Solving hardware problems with software is not a good practice. We can now see it with the huge load of bad drivers out there. Maybe it's time to change all that.
Having HW BIOS handle the hardware would also solve the last stability issue in a lot of OSes (Like, all of them except maybe OSX since it does support a very limitted set of HW).
Every board has a CPU (Actually, every chip is some kind of CPU), even if it is a dedicated CPU (Say, specialized in network access). But I get your point.
What you are describing is the core of the problem. Cards are now reduced to a simple chipset, leaving all the overhead of how to use this chipset to the OS. Cards (NICs for example) doesn't provide anything anymore. You could buy the chipset, a PCI board and solder the whole thing yourself in 1/2 day.
What we need is real cards, that include an interface to their freaking chipset. The OS shouldn't have to know if the chip on the NIC is a 3Com, Realtek or anything crappy like that.
regardless of if the actual driver code lies on a BIOS or is loaded from an HD
The difference would be that the BIOS driver would run on the board it's shipped in, hence it would be platform independant. A HD-based driver would still run on your main CPU and you would have to have one for every freaking architecture (PPC, 86, Sparc,...)
Dude the driver is in the BIOS of your board, whatever the board. That means the board should have the CPU needed to run it. The OS doesn't run the BIOS, the onboard CPU should.
I don't give a fuck what the hardware specs are, as long as I have an API to access it. Of course, as of today, the API is the direct addressing of the embedded APU, so I'd better know the chip.
What we need for that is a protocol to call H/W APIs. That's all, it's architecture independant. Of course, you need the right connector;-)
Exactly my point. If HW vendors did their jobs, we wouldn't be there today. Higher level BIOS drivers are a necessity today. USB "Plug And Play" is an heresy! You need the CD with the drivers, how could it possibly be "PnP"???!?!?
Point is moot: Any virus can easily destroy my windows drivers. And they happily do. As for *nix, just require to be root to flash the BIOS of any hardware device.
Flashing BIOS can be painful, granted. There might be better ways though, I'm sure they would find some.
But imagine the thing: No drivers to write for any OS!!! Wouldn't that be amazing? The manufacturer would write a driver embedded in its hardware, flashable, and tada! All OSes out there benefit from the full-fledge piece of hardware: Linux, Windows, BSD, BeOS, AmigaOS, MS-DOS 1.5 uh, no wait...
The thing is people writing a driver for Windows and people writing driver for Linux are pretty much doing the same thing. What a waste of ressources and time!
The problem is BIOS haven't evolved in the last 10 years. BIOS is supposed to be the layer between the software and the hardware. Nobosy uses it anymore and most of the drivers for most of the OS just bypas the BIOS altogether.
IMO, a big huge part of the Linux/Windows/CustomOS Kernel (name: the drivers) should be made part of the BIOS.
When you add a third-party card on your computer (say, a Radeon), it should have its own BIOS and be driven by that.
That's what a BIOS is for: provide an abstraction layer to the hardware. It is just failing at this role since a long time.
The point is pot is villianized when it shouldn't be
No, the point is you are trying to get a valid message across using an invalid argument. That rarely works unless your demo it tight proof, which it isn't.
Get your numbers straight, review the whole thing and try again. Or just drop it.
Do you think a lot of people on slashdot doesn't know that pot is less dangerous than cigarette or alcohol? Dude, the point has been made for a long time. The fact that it isn't legal is political and nothing else. Statistics won't help.
Dude, it's going to be released not before 2 years! The thing you are reviewing is in debug mode and probably has very little to do with what Longhorn will finally be. Forget what you see, this is just not Longhorn! You are reviewing a product that is nothing, will never be released close to the form you are reviewing it, so I am asking: What is the point?
how good will it look Nobody knows for one good reason: Nobody can predict the future. Let's review it when MS says it is good to be reviewed.
have a Nokia phone in front of me and guess what? It is a T-Mobile and doesn't have a SIM card (At least, not one I can easily get at)! Wow! How come??? I was a little too harsh, I have to agree. But you seem as unable to accept something you haven't seen than I do.
There were two types of SIM cards Thanks, I have already seen a SIM card in my life. It's not like I would talk about something I don't know;-)
In related news, I've blacklisted you for making repeated stupid statements without taking 10 seconds to STFG Good for you.
Ok, you win. This has to be a terminology problem Let's reformulate: There is no GSM network in the US, because according to your specs they would have to provide SIMS, and none of them does. So those AT&T phones we're talking about are not GSMs. End of argument.
Dude, all GSM phones need to identify themselves to the network. That's true, but they don't need a SIM card. They need to have this identification. In here (US) this is done most of the time by some internal ROM, most likely programmable by one mean or another.
My friends are dudes that got their cellphones at Wal-Mart. If they worked on CDMA/TDMA I wouldn't cite them as examples! BTW, do you live in the US? I don't know anyone with a SIM card out here... I was surprised myself when I came at first.
Did you see them all to write such a statement? I know for a fact that this is just plain wrong. Most of my friends that have cellphones don't have SIM cards.
So from what I can see, there is a lot of websites out there. Is this the new way of googling? Instead of spending 10 minutes on Google, just write a/. story in 2 minutes and come back the next day with answers?
there's going to be lots of things that Linus can think up to throw into the kernel which will take up half a gig of memory
Dude, we're not talking Kernel here, but Operating System. Add up all your favorite apps (read: X, KDE/Gnome, Mozilla, etc...) and tell me how much memory you have allocated.
Don't try to compare two dirrerent things. Memory ahd HDD are becoming more of a commodity these days. Let's be happy and enjoy our new features!
Obviously, the final release or even the beta releases will not consume this much of the system resources
What is the point of showing these numbers then?
In other news: Apple is working on a ultra-mini iPod. The pre-beta-alpha version we got our hands on weighted 20 pounds and was bigger than my G5. Of course, the final version will be smaller and lighter. One could still wonder where Apple is heading at with such a bulky product.
Foreword: If you have nothing relevant to say, don't say anything!
1) When all is said and done, I don't think that SCO will have really done much damage to Linux (or by extension, the GPL) Well, I hope. It sounds a little weird to me that this is their last card to play. We'll see, but they might have some more cards hidden.
This potentially hugely damaging to MS Why? Because people will start realizing that "MS is bad"? Didn't they already know about it?
Having failed to really slow Linux adoption or development As I said, just don't sell the idea before it's there. Nothing is out yet. There is still plenty of FUD going on and the Linux adoption has already been slowed down - altough not to a large extent. But this is very difficult to measure: How do you know how many people/companies were slowed down because of all the FUD?
Just tell me about one technology that hasn't been defeated yet. You'll figure out they fall into one of the following two categories:
a. It is a matter of time, but we'll get there b. Ratio complexity/interest to defeat the technolgy is too high. But this is not a technical limitation.
The specifity of software, is that the complexity can never be too high (Or at least it's never been). And the specifity of piracy is that the interest is very high. So the ratio complexity/interest is always going to be ultra-low. Hence, all software falling in this category is doomed to be cracked/worked around very soon.
One might think they got it with the CD. But no. One might think they got it with the DVD. Again, no luck. One might think they'll get it with the p2p. We'll see.
By "By definition, technology can be defeated", I mean that technology (It is more obvious in software) is created by man. With any sufficiently more evolved or parallel technology, you can totally control any other technology. Just a matter of time.
I don't think the answer is in the technology. By definition, technology can be defeated.
They just need to help the message pass through: Piracy is bad.
Now that I've helped them, I'll just close my Kazaa to show people that this can be done.... hmmm, my KazaaLite doesn't want to stop... Wow... My antivirus is going crazy... wait! piracy is bad!!! Arrrrrrrrrgh
I think you could easily show malicious intent if SCO doesn't own the linux code in question
No, that isn't enough. That would prove their claim was baseless. That wouldn't prove they knew it. You have to prove intent, which is to say: a. Their claim was baseless b. They knew it
They could always argue that they were mistaken!! Though Boies will probably be gone by then since he's paid in stocks and it's going to be worth peanuts when they will get there.
Alhtough you are obviously kidding (I hope), the main reason I still use Opera from time to time is... performances.
I've come to call Mozilla startup time it's "Boot time". Sure it doesn't take as long as my PC to get started, but it is slow as hell. Opera starts in 5-6 seconds on my bloated P3 1GHz, while Mozilla takes about 30 seconds. I don't count IE in there since it's part of the Kernel, I should add-up windows boot time and that would be somewhat unfair.
Runtime is almost as bad. When I leave Moz alone for 2 hours (Using heavy apps such as The Gimp, Java stuff, Cygwin X Server...) It takes forever to respond back (like 15 seconds). With Opera, it's just a matter of 2 seconds.
Again, takes these numbers with a grain of salt: 1GHz P3, usually 500-700MB used out of my 256MB chip. Like I said, bloated.
This is the main annoyance for me, by a large amount.
Yes it is. The public opinion would do the same to Linux that they did a decade ago (or was it more?) with BSD: Push it back to a nerdy/geeky tool. It would stop (or slow dramatically down) the broad adoption of Linux, effectively killing it.
That is because they can delay their loss on the Novell/RH/IBM cases up until it doesn't matter anymore because Linux has lost its credibility. Then they settle silently and we're all done. Linux is dead (Or in very bad shape). M$ Wins.
See, it's not just about loosing or winning the IBM case.
Slashdot Mirror
Solving hardware problems with software is not a good practice. We can now see it with the huge load of bad drivers out there. Maybe it's time to change all that.
Having HW BIOS handle the hardware would also solve the last stability issue in a lot of OSes (Like, all of them except maybe OSX since it does support a very limitted set of HW).
Every board has a CPU (Actually, every chip is some kind of CPU), even if it is a dedicated CPU (Say, specialized in network access). But I get your point.
What you are describing is the core of the problem. Cards are now reduced to a simple chipset, leaving all the overhead of how to use this chipset to the OS. Cards (NICs for example) doesn't provide anything anymore. You could buy the chipset, a PCI board and solder the whole thing yourself in 1/2 day.
What we need is real cards, that include an interface to their freaking chipset. The OS shouldn't have to know if the chip on the NIC is a 3Com, Realtek or anything crappy like that.
Oh well....
regardless of if the actual driver code lies on a BIOS or is loaded from an HD
...)
The difference would be that the BIOS driver would run on the board it's shipped in, hence it would be platform independant. A HD-based driver would still run on your main CPU and you would have to have one for every freaking architecture (PPC, 86, Sparc,
Dude the driver is in the BIOS of your board, whatever the board. That means the board should have the CPU needed to run it. The OS doesn't run the BIOS, the onboard CPU should.
;-)
I don't give a fuck what the hardware specs are, as long as I have an API to access it. Of course, as of today, the API is the direct addressing of the embedded APU, so I'd better know the chip.
What we need for that is a protocol to call H/W APIs. That's all, it's architecture independant. Of course, you need the right connector
Exactly my point. If HW vendors did their jobs, we wouldn't be there today. Higher level BIOS drivers are a necessity today. USB "Plug And Play" is an heresy! You need the CD with the drivers, how could it possibly be "PnP"???!?!?
Oh well...
Ever had a virus that hosed your bios
Point is moot: Any virus can easily destroy my windows drivers. And they happily do. As for *nix, just require to be root to flash the BIOS of any hardware device.
Flashing BIOS can be painful, granted. There might be better ways though, I'm sure they would find some.
But imagine the thing: No drivers to write for any OS!!! Wouldn't that be amazing? The manufacturer would write a driver embedded in its hardware, flashable, and tada! All OSes out there benefit from the full-fledge piece of hardware: Linux, Windows, BSD, BeOS, AmigaOS, MS-DOS 1.5 uh, no wait...
The thing is people writing a driver for Windows and people writing driver for Linux are pretty much doing the same thing. What a waste of ressources and time!
BTW, it is not extream, but extreme.
The problem is BIOS haven't evolved in the last 10 years. BIOS is supposed to be the layer between the software and the hardware. Nobosy uses it anymore and most of the drivers for most of the OS just bypas the BIOS altogether.
IMO, a big huge part of the Linux/Windows/CustomOS Kernel (name: the drivers) should be made part of the BIOS.
When you add a third-party card on your computer (say, a Radeon), it should have its own BIOS and be driven by that.
That's what a BIOS is for: provide an abstraction layer to the hardware. It is just failing at this role since a long time.
The point is pot is villianized when it shouldn't be
No, the point is you are trying to get a valid message across using an invalid argument. That rarely works unless your demo it tight proof, which it isn't.
Get your numbers straight, review the whole thing and try again. Or just drop it.
Do you think a lot of people on slashdot doesn't know that pot is less dangerous than cigarette or alcohol? Dude, the point has been made for a long time. The fact that it isn't legal is political and nothing else. Statistics won't help.
Dude, it's going to be released not before 2 years! The thing you are reviewing is in debug mode and probably has very little to do with what Longhorn will finally be. Forget what you see, this is just not Longhorn! You are reviewing a product that is nothing, will never be released close to the form you are reviewing it, so I am asking: What is the point?
how good will it look
Nobody knows for one good reason: Nobody can predict the future. Let's review it when MS says it is good to be reviewed.
have a Nokia phone in front of me and guess what? It is a T-Mobile and doesn't have a SIM card (At least, not one I can easily get at)! Wow! How come??? I was a little too harsh, I have to agree. But you seem as unable to accept something you haven't seen than I do.
;-)
There were two types of SIM cards
Thanks, I have already seen a SIM card in my life. It's not like I would talk about something I don't know
In related news, I've blacklisted you for making repeated stupid statements without taking 10 seconds to STFG
Good for you.
Ok, you win. This has to be a terminology problem Let's reformulate: There is no GSM network in the US, because according to your specs they would have to provide SIMS, and none of them does. So those AT&T phones we're talking about are not GSMs. End of argument.
Dude, all GSM phones need to identify themselves to the network. That's true, but they don't need a SIM card. They need to have this identification. In here (US) this is done most of the time by some internal ROM, most likely programmable by one mean or another.
My friends are dudes that got their cellphones at Wal-Mart. If they worked on CDMA/TDMA I wouldn't cite them as examples! BTW, do you live in the US? I don't know anyone with a SIM card out here... I was surprised myself when I came at first.
ALL GSM phones use SIM cards
Did you see them all to write such a statement? I know for a fact that this is just plain wrong. Most of my friends that have cellphones don't have SIM cards.
Informative? Bah...
So from what I can see, there is a lot of websites out there. Is this the new way of googling? Instead of spending 10 minutes on Google, just write a /. story in 2 minutes and come back the next day with answers?
there's going to be lots of things that Linus can think up to throw into the kernel which will take up half a gig of memory
Dude, we're not talking Kernel here, but Operating System. Add up all your favorite apps (read: X, KDE/Gnome, Mozilla, etc...) and tell me how much memory you have allocated.
Don't try to compare two dirrerent things. Memory ahd HDD are becoming more of a commodity these days. Let's be happy and enjoy our new features!
Obviously, the final release or even the beta releases will not consume this much of the system resources
What is the point of showing these numbers then?
In other news:
Apple is working on a ultra-mini iPod. The pre-beta-alpha version we got our hands on weighted 20 pounds and was bigger than my G5. Of course, the final version will be smaller and lighter. One could still wonder where Apple is heading at with such a bulky product.
Foreword: If you have nothing relevant to say, don't say anything!
1) When all is said and done, I don't think that SCO will have really done much damage to Linux (or by extension, the GPL)
Well, I hope. It sounds a little weird to me that this is their last card to play. We'll see, but they might have some more cards hidden.
This potentially hugely damaging to MS
Why? Because people will start realizing that "MS is bad"? Didn't they already know about it?
Having failed to really slow Linux adoption or development
As I said, just don't sell the idea before it's there. Nothing is out yet. There is still plenty of FUD going on and the Linux adoption has already been slowed down - altough not to a large extent. But this is very difficult to measure: How do you know how many people/companies were slowed down because of all the FUD?
Congratulations. You have no sense of humour.
Just tell me about one technology that hasn't been defeated yet. You'll figure out they fall into one of the following two categories:
a. It is a matter of time, but we'll get there
b. Ratio complexity/interest to defeat the technolgy is too high. But this is not a technical limitation.
The specifity of software, is that the complexity can never be too high (Or at least it's never been). And the specifity of piracy is that the interest is very high. So the ratio complexity/interest is always going to be ultra-low. Hence, all software falling in this category is doomed to be cracked/worked around very soon.
One might think they got it with the CD. But no. One might think they got it with the DVD. Again, no luck. One might think they'll get it with the p2p. We'll see.
By "By definition, technology can be defeated", I mean that technology (It is more obvious in software) is created by man. With any sufficiently more evolved or parallel technology, you can totally control any other technology. Just a matter of time.
I don't think the answer is in the technology. By definition, technology can be defeated.
They just need to help the message pass through: Piracy is bad.
Now that I've helped them, I'll just close my Kazaa to show people that this can be done.... hmmm, my KazaaLite doesn't want to stop... Wow... My antivirus is going crazy... wait! piracy is bad!!! Arrrrrrrrrgh
Foreword:
Piracy is bad. KazaaLite is worse.
I think you could easily show malicious intent if SCO doesn't own the linux code in question
No, that isn't enough. That would prove their claim was baseless. That wouldn't prove they knew it. You have to prove intent, which is to say:
a. Their claim was baseless
b. They knew it
They could always argue that they were mistaken!!
Though Boies will probably be gone by then since he's paid in stocks and it's going to be worth peanuts when they will get there.
Alhtough you are obviously kidding (I hope), the main reason I still use Opera from time to time is... performances.
I've come to call Mozilla startup time it's "Boot time". Sure it doesn't take as long as my PC to get started, but it is slow as hell. Opera starts in 5-6 seconds on my bloated P3 1GHz, while Mozilla takes about 30 seconds. I don't count IE in there since it's part of the Kernel, I should add-up windows boot time and that would be somewhat unfair.
Runtime is almost as bad. When I leave Moz alone for 2 hours (Using heavy apps such as The Gimp, Java stuff, Cygwin X Server...) It takes forever to respond back (like 15 seconds). With Opera, it's just a matter of 2 seconds.
Again, takes these numbers with a grain of salt: 1GHz P3, usually 500-700MB used out of my 256MB chip. Like I said, bloated.
This is the main annoyance for me, by a large amount.
No, I don't think so
Yes it is. The public opinion would do the same to Linux that they did a decade ago (or was it more?) with BSD: Push it back to a nerdy/geeky tool. It would stop (or slow dramatically down) the broad adoption of Linux, effectively killing it.
That is because they can delay their loss on the Novell/RH/IBM cases up until it doesn't matter anymore because Linux has lost its credibility. Then they settle silently and we're all done. Linux is dead (Or in very bad shape). M$ Wins.
See, it's not just about loosing or winning the IBM case.
each company they sue can countersue for defamation
And so? What exactly do they have to loose in the process?