I think the thing that Hurd is really lacking is a Linus or Theo-like figure with an established cult of personality.
And it looks like Neal's got some hope. He's reasonably pragmatic (important) and can use humor effectively (as in the comment that he hasn't done drugs)
Of course, the problem with Mach is that there's five people working on it, and it's not ready for a mess of people to start hacking on it. It has some interesting ideas, and some really cool directions that it could go in. Somebody's going to have to do some leadership and start setting it on a course from R&D to a real production project sometime.
I do know that the Microwave one is more along the lines where if you have interference from it, you probably should be worying about getting baked by standing in front of it instead of dropping packets.
But the last time it came up on Slashdot, people were chiming in that certain brands of phones really suck when used in conjunction with Wireless Ethernet. Which makes sense. They are using the exact same band, which will degrade the signal to varying extents.
Of course, consumer reports found that the gigahertz phones tend to perform worse than the equivelently featured 900MHz phone. But the manufacturers have their hearts set on the gigahertz phones, so they are making sure that the 900MHz ones suck.
Last I heard, the 802.11a gear had less range in paper, i.e. you can't get the 5x speed at especially good range, but the range within which you can get a decent signal in the multi-megabit range is actually at least as good as 802.11b. As in, it has more bandwidth to degrade over.
Of course, what could also happen is we see dual 802.11a/b gear. Which is fine for a number of reasons, including the multi-mile range with antenas, the lowered power consumption (If that is, in fact the case) for PDA usage, and upgradability.
I think you are going to be seeing business users going for this first, not consumers.
This way, you can have 20 people in a conference room get decent bandwidth and response time while they are all participating in a meeting/training session/etc and still leave bandwidth free for Ed, who's network port was acting up this morning.
I'm all for 54Mbps because I/can/ hose a 10MBs connection. Well, that, and I don't have the stuff purchased yet, so I'll just wait a little bit and get the 54Mbps gear when I have the cash ready. Maybe they'll have WEP fixed by then, too.;)
On the other hand, 802.11a is not that much more expensive than 802.11b.
And you can use the gigahertz cordless phones and microwaves without worying about it messing with your wireless connectivity, even though you mostly just have to wory about the gigahertz phones.
Which means that I have another few months to wait, so that I don't get the early-adopter tax on the 802.11a units, before I go out and see about picking up Wireless Ethernet kit.;)
The problem with VR is that it had a lot of things that would be cool, but never panned out.
Part of this is the model that made sense to the researchers didn't make sense to everybody else. Stock traders still don't use cool VR views of it, they stick to what makes sense to them, even though it could be done better, for example.
What you see the most applications of VR in are various forms of visuilization, a few choice applications that caught on, and games.
So what you have is consumer-grade 3D hardware for FPS games, and then the really expensive stuff for scientists.
I have been picking up pieces of office furnature that can be connected together. So I have a 60" table for $180. Get two of those and a corner piece, and it should cost you less than $1000, even if you get drawers and shelves for it, and you can put your stuff wherever you want. Works much better than the desks that are suposedly designed for computers.
And remember, $1000 in computing hardware is going to be obselete in a year or two. But $1000 in good desk hardware will last you for years and years.
I have TrippLite UPS, and it's great. Never had an APS one, so I can't comment. The only thing that APS UPS boxes may have is slightly better monitoring capabilities, depending on the model.
Well, to a certain extent, an Exokernel makes sense.
And many of the Exokernel ideas are just a slight refinement on the Partitioning idea used by large server systems. So by that token, it makes sense.
I think the bigger problem, however, is the idea in the academic community that they can build a perfect system for every application if they just abstract things enough. This drove Microkernels and this is driving Exokernels.
1) Use a LCD screen instead of a CRT
2) Prefer either laptop drives (decent) or CompactFlash. CompactFlash will require you to run Linux (is that a bad thing?;) ) unless you want to spend a LOT on storage. Laptop drives are generally rated to work in a bumpy car/truck ride, so they may work. Desktop drives just aren't up for it. CompactFlash will work until you drop the thing hard enough to break it.3) Try to get an AGP card that has a retention mechanism.4) Make sure that everything's properly screwed in.
5) Filter your airflow, so you don't need to wory about too much dust clogging your fans.
The story of the whole Parrot thing was:
1) April Fools Day hoax from the Perl and Python people creates a merged Perl / Python language called Parrot.
2) Perl 6 wanted to abstract the interpreter for bytecode from the rest of the system.
3) Perl 6 decided to name the bytecode interpreter "Parrot", in reference to the afformentioned joke.
But you do need a pretty recent one, because GeForce 1 and 2 cards I know have had problems with it, even though they do 1920x1080x60Hz non-interlaced.
It was going to be 1050 with slightly non-square pixels (i.e. 1920x1050) but they wised up.
And the frame rate is 24,25,29.97, or 30 progressive frames per second, depending on the source material, and twice those numbers for interlaced frames per second. Which means it will actually be able to do movies at the right frame rate so that it will look better.
You aren't going to see anything really taking advantage of the quality of HDTV for another few years. But when they start to show movies at the form factor the producer intended, it'll be great.;)
Analog HDTV is dissapearing. They had it in Japan, but not anywhere else, really.
There are two standards for HDTV. Same as analog TV, actually. There's the standard for the US and there's the standard for the rest of the world. The US standard is called ATSC. HDTV is the generic term for any higher resolution TV standard.
I've seen it tried and it's more of a pain in the rear than it's worth. As a matter of fact, I haven't seen anybody get real 1920x1080 on a consumer HTDV set, although the gas plasma ones might be a different case.
1920x1080 needs to be interlaced at 60 Hz. Your average video card doesn't do that well.
Sure, Loki's got the right idea, but it's better than booting over into windows if you are a hardcore linux junkie. Transgaming has the potential to be just like booting over to windows, without the whole booting over nonsense.
The problem is that most people who are hardcore linux junkies are also used to running multiple partitions for other reasons, which is why Loki's having problems.
Well, for one, the linkers ARE build on the old procedural model. I think the major reason for this is because the current model is language-neutral. You don't have to exactly pair the linker with the compiled language.
And that wouldn't really help, because different C++ compilers do things a little differently. Differences in RTTI, exception handling, memory management, etc. come in to play.
Which is why a system like CORBA, COM, or the like is so useful.
Except that all of those systems are slightly annoying on some level, so people want to avoid using them. So I'm not sure what the best answer to that is.
A better project for them would have been to write a new BeOS like API to replace X and wrap over Linux.
I have a little concern that BeOS and AtheOS are both going to suffer the same problem. By not using CORBA or COM, they gain in simplicity, but C++ linkages are just not stable, especially with respect to future expandability and changing compilers.
It doesn't work too well, especially if you are developing software for the PC, to lock down developers.
Most places allow developers to disable the lock-down, and generally won't debug your machine if you screw it up, just rebuild it from scratch. Even the big guys do that.
With respect to the Hammer, it still works in a SMP-model-like, not a NUMA model. It's not entirely SMP, but enough like SMP to make the optimizations not as hard.
Between this and Hyperthreads, new OS designs should be able to take advantage of at least multiple processors, even on the desktop. Of course, the Pentium was supposed to be the first CPU to enable SMP For The Rest Of Us, so we'll just have to see what happens.
It's likely that the K9 (I hope they have a dog or dental-related sort of codename for it.;) ) will be dependent on how the Itanium and Hammer do on the market.
If the Hammer cleans up, the K9 will build on it, leaving any possibility for a whole new platform to the K10. If the Itanic architecture starts to gain speed, the K9 will probably be an IA64 machine.
I think the key thing is that the instruction sets are mattering less. You can put optimizers in hardware that convert a messy x86 architecture to a nice RISC one. Think of the x86 arhictecture as a compression format for nice RISC opcodes. Or you can do various kinds of software morphing, which are getting more advanced as time goes on. The only real advantage to the IA64 is that it has the likelyhood of allowing the compiler to make better optimizations that will leaverage the processor more.
With regards to locking a processor into a particular memory architecture, that shouldn't be a huge issue. For one, most processor architectures stay with the same memory architecture in the chipsets for a useful span of time. So a non-issue that way, IMHO.
Now, about changing CPUs and getting a better memory architecture, that's not extremely likely. A newer memory architecture will probably have different shielding/terminating/etc. requirements. The l33t motherboard manufacturers will probably make their boards have enough headroom that it might be able to take the new memory architectures.
But that's virtually impossible to work. If it works on my buddy's p1mp ASUS motherboard, so if I have a cheappie bargan-basement motherboard, I'll expect it to work. Except that the cheappie motherboard wasn't designed with headroom.
AMD nets one happy customer and one very pissed off customer. So they will probably change things or put configuration pins in there so that the first crop of DDR333 motherboards will do a maximum of DDR333, no matter what.
Plus, most rational people upgrade processor and motherboard at the same time anyways.
So it's probably a non-issue. I personally think the integrated northbridge has been a good idea for a while. I want a 4 or 8 CPU Hammer.;)
Slashdot Mirror
True.
I think the thing that Hurd is really lacking is a Linus or Theo-like figure with an established cult of personality.
And it looks like Neal's got some hope. He's reasonably pragmatic (important) and can use humor effectively (as in the comment that he hasn't done drugs)
Of course, the problem with Mach is that there's five people working on it, and it's not ready for a mess of people to start hacking on it. It has some interesting ideas, and some really cool directions that it could go in. Somebody's going to have to do some leadership and start setting it on a course from R&D to a real production project sometime.
I do know that the Microwave one is more along the lines where if you have interference from it, you probably should be worying about getting baked by standing in front of it instead of dropping packets.
But the last time it came up on Slashdot, people were chiming in that certain brands of phones really suck when used in conjunction with Wireless Ethernet. Which makes sense. They are using the exact same band, which will degrade the signal to varying extents.
Of course, consumer reports found that the gigahertz phones tend to perform worse than the equivelently featured 900MHz phone. But the manufacturers have their hearts set on the gigahertz phones, so they are making sure that the 900MHz ones suck.
Last I heard, the 802.11a gear had less range in paper, i.e. you can't get the 5x speed at especially good range, but the range within which you can get a decent signal in the multi-megabit range is actually at least as good as 802.11b. As in, it has more bandwidth to degrade over.
Of course, what could also happen is we see dual 802.11a/b gear. Which is fine for a number of reasons, including the multi-mile range with antenas, the lowered power consumption (If that is, in fact the case) for PDA usage, and upgradability.
I think you are going to be seeing business users going for this first, not consumers.
/can/ hose a 10MBs connection. Well, that, and I don't have the stuff purchased yet, so I'll just wait a little bit and get the 54Mbps gear when I have the cash ready. Maybe they'll have WEP fixed by then, too. ;)
This way, you can have 20 people in a conference room get decent bandwidth and response time while they are all participating in a meeting/training session/etc and still leave bandwidth free for Ed, who's network port was acting up this morning.
I'm all for 54Mbps because I
On the other hand, 802.11a is not that much more expensive than 802.11b.
;)
And you can use the gigahertz cordless phones and microwaves without worying about it messing with your wireless connectivity, even though you mostly just have to wory about the gigahertz phones.
Which means that I have another few months to wait, so that I don't get the early-adopter tax on the 802.11a units, before I go out and see about picking up Wireless Ethernet kit.
You really didn't specify what you were trying to do as far as audio codecs go..
l
a ce .htmlhttp://www.firstpr.com.au/audiocomp/lossless/ http://www.cs.tut.fi/~albert/Dev/pucrunch/
But here's some nice compression resources. Boning up on that should help you write any sort of codecs..
http://www.dogma.net/markn/
http://www.compressconsult.com/
http://geocities.com/eri32/
http://www.internz.com/compression-pointers.htm
http://www.cs.mu.oz.au/~alistair/abstracts/inpl
http://www.cs.auckland.ac.nz/~peter-f/
http://www.arturocampos.com/
Hope that some or all of these URLs are useful to some area of what you want to do. Have fun.
The problem with VR is that it had a lot of things that would be cool, but never panned out.
Part of this is the model that made sense to the researchers didn't make sense to everybody else. Stock traders still don't use cool VR views of it, they stick to what makes sense to them, even though it could be done better, for example.
What you see the most applications of VR in are various forms of visuilization, a few choice applications that caught on, and games.
So what you have is consumer-grade 3D hardware for FPS games, and then the really expensive stuff for scientists.
I have been picking up pieces of office furnature that can be connected together. So I have a 60" table for $180. Get two of those and a corner piece, and it should cost you less than $1000, even if you get drawers and shelves for it, and you can put your stuff wherever you want. Works much better than the desks that are suposedly designed for computers.
And remember, $1000 in computing hardware is going to be obselete in a year or two. But $1000 in good desk hardware will last you for years and years.
I have TrippLite UPS, and it's great. Never had an APS one, so I can't comment. The only thing that APS UPS boxes may have is slightly better monitoring capabilities, depending on the model.
Well, to a certain extent, an Exokernel makes sense.
And many of the Exokernel ideas are just a slight refinement on the Partitioning idea used by large server systems. So by that token, it makes sense.
I think the bigger problem, however, is the idea in the academic community that they can build a perfect system for every application if they just abstract things enough. This drove Microkernels and this is driving Exokernels.
1) Use a LCD screen instead of a CRT ;) ) unless you want to spend a LOT on storage. Laptop drives are generally rated to work in a bumpy car/truck ride, so they may work. Desktop drives just aren't up for it. CompactFlash will work until you drop the thing hard enough to break it.3) Try to get an AGP card that has a retention mechanism.4) Make sure that everything's properly screwed in.
2) Prefer either laptop drives (decent) or CompactFlash. CompactFlash will require you to run Linux (is that a bad thing?
5) Filter your airflow, so you don't need to wory about too much dust clogging your fans.
The story of the whole Parrot thing was:
1) April Fools Day hoax from the Perl and Python people creates a merged Perl / Python language called Parrot.
2) Perl 6 wanted to abstract the interpreter for bytecode from the rest of the system.
3) Perl 6 decided to name the bytecode interpreter "Parrot", in reference to the afformentioned joke.
Ahh.. Then I stand corrected.
But you do need a pretty recent one, because GeForce 1 and 2 cards I know have had problems with it, even though they do 1920x1080x60Hz non-interlaced.
HDTV is 1920x1080 at a few different frame rates.
;)
It was going to be 1050 with slightly non-square pixels (i.e. 1920x1050) but they wised up.
And the frame rate is 24,25,29.97, or 30 progressive frames per second, depending on the source material, and twice those numbers for interlaced frames per second. Which means it will actually be able to do movies at the right frame rate so that it will look better.
You aren't going to see anything really taking advantage of the quality of HDTV for another few years. But when they start to show movies at the form factor the producer intended, it'll be great.
Well..
Analog HDTV is dissapearing. They had it in Japan, but not anywhere else, really.
There are two standards for HDTV. Same as analog TV, actually. There's the standard for the US and there's the standard for the rest of the world. The US standard is called ATSC. HDTV is the generic term for any higher resolution TV standard.
I've seen it tried and it's more of a pain in the rear than it's worth. As a matter of fact, I haven't seen anybody get real 1920x1080 on a consumer HTDV set, although the gas plasma ones might be a different case.
1920x1080 needs to be interlaced at 60 Hz. Your average video card doesn't do that well.
Yes, but the Hauppauge card doesn't decode the full HDTV stream. They just convert it to NTSC and let you see that.
He's only half right.
Sure, Loki's got the right idea, but it's better than booting over into windows if you are a hardcore linux junkie. Transgaming has the potential to be just like booting over to windows, without the whole booting over nonsense.
The problem is that most people who are hardcore linux junkies are also used to running multiple partitions for other reasons, which is why Loki's having problems.
Well, for one, the linkers ARE build on the old procedural model. I think the major reason for this is because the current model is language-neutral. You don't have to exactly pair the linker with the compiled language.
And that wouldn't really help, because different C++ compilers do things a little differently. Differences in RTTI, exception handling, memory management, etc. come in to play.
Which is why a system like CORBA, COM, or the like is so useful.
Except that all of those systems are slightly annoying on some level, so people want to avoid using them. So I'm not sure what the best answer to that is.
I dono why they are bothering with this.
A better project for them would have been to write a new BeOS like API to replace X and wrap over Linux.
I have a little concern that BeOS and AtheOS are both going to suffer the same problem. By not using CORBA or COM, they gain in simplicity, but C++ linkages are just not stable, especially with respect to future expandability and changing compilers.
It doesn't work too well, especially if you are developing software for the PC, to lock down developers.
Most places allow developers to disable the lock-down, and generally won't debug your machine if you screw it up, just rebuild it from scratch. Even the big guys do that.
Ever drop a tape while taking it out of the bay and stuffing it into the tape store? I have. The tape was fine.
Even rugedized drives, when dropped from arm's length, are not going to hold up too well. Cheap drives will definately not hold up.
The deck is stacked against windows.
It's a large-scale Sun or the like server with "Smart Terminals" a.k.a. Dick..err.. diskless workstations a.k.a. X-terminals vs. a PC network.
I would like to see a comparison in there that also includes Linux workstations and either Unix or Linux servers.
With respect to the Hammer, it still works in a SMP-model-like, not a NUMA model. It's not entirely SMP, but enough like SMP to make the optimizations not as hard.
Between this and Hyperthreads, new OS designs should be able to take advantage of at least multiple processors, even on the desktop. Of course, the Pentium was supposed to be the first CPU to enable SMP For The Rest Of Us, so we'll just have to see what happens.
It's likely that the K9 (I hope they have a dog or dental-related sort of codename for it. ;) ) will be dependent on how the Itanium and Hammer do on the market.
If the Hammer cleans up, the K9 will build on it, leaving any possibility for a whole new platform to the K10. If the Itanic architecture starts to gain speed, the K9 will probably be an IA64 machine.
I think the key thing is that the instruction sets are mattering less. You can put optimizers in hardware that convert a messy x86 architecture to a nice RISC one. Think of the x86 arhictecture as a compression format for nice RISC opcodes. Or you can do various kinds of software morphing, which are getting more advanced as time goes on. The only real advantage to the IA64 is that it has the likelyhood of allowing the compiler to make better optimizations that will leaverage the processor more.
I have considered this myself.
;)
With regards to locking a processor into a particular memory architecture, that shouldn't be a huge issue. For one, most processor architectures stay with the same memory architecture in the chipsets for a useful span of time. So a non-issue that way, IMHO.
Now, about changing CPUs and getting a better memory architecture, that's not extremely likely. A newer memory architecture will probably have different shielding/terminating/etc. requirements. The l33t motherboard manufacturers will probably make their boards have enough headroom that it might be able to take the new memory architectures.
But that's virtually impossible to work. If it works on my buddy's p1mp ASUS motherboard, so if I have a cheappie bargan-basement motherboard, I'll expect it to work. Except that the cheappie motherboard wasn't designed with headroom.
AMD nets one happy customer and one very pissed off customer. So they will probably change things or put configuration pins in there so that the first crop of DDR333 motherboards will do a maximum of DDR333, no matter what.
Plus, most rational people upgrade processor and motherboard at the same time anyways.
So it's probably a non-issue. I personally think the integrated northbridge has been a good idea for a while. I want a 4 or 8 CPU Hammer.