The compiler can know pretty much. A compiler can be updated and modified much easier than a hardware optimiser. And a ridiculous proportion of the typical x86 chip is doing optimisation rather than actual processing.
x86-64 is not better, it's more kludgy and horrible than even x86. There is at least 3 levels of emulation going on there, it might seem nicer from a high level perspective but working with it it's really horrible. Yes, the lack of a good compiler for a while is what killed itanium, but it really didn't deserve to die, especially now icc is working and working well.
Compiler development can lag the CPU, but it shouldn't. The compiler knows how to optimize for the CPU better than the CPU, because the compiler is dedicated to optimizing and designed to be, wheras CPUs were originally meant to ocassionally find time to, you know, actually do the processing. Assessing run-time behaviour is overrated. The CPU is deterministic, anything you can do at run-time you can do at compile-time - just add a detection routine if you really need to, most of the time you won't. Certainly there is no way run-time optimisation can be faster than compile-time, because you can just compile to the same opcodes that would have been executed when run-time optimizing in the case that that's the fastest way to do things. But most of the time it won't be.
...who knows what he does not know. A manager who doesn't know tech is fine, provided he understands this and is willing to trust the technical people. As long has he understands that I am the specialist, so will accept when I say something is impossible or I know a better way to do things, there's no problem. It's when a non-technical manager dictates on technical matters that problems start.
Because dolts like it. They think their new super-leet bios is cool. Thus, at relatively little cost to the makers, it increases sales. It's useless, but it's a uselessness that sells.
And the answer to that is that some people still use DOS. To some people, DOS is business-critical. If bios makers drop the "cruft", they will lose sales to these people, wheras they gain little (nothing?) from dropping it.
Legacy compatibility. If I can't boot OS/2 on your fancy BIOS, I don't want it. Granted, the torturous current boot process is a horror. But I don't think there are any elements which can be eliminated without disabling something which, to someone, somewhere, is vital.
I think OpenFirmware is a good enough compromise. As you said, a good maximalist approach is very hard - too hard. But OpenFirmware will probably be maximalist enough for DOS, maximalist enough for hobbyist OSes to get working until they're mature enough to write full drivers for things. It's certainly better than the current x86 bios.
The GPL requires the source to a to be available if it is legally a derivative work, so it depends how closely 'a' is linked to 'b' and 'c'. If they just talk through pipes you're fine, if they dlopen each other it's a bit unclear, if they're linked in at compile time you probably need to release source for 'a'. But if you wrote 'b' and 'c', then you can release 'abc' under any license you like - you don't have to give up any rights by releasing 'b' under gpl.
Also note that many, perhaps most, libraries are lgpl for this very reason. If 'b' and 'c' are lgpl libraries, then when you release 'abc' you have to include the source for 'b' and 'c' but not for 'a'. This is usually done for libraries which duplicate existing commercial alternatives - it's only the really innovative ones where you'll find libraries being gpled without an lgpl or bsd alternative.
I'm fed up of seeing this nonsense argument getting modded up. It's not us who are going to pay it, directly or otherwise. If MS gets fined, it is their shareholders who will lose money. No-one else. Just think about it for a minute. MS is already charging as much as they think they can get away with. Their prices are optimised - if they charge more, there'll be enough fewer people buying their product that it isn't worth it. So just how would we pay?
Fining companies punishes the company, in terms of its shareholders who are its owners. If they are fined enough they go bankrupt. It's nonsense to claim fining them is pointless.
That's not what we need to do, you can still get sued. What we need to do is make it easy for innocent people to get acquitted without losing huge amounts of money.
I take it you'd prefer to have your file manager start spewing windows everywhere, your dialogs backwards, and NO WAY TO TURN IT OFF?
Seriously, the "offer an option" approach is one of my favourite things about kde. There will be times when icon caching makes sense and there will be times when it doesn't, there will be users who want it and users who don't, so let the user choose!
Are you sure that's not your distro moving your config files? Many will use a different config for each new version, because sometimes KDE gets confused when upgrading and they reckon it's better to just start afresh. The control centre still has those two different forms (module at a time or whole thing, I presume you mean), and having it load the "whole thing" off the panel to be consistent with the menu was a change - would you have preferred that to stay the same?
I haven't noticed such changes. 3.1 and 3.3 seem to behave pretty much the same way, and where it's changed, it's usually the case that the new behaviour is what I thought it "should" be. KDE has certainly never done anything like forcing the file browser into "spatial" mode or reversing the order of buttons in the confirmation dialog, and every kde change has been reversable by hand with ordinary configuration changes.
Have you looked at Linux recently? If you can afford to, get one of those cheap HP towers with Mandrake preinstalled. It "just work"s fine, really. Throw a random "grandmother" at it and it's as easy to use as windows, any day.
Looking at what other people have experienced I seem to have been insanely lucky, but the fact is, Linux supported my hardware - random hardware I'd bought before I even heard of Linux - better than windows. Graphics card - working a little by default, well after downloading the drivers from the nvidia website, on both OSes. Printer - on windows, worked fine after installing drivers from cd. On linux, mostly worked (ink level monitoring a bit fiddly was all) out of the box, perfect after downloading drivers from the epson website. Sound (onboard via82c686b) was an absolute nightmare in windows. Have to install from a driver CD, but 4 out of five times (that I tried, I stopped after the fifth which was the only success) the driver installer says "you don't have this type of card". Then, it will refuse to run again until a reboot. And when I reboot (2 minutes each time), I get a popup message saying "congratulations on successfully installing the driver", even though it didn't install, and I can't run the installer until I've rebooted again. Arrgh. Under linux, and not even a friendly linux, slackware, it's set up on install without me doing anything at all. I just have to turn up the volume and it works. Hard drives - I had two drives thinking they were master on the same channel. This caused windows to randomly swap around the two drives on the other channel, and treat one as a "removeable device". Linux had no difficulty at all. Granted this is a hardware problem, but Linux handled it much better.
Maybe I'm just lucky. But on my system, on my hardware, Linux handled it much better.
Well, anecdotally, I use linux at home maybe 4 hours a day, and windows at school for maybe 2 (probably slightly less). Since September I have seen around the same number (3) of crashes on both. But, crucially, I can reproduce all of the Linux ones every time. And two were due to binary-only software, while the other was a beta for which I filed a bugreport and it was fixed in the next release. Wheras when windows goes down, it does so randomly, with no indication of what caused it, and no way to fix it.
Binary-only code. That's the only way I know to make Linux unstable. Unreal Tournament for me and the flash player for you.
Linux still crashes, but it doesn't do so randomly. I've had 3 crash conditions that didn't involve UT. All of them I can reproduce every single time, and thus I can avoid them.
Is a felon that has served its sentence entitled to the same rights as others?
Yes. 100%. They are citizens and they have served their sentence. That is it. If we think they should be punished more, a longer sentence would be the place to do that. It's not fair to punish someone for their entire life for a mistake when they were 20.
He didn't say hatred of Christ, just of Christians. I take Ghandi's view - Christ was a good man, and if Christians were like him I would respect them. But the fact is the average Christian behaves less "Christainly" than the average man in the street. I have a dislike of Christians that has no basis in my disagreement with their faith - it's based entirely on the way they act. Yes, there are exceptions, but on the whole Christians are not nice people.
Slashdot Mirror
The compiler can know pretty much. A compiler can be updated and modified much easier than a hardware optimiser. And a ridiculous proportion of the typical x86 chip is doing optimisation rather than actual processing.
x86-64 is not better, it's more kludgy and horrible than even x86. There is at least 3 levels of emulation going on there, it might seem nicer from a high level perspective but working with it it's really horrible. Yes, the lack of a good compiler for a while is what killed itanium, but it really didn't deserve to die, especially now icc is working and working well.
Compiler development can lag the CPU, but it shouldn't. The compiler knows how to optimize for the CPU better than the CPU, because the compiler is dedicated to optimizing and designed to be, wheras CPUs were originally meant to ocassionally find time to, you know, actually do the processing. Assessing run-time behaviour is overrated. The CPU is deterministic, anything you can do at run-time you can do at compile-time - just add a detection routine if you really need to, most of the time you won't. Certainly there is no way run-time optimisation can be faster than compile-time, because you can just compile to the same opcodes that would have been executed when run-time optimizing in the case that that's the fastest way to do things. But most of the time it won't be.
...who knows what he does not know. A manager who doesn't know tech is fine, provided he understands this and is willing to trust the technical people. As long has he understands that I am the specialist, so will accept when I say something is impossible or I know a better way to do things, there's no problem. It's when a non-technical manager dictates on technical matters that problems start.
Boot freedos. It's still free, and rebooting to do something so low-level isn't unreasonable.
And the answer to that is that some people still use DOS. To some people, DOS is business-critical. If bios makers drop the "cruft", they will lose sales to these people, wheras they gain little (nothing?) from dropping it.
Legacy compatibility. If I can't boot OS/2 on your fancy BIOS, I don't want it. Granted, the torturous current boot process is a horror. But I don't think there are any elements which can be eliminated without disabling something which, to someone, somewhere, is vital.
I think OpenFirmware is a good enough compromise. As you said, a good maximalist approach is very hard - too hard. But OpenFirmware will probably be maximalist enough for DOS, maximalist enough for hobbyist OSes to get working until they're mature enough to write full drivers for things. It's certainly better than the current x86 bios.
Also note that many, perhaps most, libraries are lgpl for this very reason. If 'b' and 'c' are lgpl libraries, then when you release 'abc' you have to include the source for 'b' and 'c' but not for 'a'. This is usually done for libraries which duplicate existing commercial alternatives - it's only the really innovative ones where you'll find libraries being gpled without an lgpl or bsd alternative.
Fining companies punishes the company, in terms of its shareholders who are its owners. If they are fined enough they go bankrupt. It's nonsense to claim fining them is pointless.
I seriously doubt that, why would slackware optimise for 486 if 386 ran better on modern machines?
That's not what we need to do, you can still get sued. What we need to do is make it easy for innocent people to get acquitted without losing huge amounts of money.
Seriously, the "offer an option" approach is one of my favourite things about kde. There will be times when icon caching makes sense and there will be times when it doesn't, there will be users who want it and users who don't, so let the user choose!
Burning karma here, but ffs why are you putting an apostrophe on cds?
Are you sure that's not your distro moving your config files? Many will use a different config for each new version, because sometimes KDE gets confused when upgrading and they reckon it's better to just start afresh. The control centre still has those two different forms (module at a time or whole thing, I presume you mean), and having it load the "whole thing" off the panel to be consistent with the menu was a change - would you have preferred that to stay the same?
I haven't noticed such changes. 3.1 and 3.3 seem to behave pretty much the same way, and where it's changed, it's usually the case that the new behaviour is what I thought it "should" be. KDE has certainly never done anything like forcing the file browser into "spatial" mode or reversing the order of buttons in the confirmation dialog, and every kde change has been reversable by hand with ordinary configuration changes.
Wrong. (Notice IA64 sitting quite happily in the architecture list)
It's all relative though. It would take him just as much acceleration to "hover above the ground" as to fly around the world the other way.
Can you still press ctrl-N and have it work? That's sometimes the case.
Have you looked at Linux recently? If you can afford to, get one of those cheap HP towers with Mandrake preinstalled. It "just work"s fine, really. Throw a random "grandmother" at it and it's as easy to use as windows, any day.
Maybe I'm just lucky. But on my system, on my hardware, Linux handled it much better.
Well, anecdotally, I use linux at home maybe 4 hours a day, and windows at school for maybe 2 (probably slightly less). Since September I have seen around the same number (3) of crashes on both. But, crucially, I can reproduce all of the Linux ones every time. And two were due to binary-only software, while the other was a beta for which I filed a bugreport and it was fixed in the next release. Wheras when windows goes down, it does so randomly, with no indication of what caused it, and no way to fix it.
Linux still crashes, but it doesn't do so randomly. I've had 3 crash conditions that didn't involve UT. All of them I can reproduce every single time, and thus I can avoid them.
Yes. 100%. They are citizens and they have served their sentence. That is it. If we think they should be punished more, a longer sentence would be the place to do that. It's not fair to punish someone for their entire life for a mistake when they were 20.
He didn't say hatred of Christ, just of Christians. I take Ghandi's view - Christ was a good man, and if Christians were like him I would respect them. But the fact is the average Christian behaves less "Christainly" than the average man in the street. I have a dislike of Christians that has no basis in my disagreement with their faith - it's based entirely on the way they act. Yes, there are exceptions, but on the whole Christians are not nice people.