000th PCI card working.
You're the kind of person I'm talking about; willing to trade freedom for gaudy buttons and the illusion of occasional convenience, no matter how irrelevant that vaunted "convenience" actually is. I'm glad you fanboys are leaving Linux; you never understood the value of it it the first place.
I've been a corporate slave before, paying the yearly tithe and begging for scraps at the altar of Jobs. Jobs lets you "use" his software but when it suits him he pulls the rug out from under you. No longer. I'm more pragmatic than that. I figured it out long ago; freedom is more important.
If you want to keep Linux as a club for people who enjoy writing their own drivers, good luck to you. I think you'll achieve your goal.
You haven't figured it out yet. That's your problem. Linux will still be here when Jobs burns you all again and you finally get a clue.
Linux has lost momentum and OS X has gained it. More and more people have decided that there's no point in waiting for Linux to provide a good user friendly nix desktop where things just work, when OS X already offers it. People have waited long enough for Linux already.
Linux was never supposed to provide a "good user friendly nix desktop". It's supposed to be Free Software. Did you long for the days when men were real men and wrote their own drivers? That's what Linux is about. Not about some namby-pamby "I can't get play two sounds at once" whining.
The problem is that in the late-90s the ranks of Linux were swelled by disillusioned Windows fanboys who thought Linux was going to be some Microsoft killer. Now these same fanboys are returning to their proprietary non-free OS roots by migrating to MacOS X. Good riddance, I say.
The developers are "too important" to listen to the concerns of users. They have too much pride. They have their idea of how things should be done and are insulted if anyone suggests a different way.
I don't know... let's look at your examples.
How long has the GIMP had a crappy user interface?
The developers listened to the users and the latest version of GIMP sports a shiny menubar across the top of each window. It also uses mostly the same shortcuts as Photoshop.
Why does GNOME have this spatial paradigm as opposed to the more popular navigation paradigm?
Because that was the paradigm that Apple found to be more intuitive after more than a decade of research. But listening to "popularity" as if any good ever came of that (American Idol, Brittany Spears) the developers implemented a very visible preference to implement the old behaviour. Go to Edit, Preferences, Behaviour and tick the box for navigational windows.
This is the major flaw with open source software. Most of the developers are volunteering their time so they care about what interests them. Thats fine, no one should tell them what they should be spending their own time doing. But until Open Source "grows up" and starts listening to its users it will never be popular and shouldn't expect to be.
The thing is, you're clearly not a user or you would know these things were already fixed. In the case of the GIMP they were fixed 12 MONTHS AGO.
Well apparently you are. The grandparent was pointing out that all platforms - including Windows and OSX - sometimes have problems working with various pieces of hardware. Your response is to start spouting techno bibblety-babble about how many accessories you own.
Here's one for you. I've got a Mac here which doesn't support a PCI TV Tuner card or a USB TV Tuner dongle. Both pieces work in Windows XP and in Linux (Debian). Now if I was to do a JWZ (JWZ being a verb for having a pretentious hissy fit) then I'd write a scathing blog about how crappy OSX is and how I'm switching to Linux.
Or the more reasonable person that I am would say that sometimes you should check the supported hardware list before blaming the OS for not supporting some obscure piece of hardware.
Can a driver determine the tuner type by querying the card?
No. It's hidden behind an i2C bus at an unknown address and is not discoverable. Even within the same model of card, even using the same PCI vendor and product codes, the manufacturers have changed tuner chips and i2c addresses.
Like most things at Slashdot, there is a double standard at play here. In other words, the Slashdot fanboys are not as pue as they like to think of themselves as. If it's bad for MS, it's good "just because". Pay backs, you know? Like little children...
Take your smarmy pseudo-intellectual "I'm so much smarter than all the plebs on Slashdot" attitude and shove it.
And NO ONE, will run Linux on Apple-intel hardware.
Why not? People run Linux on Apple/PPC hardware. The Apple laptops are simply niftier than the x86 equivalents and they are roughly the same price. If Apple's x86 laptop is also nifty and within the same ballpark price (say 5-10%) I will buy one.
Do you own a car? You do realize that there are various patents and copyrights covering your car?
Do you own a microwave oven? Once again, there are patents and copyright limitations covering it too.
Do you own a TV? Same thing.
Why is it that when it comes to software you demand complete freedom, but when it comes to everything else, such freedom is irrelevant?
Because software isn't a car. Software isn't a TV. Software isn't a microwave.
Software already has protection via copyright and trade secrets. Thanks to copyright, and the very nature of source code vs machine code, we can't see how closed software works. We can't modify it. We can't improve it. We can't learn from it. It's a black box, never to be opened. And thanks to patents we can't even make another piece of software that WORKS like the original.
My car might have patents but nobody owns the copyright to my car. My television might use a radical new form of electron gun but nobody will sue me for building my own TV. With patents I'm supposed to be able to see how the invention works; that's the balance that patents are supposed to provide. Where's that balance with software patents?
Books have copyrights but I don't see anybody claiming a patent for murder mysteries. Music can be copyrighted but nobody owns a patent on Rock Ballads. With copyright the original is supposed to pass into the public domain for the good of all humanity. With software copyright, where is the balance? The knowledge is still locked up in the source code which we NEVER SEE.
The software manufacturers are simply greedy; they want copyright protection AND patent protection AND trade secrets. They want copyright on the machine code and trade secrets for the source, so the public NEVER receives the intended balance. They want patents on the algorithms so nobody can compete, but if you can't see the code then how can you know when you infringe? Once again, where is the balance?
I'd like to see a simple rule applied here; software can have patent protection, or copyright protection, but not both. If you choose patents then you must publish your source code and in 20 years time it's in the public domain. If you choose copyright then you can keep the source code a secret but you can't enforce patents. That would go some way towards restoring the balance.
Err, you mean a Mach kernel with a BSD userspace, right? With a nice GUI written predominantly in ObjC? You know what architecture OS X is, don't you? (And which strange but wonderful C derivative is predominantly used on OS X?)
Can you put two+two together?
I think he knows that. And I think you missed his point.
So not only has Apple dumped IBM, they also appear to be planning to dump gcc.
That's certainly possible, though in hindsight it's now obvious why they used GCC in the first place. They were simultaneously developing for PowerPC and x86 platforms. The Intel and IBM compilers didn't support both; GCC did.
You are in way over your head. You don't know the difference between an architecture and a specific CPU.
How amusing; yet another thing you don't know anything about that yet you feel to need to speak with authority. For the record, I'm a graduate engineer with specific training in embedded system design and over a decade of real-world experience.
...frothing at the mouth ranting...
Certainly seems that I hit a nerve there. You might want to clean the spittle off your keyboard.
That's why I've been paid to design, and review designs for, embedded systems.
Then I advise you to give them back their $3.50 because you were apparently overpaid.
0 / any$ = 0. If the CPU has no value in a given application, it does not matter how cheap it is. A Zilog Z80 can be had for probably $5. That doesn't mean that it has great bang for the buck if your goal is to build a desktop system.
You keep changing your tune. First the chip isn't adequate for a desktop. Then when you're proven wrong you talk about price. Then when several people point out that PowerPC is cheaper than AMD or Intel you say the price is irrelevant and what matters is "bang for buck". Then when I point out that price can never be irrelevant when you're talking "bang for buck" you go back to your initial argument that the PPC isn't adequate for a desktop. Hadn't you noticed that PowerPC desktops are rather popular? Apple is selling millions of them.
I think I spot the underlying problem here; you don't actually know how much the PowerPC costs. The PowerPC is one of the most popular embedded CPUs - used in cars, phones, routers, consoles, multimedia devices - because systems designed around the PowerPC are inexpensive. Embedded systems are the litmus test for cost efficiency. They measure CPUs in $ per MIP; that's "bang for buck" by another name. Are all the embedded designers stupid? Would they save more money by listening to you?
Of course not. You're just wrong. You don't know what you're talking about.
But it doesn't mean that it's got adequate performance for a modern desktop system.]
The PowerPC is certainly adequate for a modern desktop system. The popularity of Apple desktops is proof enough of that. The CPU hasn't been a hinderance to desktop performance for many years now; even the lowliest Celeron or G4 has enough computing power for word processing, web browsing and email.
So, in that usage, it has incredibly poor bang for the buck, regardless of price.
I don't see how it's possible for you to argue "bang for buck" while disregarding the price.
Slashdot Mirror
That is incorrect. There are 100% open source drivers for several 3D video cards. See DRI.
No, my position is that the long-term success of Linux depends on the distribution remaining open source. Your desire to play 3D video games using $600 video cards is relatively irrelevant.
A single non-free binary driver means the system as a whole is not free. Haven't you ever wondered why many Linux distros won't bundle the nvidia driver, instead requiring the users to download it as a separate step? It's because the nvidia driver isn't free.
All the nvidia and ati cards have open source drivers. The open source drivers sometimes lack features, or don't perform as well, but they are open source. It is not correct to say that the "video card market is largely binary". That is a choice that the end-user makes.
With Linux, open source *is* the real issue.
The core value of Linux is that it is open source.
Good. Linux will be better off without them if the mere mention of "open source" is enough to scare them away.
With Linux, yes it is. If you don't care about open-source then use something else.
Idealism is not a dirty word. Idealism means seeing the bigger picture and foregoing fleeting fancies in the pursuit of long-term success. If you're too short-sighted to understand that then perhaps Linux is not for you.
Reward nvidia for releasing *binary* drivers? Anybody with an interest in the long-term success of Linux should be *punishing* nvidia by NOT buying their hardware.
Reward the companies that produce open source drivers, or publish specs, or help the developers of open-source drivers. Don't reward companies who are destroying the core value of Linux.
Go Blender!
There are multiple simple solutions. He could have enabled dmix. He could have installed Fedora Core 4. He could have bought a $10 card that supports the feature he wanted. He chose a complex solution; changing all of his applications and operating system and hardware platform to solve an insignificant configuration issue. Classic prima donna behaviour.
You've got the rose-coloured glasses on. If you truly used Macs during the 68k era then they didn't Just Work(tm). They crashed. Extensions were a nightmare. There was more black magic involved in MacOS than there was in DOS and Windows combined. Hardware support was spotty at best. Even when the hardware was meant to work, that was no guarantee it would work. I endlessly cursed SCSI scanners that didn't scan, or Local Talk networks that didn't talk, or soft-modems that dropped connections and ran slowly. Apple had a supported hardware list that makes Linux look enriched by comparison. The only peripherals that ever worked "perfectly" were serial-based laserwriters, even though they were SLOW. Apple's TCP/IP strategy was a joke. Applications were notoriously buggy throughout MacOS 7 and 8 days. And the 68k to PPC transition was NOT smooth, no matter what the hazy memories on Slashdot might say.
The thing is, MacOS still doesn't work. I've got hardware here that doesn't work on MacOS X. TV tuners, USB wireless dongles, bluetooth phones. They've all got issues with OS X. Yet we still have fanboys claiming It Just Works. What a load of rot. More rose-coloured glasses.
I haven't learn the lesson? On the contrary, I have learnt the lesson. I've learnt that NO computer Just Works. They've all got issues and complications and brain-damaged behaviour. What I have learnt is that you need specialised knowledge no matter what OS you use. MacOS, OS X, DOS, Windows, Solaris, Linux, they're all roughly the same in terms of complexity. I don't find any of them to be a panacea for ease of use. I don't find any of them to be exceptionally difficult. They're all equally annoying in their own unique and special ways.
So given that not one of them is a technical marvel, what does matter? Applications, of course. But all of them have roughly the same level of applications. So what else? Price, that's very important, but they're all in roughly the same ballpark for price too. So what's left? Freedom. That's the only distinguishing feature left. With Windows or MacOS X you're simply not free. You call that ideology. I call it pragmatism. If you don't know why freedom is important, then that is YOUR LOSS. You obviously haven't been burnt before. When you do figure it out, Linux will be here waiting for you. I promise I won't even say "I told you so".
Close. It's not the case that the Windows driver looks at a magic I2C location and is told the tuner type. The Windows driver already knows both the address and the tuner type. It can just blindly use the chip, secure in the knowledge that it will work. The reason the Windows driver works neatly is that the driver only needs to support a limited number of card variants. Every vendor has their own driver with their own vendor-specific tweaks. The Linux drivers are vendor-agnostic and they need to support 100s of cards.
Many of the tuner chips have "signatures", so the Linux drivers can probe the I2C bus to find vendor-specific variations. Unfortunately some of the tuners don't have signatures, or they lockup when you try and probe them. This is why in a few rare cases the user is still asked to add their tuner model as a module parameter. It's not always the case; for example, the driver for my TV card auto-detects the tuner from 3 known possibilities.
Anyway, I think that answers your question in more detail than you needed. You asked why the driver doesn't figure this out automatically. Now you know.
Joe User is not being asked to answer those questions. Guy Harris seemed to think it was a trivial matter of querying the card. My point is that it doesn't work like that. TV tuner cards are notoriously difficult to write drivers for. They don't use discoverable busses such as PCI or USB to connect the tuner chips; they use busses like I2C.
Right here. http://www.emperorlinux.com/
Yeah, you'll pay a premium alright, but it'll Just Work.
You're the kind of person I'm talking about; willing to trade freedom for gaudy buttons and the illusion of occasional convenience, no matter how irrelevant that vaunted "convenience" actually is. I'm glad you fanboys are leaving Linux; you never understood the value of it it the first place.
I've been a corporate slave before, paying the yearly tithe and begging for scraps at the altar of Jobs. Jobs lets you "use" his software but when it suits him he pulls the rug out from under you. No longer. I'm more pragmatic than that. I figured it out long ago; freedom is more important.
You haven't figured it out yet. That's your problem. Linux will still be here when Jobs burns you all again and you finally get a clue.
Linux was never supposed to provide a "good user friendly nix desktop". It's supposed to be Free Software. Did you long for the days when men were real men and wrote their own drivers? That's what Linux is about. Not about some namby-pamby "I can't get play two sounds at once" whining.
The problem is that in the late-90s the ranks of Linux were swelled by disillusioned Windows fanboys who thought Linux was going to be some Microsoft killer. Now these same fanboys are returning to their proprietary non-free OS roots by migrating to MacOS X. Good riddance, I say.
I don't know... let's look at your examples.
The developers listened to the users and the latest version of GIMP sports a shiny menubar across the top of each window. It also uses mostly the same shortcuts as Photoshop.
Because that was the paradigm that Apple found to be more intuitive after more than a decade of research. But listening to "popularity" as if any good ever came of that (American Idol, Brittany Spears) the developers implemented a very visible preference to implement the old behaviour. Go to Edit, Preferences, Behaviour and tick the box for navigational windows.
The thing is, you're clearly not a user or you would know these things were already fixed. In the case of the GIMP they were fixed 12 MONTHS AGO.
Well apparently you are. The grandparent was pointing out that all platforms - including Windows and OSX - sometimes have problems working with various pieces of hardware. Your response is to start spouting techno bibblety-babble about how many accessories you own.
Here's one for you. I've got a Mac here which doesn't support a PCI TV Tuner card or a USB TV Tuner dongle. Both pieces work in Windows XP and in Linux (Debian). Now if I was to do a JWZ (JWZ being a verb for having a pretentious hissy fit) then I'd write a scathing blog about how crappy OSX is and how I'm switching to Linux.
Or the more reasonable person that I am would say that sometimes you should check the supported hardware list before blaming the OS for not supporting some obscure piece of hardware.
No. It's hidden behind an i2C bus at an unknown address and is not discoverable. Even within the same model of card, even using the same PCI vendor and product codes, the manufacturers have changed tuner chips and i2c addresses.
Any other questions?
Feed Me, Seymour!
Take your smarmy pseudo-intellectual "I'm so much smarter than all the plebs on Slashdot" attitude and shove it.
Why not? People run Linux on Apple/PPC hardware. The Apple laptops are simply niftier than the x86 equivalents and they are roughly the same price. If Apple's x86 laptop is also nifty and within the same ballpark price (say 5-10%) I will buy one.
Because software isn't a car. Software isn't a TV. Software isn't a microwave.
Software already has protection via copyright and trade secrets. Thanks to copyright, and the very nature of source code vs machine code, we can't see how closed software works. We can't modify it. We can't improve it. We can't learn from it. It's a black box, never to be opened. And thanks to patents we can't even make another piece of software that WORKS like the original.
My car might have patents but nobody owns the copyright to my car. My television might use a radical new form of electron gun but nobody will sue me for building my own TV. With patents I'm supposed to be able to see how the invention works; that's the balance that patents are supposed to provide. Where's that balance with software patents?
Books have copyrights but I don't see anybody claiming a patent for murder mysteries. Music can be copyrighted but nobody owns a patent on Rock Ballads. With copyright the original is supposed to pass into the public domain for the good of all humanity. With software copyright, where is the balance? The knowledge is still locked up in the source code which we NEVER SEE.
The software manufacturers are simply greedy; they want copyright protection AND patent protection AND trade secrets. They want copyright on the machine code and trade secrets for the source, so the public NEVER receives the intended balance. They want patents on the algorithms so nobody can compete, but if you can't see the code then how can you know when you infringe? Once again, where is the balance?
I'd like to see a simple rule applied here; software can have patent protection, or copyright protection, but not both. If you choose patents then you must publish your source code and in 20 years time it's in the public domain. If you choose copyright then you can keep the source code a secret but you can't enforce patents. That would go some way towards restoring the balance.
I think he knows that. And I think you missed his point.
That's certainly possible, though in hindsight it's now obvious why they used GCC in the first place. They were simultaneously developing for PowerPC and x86 platforms. The Intel and IBM compilers didn't support both; GCC did.
How amusing; yet another thing you don't know anything about that yet you feel to need to speak with authority. For the record, I'm a graduate engineer with specific training in embedded system design and over a decade of real-world experience.
Certainly seems that I hit a nerve there. You might want to clean the spittle off your keyboard.
Then I advise you to give them back their $3.50 because you were apparently overpaid.
You keep changing your tune. First the chip isn't adequate for a desktop. Then when you're proven wrong you talk about price. Then when several people point out that PowerPC is cheaper than AMD or Intel you say the price is irrelevant and what matters is "bang for buck". Then when I point out that price can never be irrelevant when you're talking "bang for buck" you go back to your initial argument that the PPC isn't adequate for a desktop. Hadn't you noticed that PowerPC desktops are rather popular? Apple is selling millions of them.
I think I spot the underlying problem here; you don't actually know how much the PowerPC costs. The PowerPC is one of the most popular embedded CPUs - used in cars, phones, routers, consoles, multimedia devices - because systems designed around the PowerPC are inexpensive. Embedded systems are the litmus test for cost efficiency. They measure CPUs in $ per MIP; that's "bang for buck" by another name. Are all the embedded designers stupid? Would they save more money by listening to you?
Of course not. You're just wrong. You don't know what you're talking about.
The PowerPC is certainly adequate for a modern desktop system. The popularity of Apple desktops is proof enough of that. The CPU hasn't been a hinderance to desktop performance for many years now; even the lowliest Celeron or G4 has enough computing power for word processing, web browsing and email.
I don't see how it's possible for you to argue "bang for buck" while disregarding the price.