The winner was a guy who wasn't even competing using his off-the-shelf laptop.
That says a lot about the lack of skill of the participants.
PCs CAN be made more efficient than notebooks. Using a notebook CPU (Turion/ULV Pentium-M), 80%+ efficient PSU, perhaps a notebook HDD, and then a small LCD. preferably without a backlight.
Not to mention, using a notebook every day isn't pleasant.
Except, of course, the goal was to put together a system fast enough to be useful.
It's sad how far VIA has gone on their bullshit claims of energy efficiency, when certain older, faster CPUs from both Intel and AMD are lower power. Not to mention the top-of-the-line mobile CPUs like the Turion only uses 2X the power VIA claims it's processors use, yet AMD has CnQ power management, which will put the Turion well below VIA's chips in average use.
There's good reason you don't see VIA CPUs in notebooks. Their entire business model is fooling unsuspecting people, with clever marketing, and out-of-context numbers. They don't actually WORK.
Just thinking about it makes me want to sell my desktop on ebay and use the money to stock up on old Thinkpads to save for the future.
Using a Laptop as your desktop machine is not fun at all, nor cheap.
Meanwhile, you can put together a desktop system, that is almost as low power as a notebook.
Thinkpads commonly used ultra low voltage Pentium 3s. You too can buy a P3 motherboard, buy something like a PIII 933, and undervolt it. It will run fanless with decent (passive) airflow.
But for something that doesn't have horrible performance, you can pick up a 2GHz+ 25W Turion for $80, and a $40 Socket 754 motherboard.
In either case, use the most efficient PSU you can get, and a small LCD, and together they will average maybe 30W, possibly less. And you can still use as much desktop hardware as you want.
According to the elliptical machine at the gym, I burn through about 1000 calories in the first hour of cardio.
That's probably some sort of estimate based on how much heat your body is generating, and such things. Definately NOT actual energy being dissipated by the machine.
The How Stuff Works page you linked is quite accurate, you should have read it before posting:
"A "normal person" might be able to sustain a third of a horsepower for half an hour before falling off the bike from exhaustion."
and
"1 horsepower is equal to 746 watts."
Even the strongest athletes in the world couldn't possibly generate half a kW an hour, while you believe you are doing more than double that...
Even someone in incredibly physical shape, able to sustain 1/5th horsepower (1/10th is more accurate), and using the most inflated number for electricity you've given, you're still talking less than 1 cent per hour. And that's as hopelessly optimistic it gets, on all counts. No mention of cost, losses, rates, etc.
Before computers were used in real engineering, we could get away with "k" sometimes meaning 1024 (like in memory addresses) and sometimes meaning 1000
BEFORE? So the ENIAC was built to play solitare?
Engineering was one of the very FIRST uses of computers, it's the frivolous crap that is the recent addition.
First off, it's nice to have energy-efficient everything. However, a world of attention is being paid to squeezing a few more percent out of PCs, which are using about as much power on average as a lightbulb. I'd really suggest you take a look at your refrigerator before spending lots of money on low-power PC components.
Just a few days ago, I was looking for new power supplies. The cheapest I found 80%+ PSUs like Seasonic were over $40, meanwhile, 70% efficient PSUs are $10 (both prices including S+H). It will take quite a while to pay off the difference in electricity, even here in CA, and my PSUs don't seem to survive very long to begin with.
Incidentally, is this guy a complete moron???
From TFA:
A PC uses 200 to 400 watts, depending on its configuration and use.
They certainly can, but most don't. Mine max-out around 90W + 30W LCD.
In fact, most computers drain more power than they need during normal operation,
Everything drains more power than needed. Nothing is 100% efficient, nor can it possibly be.
If a PSU meets the certification [...] Only when the PC requires full power will the PSU run at the full wattage load.
As opposed to non-certified PSUs that run at 500W when the system only needs 20? What? That doesn't even remotely make any sense.
an inefficient 500-watt PSU typically drains more than 500 watts of power.
An efficient 500-watt PSU always drains more than 500 watts of power as well...
If you've got the money, upgrade your green PC with a better drive once they are released.
Well, he's just completely defeated the purpose of this "green PC" by telling people to throw away perfectly good working components. Good job.
Also, it's hard to take his "green PC" seriously with a Core 2 Duo, instead of something like a Turion (or a Geode like the OLPC), which would uses about 1/4th the power. Saying it's "green" because it is lower power than a P4 is setting the bar pretty low...
I wasn't asking you to believe anything. I was asking you to prove your point, and show you aren't just talking complete bullshit.
Again supporting my original (great grand parent) post: for the situations (L)GPL is better.
That doesn't "support" a thing. You have to _prove_ that a similar but BSD licensed project would fail, to make any such claims. The fact that the GPL and LGPL are popular is not proof of superiority.
Tremor is for "low-end" hardware - DSPs & FPGAs.
It's for anything that can do integer math. In fact, it's faster than libvorbis on standard PCs.
libvorbis itself is used by many embedded systems.
By all means, point out a few.
There, vorbis isn't considered software - it is part of firmware and thus part of hardware.
And what about hardware that needs HTML rendering? Shouldn't we BSD license Gecko and KHTML for hardware, now?
"Too many watermills: water dissappears from rivers!"
It wouldn't disappear, of course. It would be trapped upstream, causing plenty of environmental issues there, and completely drying up the river further downstream.
As long as you make sure the new hardware fits into the slots that come with your computer and does not overburden its power supply, it's usually just a case of carefully lifting out the old and slotting in the new piece of kit.
He should tell that to ATI... My 8500DV won't power-up in my VIA system, but it works in my SIS system. Ditto for an NVidia PCI videocard. No rhyme, no reason, just screwed.
Many problems with RAM and CPU upgrades... Most people can't figure out setting hard drive/CD/DVD jumpers...
So, yeah, I guess if there's anything else in your system left, upgrading it won't be very hard. Otherwise, consult your local geek before you start, and risk doing serious damage.
I had read somewhere that a Vista installation would take 20 minutes. Not if you upgrade from XP.
Everyone should know by now that Microsoft has NEVER been able to handle upgrades worth a damn. Though I know of many that have tried, I haven't seen a single successful upgrade from any version of Windows, to any other.
Once online, Creative's website told me that my sound card was a write-off. No Vista support would be forthcoming.
Am I crazy? Did I miss something? Or is the rest of the world perhaps stupid? From Windows 2000 and on, backwards driver backwards compatibility was built-in, and it's not unusually for Windows to select (by default) a driver for a different Windows version, before the one you would expect (signed/unsigned issues). Has Vista stopped this? I can't imagine there would be any problems with something as simple as soundcard drivers.
You can't "run out" of wind or "stop" wind by using wind energy any more than you can "use up" a river with a mill wheel.
You can indeed run out of wind, slowing it down to nearly nothing with enough windmills. Ditto for "a mill wheel". Each one removes energy from the water, changes the environment up and downstream. With enough of them, the water will practically come to a stop and never reach its destination.
The OP's claims are still bullshit. Using wind-power doesn't mean we need to use 100% of it... Things like dams have worked out wonderfully, and although we can't expand on them much more than we can, what they provide is great, and far better than the alternatives, NOT utilizing hydro.
What if they started "paying" (like 1 buck per somethingsomething) people for producing energy.
Humans can sustain about 1/10th horsepower. At that, you'd have to work out for 12+ hours, non-stop, to make maybe 2 cents worth of electricity.
That's with no mention of efficiency, maintenance, infrastructure, etc.
I don't think 1/20th of a cent per hour you exercise on a stationary bicycle is going to be much motivation to people... especially when a gym membership is $20+ per month to start with.
It's probably just barely worth harnessing it when they're not paying people anything for their effort.
kHTML people proved the important point: HTML/DOM/CSS/JavaScript are no rocket science and can be implemented from ground up with some effort applied. What gained Apple? It has probably saved couple of years of works for several engineers. What gained KDE? It had become known in MacOS (and Wintel PC) world as providers of good quality software. End result, is that KDE gained more from Apple, than Apple gained from KDE.
And how would that have been difficult/impossible with the BSD license, rather than the GPL?
Vorbis is intended to be implemented in hardware: attaching any kind of license is hurdle.
If that was the case, they would have left libVorbis GPL'd, and just BSD licensed Tremor. They did not.
It seems pointless to seek ideas and feedback if you're going to ignore and delete the opinions you don't like. That's exactly what Intel and Slashdot has done with its "Opinion Center" website.
For example, a retailer charges $15 up front, must accept hardware for recycling, and gives you $10 back for each computer turned in for recycling.
I agree completely, except I'd make it a smaller value. Even $5 for things like car batteries and a couple cents a piece for aluminum cans is more than enough to have the homeless/unemployed going around, collecting them for the money. No doubt somebody could make a good business out of digging them out of landfills, and recycling them as well.
an this actually be considered a satellite transmitter as it is only beaming signal to a receiver on top a hill?
For less strict values of "satellite" yes... As in, not orbiting the earth.
Satellite: "3. something, as a branch office or an off-campus facility of a university, that depends on, accompanies, or serves something else. 4. an attendant or follower of another person, often subservient or obsequious in manner. "
It is just confirmation of old statement that GPL(v2) provides better (at moment best) ground for cooperation between vendors.
Not even slightly. FreeBSD has plenty of vnedor support as well.
But even more so, GPL allows Linux to "merge" back possible code base "forks". That's next to impossible with BSD licensed code most tend to keep closed.
Not at all. Following the strictest of definitions of the GPL, you can easily create something that is difficult or near impossible to merge back into the base code. This was a big complaint of KHTML developers with Apple.
Those who contribute back do it because they want to, NOT because they are FORCED to.
Whats more the BSD license allows MORE people and companies to use the software for more purposes... RMS understands this, and gave the okay to Xiph.org to BSD Vorbis and Theora... If you're forced to release the source, you're prevented from using it in a lot of places.
As such, I can think of many BSD (and MIT) licensed programs that have become defacto standards (eg. Telnet, FTP, SMTP, DNS, NFS, SSH, and many more) but none that were GPL licensed... How many can you think of?
Sure, you can GPL your high-tech encrypted network filesystem, and force everyone who uses it to release their code changes, but as a results, you'll just not get ANYONE using it to begin with. No matter how much better than NFS it might be, it will be soon forgotten, and you can continue to be smug about how the GPL help you stop others from stealing your code...
(Note: Use whatever license you want, but don't make moronic claims about how perfect your choice is over someone else's.)
Here on the east coast, we don't have fry's electronics.
No Fry's within 100 miles here on the west coast, either...
the best buy usually has like 3 video cards, 1 sub-par motherboard, and approximately 23 hojillion music cd's, dvd's, and washing machines.
Yeah, it's the same everywhere... and it has been exactly that for 10+ years. Don't even get me started on their selection of monitors, and the utter shittyness of what they chose to stock.
For all its shortcomings, when my hard drive failed, i went to compusa to get back up and running the same day,
Worst Buy does, however, always have numerous hard drives on hand, at decent prices...
I called the Microsoft authentication hotline, which had an automated voice that insisted that I read out loud the 54-digit code that Windows provided. I tried typing it in, but it ignored me. When I read it out loud, the Hotline told me that number was invalid and hung up on me.
Your experience closely mirrors my own, just a month ago. Except in my case it was an HP system, and after getting fucked-over once, I went to http://astalavista.box.sk/ and downloaded and used WGA crack v2.0. Problem solved.
If there's a technological measure preventing me from using my property, I have the absolute right to bypass it, and will proudly defend that in court if Microsoft or anyone else even considers coming after me for a "pirated" version of XP. If the next update breaks anything, you can bet I'm going to be as big a nuisance to them, as they've been to me.
The other suggestions are quite apt. Even with a dial-up or wireless connection, remotely accessing a server should be just fine, as long as your apps aren't games, high resolution videos, or something else EXTREMELY graphically intensive.
You should see if there's any other way to get your job done with better tools, that aren't such RAM hogs. The fact that notebooks with 4GBs is hard to find is a good indication that nobody else has such a problem, and there's probably a better alternative.
But, if those really aren't possibilities: I would go for a home-built lugable Opteron desktop. It's pretty easy to find low-profile components like heatsinks, PSUs, etc, and even 1/2U servers. Throw in 16GBs of DDR, and find a simple way to clip the LCD and keyboard to the unit. It probably wouldn't be any heavier than the larger "desktop-replacement" notebooks.
AMD can, for limited periods of time, pull ahead of Intel. That initially surprised a bunch of people
It surprised people... 10 years ago? When AMD released the K6? And the short period of time is EVER SINCE THEN? It's been Intel that has shortly had the better performing products in that time-period, and then, it was usually only on floating point performance, while most apps are integer-based...
Even now, I don't believe that Intel has the lead, or at least not much of one... Intel's chips usually don't have 64-bit support, and where they do, they usually have bugs that make it useless, like disabling DMA... AMD chips, however, see a big performance boost in 64-bit mode, on the order of 20%.
*Note: Not a fanboi by any stretch of the imagination. I rake either company over the coals when they do stupid shit.
- Ogg is a container like Matroshka (MSK) or AVI (but better than that one. Almost anything is better than AVI)
Gah!
First of all, it's Matroska and the extension is MKV.
Second, whether AVI is better or worse than Ogg is debatable. Any who has ever written an Ogg (de)muxer curses it's name frequently. It's extremely codec-specific, and the format is rather loosely defined, with no consistent standard way to do much of anything....
AVI has it's limitations, but they are few. Most of the problems people experience with AVI is due entirely to limited software which doesn't properly handle AVIs. The rest of the problems tend to be a result of lack of standards... For instance, Vorbis can fit into AVI just fine, but unfortunately, Xiph didn't define HOW exactly, so everyone has started doing it in their own, mutually incompatible way. Ditto for subtitles, and other meta-data.
So, the biggest problem with AVI is lack of any single official standards authority. Ogg has the same problem, but worse, since Xiph have ignored all efforts to extend Ogg to handle other formats, and now nothing is compatible.
I believe it's also more processor-intensive, so you need faster hardware to do the decoding. It's certainly harder on the batteries in any device that can handle it.
Vorbis is, but Musepack isn't. Quite the opposite. MPC is faster than MP3s for both encoding and decoding. It's patent-free, and sounds slightly better than Vorbis in most cases... Significantly better in certain cases.
Slashdot Mirror
That says a lot about the lack of skill of the participants.
PCs CAN be made more efficient than notebooks. Using a notebook CPU (Turion/ULV Pentium-M), 80%+ efficient PSU, perhaps a notebook HDD, and then a small LCD. preferably without a backlight.
Not to mention, using a notebook every day isn't pleasant.
Except, of course, the goal was to put together a system fast enough to be useful.
It's sad how far VIA has gone on their bullshit claims of energy efficiency, when certain older, faster CPUs from both Intel and AMD are lower power. Not to mention the top-of-the-line mobile CPUs like the Turion only uses 2X the power VIA claims it's processors use, yet AMD has CnQ power management, which will put the Turion well below VIA's chips in average use.
There's good reason you don't see VIA CPUs in notebooks. Their entire business model is fooling unsuspecting people, with clever marketing, and out-of-context numbers. They don't actually WORK.
(*Unhappy owner of a VIA CPU.)
Particularly when it costs you $2,000 to save that 12 watts per hour.
Oh, you meant a POSITIVE difference? Sorry.
Guess what? You're generating lots of CO2 every hour you're working to pay off the price tag of that not-so "green PC".
Using a Laptop as your desktop machine is not fun at all, nor cheap.
Meanwhile, you can put together a desktop system, that is almost as low power as a notebook.
Thinkpads commonly used ultra low voltage Pentium 3s. You too can buy a P3 motherboard, buy something like a PIII 933, and undervolt it. It will run fanless with decent (passive) airflow.
But for something that doesn't have horrible performance, you can pick up a 2GHz+ 25W Turion for $80, and a $40 Socket 754 motherboard.
In either case, use the most efficient PSU you can get, and a small LCD, and together they will average maybe 30W, possibly less. And you can still use as much desktop hardware as you want.
That's probably some sort of estimate based on how much heat your body is generating, and such things. Definately NOT actual energy being dissipated by the machine.
The How Stuff Works page you linked is quite accurate, you should have read it before posting:
"A "normal person" might be able to sustain a third of a horsepower for half an hour before falling off the bike from exhaustion."
and
"1 horsepower is equal to 746 watts."
Even the strongest athletes in the world couldn't possibly generate half a kW an hour, while you believe you are doing more than double that...
Even someone in incredibly physical shape, able to sustain 1/5th horsepower (1/10th is more accurate), and using the most inflated number for electricity you've given, you're still talking less than 1 cent per hour. And that's as hopelessly optimistic it gets, on all counts. No mention of cost, losses, rates, etc.
BEFORE? So the ENIAC was built to play solitare?
Engineering was one of the very FIRST uses of computers, it's the frivolous crap that is the recent addition.
So, yeah, your whole post is utter nonsense.
Just a few days ago, I was looking for new power supplies. The cheapest I found 80%+ PSUs like Seasonic were over $40, meanwhile, 70% efficient PSUs are $10 (both prices including S+H). It will take quite a while to pay off the difference in electricity, even here in CA, and my PSUs don't seem to survive very long to begin with.
Incidentally, is this guy a complete moron???
From TFA:
They certainly can, but most don't. Mine max-out around 90W + 30W LCD.
Everything drains more power than needed. Nothing is 100% efficient, nor can it possibly be.
As opposed to non-certified PSUs that run at 500W when the system only needs 20? What? That doesn't even remotely make any sense.
An efficient 500-watt PSU always drains more than 500 watts of power as well...
Well, he's just completely defeated the purpose of this "green PC" by telling people to throw away perfectly good working components. Good job.
Also, it's hard to take his "green PC" seriously with a Core 2 Duo, instead of something like a Turion (or a Geode like the OLPC), which would uses about 1/4th the power. Saying it's "green" because it is lower power than a P4 is setting the bar pretty low...
This is an awfully brain-dead article for
I wasn't asking you to believe anything. I was asking you to prove your point, and show you aren't just talking complete bullshit.
That doesn't "support" a thing. You have to _prove_ that a similar but BSD licensed project would fail, to make any such claims. The fact that the GPL and LGPL are popular is not proof of superiority.
It's for anything that can do integer math. In fact, it's faster than libvorbis on standard PCs.
By all means, point out a few.
And what about hardware that needs HTML rendering? Shouldn't we BSD license Gecko and KHTML for hardware, now?
It wouldn't disappear, of course. It would be trapped upstream, causing plenty of environmental issues there, and completely drying up the river further downstream.
He should tell that to ATI... My 8500DV won't power-up in my VIA system, but it works in my SIS system. Ditto for an NVidia PCI videocard. No rhyme, no reason, just screwed.
Many problems with RAM and CPU upgrades... Most people can't figure out setting hard drive/CD/DVD jumpers...
So, yeah, I guess if there's anything else in your system left, upgrading it won't be very hard. Otherwise, consult your local geek before you start, and risk doing serious damage.
Everyone should know by now that Microsoft has NEVER been able to handle upgrades worth a damn. Though I know of many that have tried, I haven't seen a single successful upgrade from any version of Windows, to any other.
Am I crazy? Did I miss something? Or is the rest of the world perhaps stupid? From Windows 2000 and on, backwards driver backwards compatibility was built-in, and it's not unusually for Windows to select (by default) a driver for a different Windows version, before the one you would expect (signed/unsigned issues). Has Vista stopped this? I can't imagine there would be any problems with something as simple as soundcard drivers.
"In this house, we obey the laws of thermodynamics!"
You can indeed run out of wind, slowing it down to nearly nothing with enough windmills. Ditto for "a mill wheel". Each one removes energy from the water, changes the environment up and downstream. With enough of them, the water will practically come to a stop and never reach its destination.
The OP's claims are still bullshit. Using wind-power doesn't mean we need to use 100% of it... Things like dams have worked out wonderfully, and although we can't expand on them much more than we can, what they provide is great, and far better than the alternatives, NOT utilizing hydro.
Humans can sustain about 1/10th horsepower. At that, you'd have to work out for 12+ hours, non-stop, to make maybe 2 cents worth of electricity.
That's with no mention of efficiency, maintenance, infrastructure, etc.
I don't think 1/20th of a cent per hour you exercise on a stationary bicycle is going to be much motivation to people... especially when a gym membership is $20+ per month to start with.
It's probably just barely worth harnessing it when they're not paying people anything for their effort.
And how would that have been difficult/impossible with the BSD license, rather than the GPL?
If that was the case, they would have left libVorbis GPL'd, and just BSD licensed Tremor. They did not.
Many.
It seems pointless to seek ideas and feedback if you're going to ignore and delete the opinions you don't like. That's exactly what Intel and Slashdot has done with its "Opinion Center" website.
Hmmm...
I agree completely, except I'd make it a smaller value. Even $5 for things like car batteries and a couple cents a piece for aluminum cans is more than enough to have the homeless/unemployed going around, collecting them for the money. No doubt somebody could make a good business out of digging them out of landfills, and recycling them as well.
For less strict values of "satellite" yes... As in, not orbiting the earth.
Satellite:
"3. something, as a branch office or an off-campus facility of a university, that depends on, accompanies, or serves something else.
4. an attendant or follower of another person, often subservient or obsequious in manner. "
Not even slightly. FreeBSD has plenty of vnedor support as well.
Not at all. Following the strictest of definitions of the GPL, you can easily create something that is difficult or near impossible to merge back into the base code. This was a big complaint of KHTML developers with Apple.
Those who contribute back do it because they want to, NOT because they are FORCED to.
Whats more the BSD license allows MORE people and companies to use the software for more purposes... RMS understands this, and gave the okay to Xiph.org to BSD Vorbis and Theora... If you're forced to release the source, you're prevented from using it in a lot of places.
As such, I can think of many BSD (and MIT) licensed programs that have become defacto standards (eg. Telnet, FTP, SMTP, DNS, NFS, SSH, and many more) but none that were GPL licensed... How many can you think of?
Sure, you can GPL your high-tech encrypted network filesystem, and force everyone who uses it to release their code changes, but as a results, you'll just not get ANYONE using it to begin with. No matter how much better than NFS it might be, it will be soon forgotten, and you can continue to be smug about how the GPL help you stop others from stealing your code...
(Note: Use whatever license you want, but don't make moronic claims about how perfect your choice is over someone else's.)
No Fry's within 100 miles here on the west coast, either...
Yeah, it's the same everywhere... and it has been exactly that for 10+ years. Don't even get me started on their selection of monitors, and the utter shittyness of what they chose to stock.
Worst Buy does, however, always have numerous hard drives on hand, at decent prices...
Sad but true... HOWEVER, rockbox firmware supports musepack, and works on almost all models of iPod (which has, what? 80% of the market or so?).
Your experience closely mirrors my own, just a month ago. Except in my case it was an HP system, and after getting fucked-over once, I went to http://astalavista.box.sk/ and downloaded and used WGA crack v2.0. Problem solved.
If there's a technological measure preventing me from using my property, I have the absolute right to bypass it, and will proudly defend that in court if Microsoft or anyone else even considers coming after me for a "pirated" version of XP. If the next update breaks anything, you can bet I'm going to be as big a nuisance to them, as they've been to me.
The other suggestions are quite apt. Even with a dial-up or wireless connection, remotely accessing a server should be just fine, as long as your apps aren't games, high resolution videos, or something else EXTREMELY graphically intensive.
You should see if there's any other way to get your job done with better tools, that aren't such RAM hogs. The fact that notebooks with 4GBs is hard to find is a good indication that nobody else has such a problem, and there's probably a better alternative.
But, if those really aren't possibilities: I would go for a home-built lugable Opteron desktop. It's pretty easy to find low-profile components like heatsinks, PSUs, etc, and even 1/2U servers. Throw in 16GBs of DDR, and find a simple way to clip the LCD and keyboard to the unit. It probably wouldn't be any heavier than the larger "desktop-replacement" notebooks.
It surprised people... 10 years ago? When AMD released the K6? And the short period of time is EVER SINCE THEN? It's been Intel that has shortly had the better performing products in that time-period, and then, it was usually only on floating point performance, while most apps are integer-based...
Even now, I don't believe that Intel has the lead, or at least not much of one... Intel's chips usually don't have 64-bit support, and where they do, they usually have bugs that make it useless, like disabling DMA... AMD chips, however, see a big performance boost in 64-bit mode, on the order of 20%.
*Note: Not a fanboi by any stretch of the imagination. I rake either company over the coals when they do stupid shit.
Gah!
First of all, it's Matroska and the extension is MKV.
Second, whether AVI is better or worse than Ogg is debatable. Any who has ever written an Ogg (de)muxer curses it's name frequently. It's extremely codec-specific, and the format is rather loosely defined, with no consistent standard way to do much of anything....
AVI has it's limitations, but they are few. Most of the problems people experience with AVI is due entirely to limited software which doesn't properly handle AVIs. The rest of the problems tend to be a result of lack of standards... For instance, Vorbis can fit into AVI just fine, but unfortunately, Xiph didn't define HOW exactly, so everyone has started doing it in their own, mutually incompatible way. Ditto for subtitles, and other meta-data.
So, the biggest problem with AVI is lack of any single official standards authority. Ogg has the same problem, but worse, since Xiph have ignored all efforts to extend Ogg to handle other formats, and now nothing is compatible.
Vorbis is, but Musepack isn't. Quite the opposite. MPC is faster than MP3s for both encoding and decoding. It's patent-free, and sounds slightly better than Vorbis in most cases... Significantly better in certain cases.