The kernel already halts processes to free up resources for more pressing functions. Thats kind of the defenition of multitasking. What this allows is for the kernel to halt itself if something in userspace is more deserving of processor time.
I know a lot of other people have corrected you already but the one thing no one has mentioned yet is that because this version of PalmOS runs on ARM (a 32 bit processor), then the ARM version will most likely have 32 bit addressing.
Actually if I remember correctly it came from a sort of fight between the Win9x and WinNT teams. You're right that OpenGL was touted as a feature of NT. As a result Microsoft originally decided not to put it in 9x, as they believed at the time that by the time opengl was needed NT would be the standard. So the 9x team started making direct3d, targeting it at games, and because no game developer expected to be run on NT it started catching on. By the time NT had d3d and 9x had opengl (which wasn't too long afterwards) ms decided to use d3d as a tactical technology and proceeded to ram it down our throats.
No it means the preservation of loki's already open source code (SDL, OpenAL). Loki can't open their game code. They licensed the source from the original Windows versions under proprietary terms so they could port them. Unless Loki managed to convince every company they ported a game for to open the Windows versions as well then its not going to happen.
Actually the Power4 and the G4 are two completely different beasts. If the Mac were using Power4s I'd buy one in a second, but Motorola has completely screwed up the PowerPC line since diverging from IBM (the G3 is the same as the Power 750, the last time IBM and Motorola made the same chip).
And Darwin is just the kernel. That's all that's been ported to X86. The Quarts graphic engine is highly optimized for the PowerPC architechture, and only really performs decently when Altivec is present. That's one of the reasons OSX on a G3 is so dismal.
OK true windows isn't the greatest at that. At best windows has some constants defined that application developers are supposed to use. MacOS on the other hand is very different in this respect. At least in previous versions of MacOS everything regarding a keybinding is pulled from some system resource fork somewhere (not sure which one...haven't used macs in a while) and I'm pretty damn sure OSX is the same (or probably a bit more powerfull with an XML file somewhere doing that job). With linux however, yes you and I both love our emacs keys, but I can guarantee that there's someone you know lurking in the shadows that wants to push escape and use h j k and l to move around. And many KDE and GNOME apps do copy those keybindings. But staroffice isn't a KDE or GNOME app, it uses its own widget set. Netscape 4.x isn't (yes I know you should be using mozilla but the point is still valid), its motif based (as are a lot of older apps that people rely on). Xpdf is an Xaw3d app, so not only does it have realy spartan keybindings but its scrollbars don't even work in remotely the same way as modern GUI users have come to expect.
And keybindings aren't even the point, they're just an example. A gtk 1.0 app might look like a gnome app, but when the user realizes that in gnome he can click one menu to drop it down and then move the mouse over another menu to have it open and the old one close automatically, he'll expect the gtk 1.0 app to work the same way. But it won't. And don't say that he should have all his apps up to date because that's not a responsibility that should be placed on the user just to have a functioning environment. If anything that shouldn't be a responsibilty simply because it wasn't with windows. When a user upgraded from win3.1 to win95 they got those nice new mouse following windows on every application, not just new win95 designed apps. Hopefully we'll have this with api's being a little more standardized in gnome2.0 and KDE 3.0. But even then you have a user who's only used gnome accidentaly install a kde program (or an Xaw or strait Xlib program) and not understand why it doesn't look and act like his others. Knowledge of which api a programmer used should not be a prerequisit for using a program!
And again, I am a linux user and I do appreciate the choice between KDE and GNOME and others. I however would really like it if at least someone came up with a common display library that could be wrapped by qt and gtk, just providing them with the same look and feel regardless of whether the developer liked DCOP or bonobo better. We have made some progress in this respect. We've got DnD interoperability. We have KDE loading pixmap based gnome themes. We have Gnome making motif apps adopt the same color schemes and at this point most window managers support both environments. We still have a ton of apps out there that just don't fit with anything (much as I love it xemacs is one of these), but those are getting replaced or updated over time. Until then I have to make excuses for things when I show people my linux machines.
It means that the OS or the Toolkit or the window manager or desktop environment doesn't force this on the applications. In MacOS (classic and OSX) and windows the api you pretty much have to use makes the programmer get keyboard shortcut bindings from some centralized part of the system. Linux on the other hand doesn't exactly have that. Just now are environments like KDE and GNOME getting capabilities like that. But, because there's no one forcing anyone to use just one of those environments (or any at all) this consistency breaks down. Now this does have its benefits in some people's eyes in that it allows applications developers to experiment with new types of interfaces without rewriting the entire system, this is only beneficial from an experementer's point of view, not a user's. Now I tend to be a person that likes to experiment with my computer so this doesn't bother me. But the vast majority of people out there don't care.
The windows quicktime works flawlessly in wine now and Codeweavers makes a wine enabled netscape plugin for linux (basically a linux netscape plugin that's sole purpose is to run windows netscape plugins plugged into a program running under wine which is embedded by the linux netscape plugin....a pretty cool setup) who's main use right now is getting the quicktime plugin to work on linux. The codeweavers plugin is $19.95 but its well worth it (with realplayer for linux and mplayer I can see anything windows people can).
If you could be persuaded into getting another console however, Smash Brothers Melee (for the game cube....due to be released tommorow) is an exceptionally cool game. Just picture the four player chaos of the original Smash Brothers with graphics that really show off what the game cube can do and about ten times as much hidden crap to unlock. They've added a new "Adventure" mode for single player where you get to play through a prettied up version of a level from each characters original game before you actually fight them (and going through the Mushroom Kingdom as Samus (from Metroid) is a very interesting experience). Lots more characters and items and stages, and all around just a hell of a fun game.
Actually iPlanet's calendar server is pretty nice. From what I've seen (my company uses it) the server itself will do anything that Exchange does, only with a more open protocal. Unfortunately the only client right now for that protocal is the web application they bundle with it. Its that web app thats not so impressive, the underlying server is much nicer. Thing is, there has been talk on the evolution mailing list about giving evolution iPlanet support. If they ever get around to finishing this I will be so unbelievably happy. Besides the support for a real calendaring server (right now it does calendaring through email attachments) I would use evolution over outlook any day.
I hate to tell you this but SGI didn't "make the Cray". SGI bought Cray along with their product line and then proceeded to slowly kill off their product line (while gradually trying to transition their customers to Origins). Thing is conventional systems like Origins (conventional relative to Crays at least) just really can't do what a true vector supercomputer can. And anyways, making a multiprocessor system is nothing like making a vector machine.
What about instead of vibrating a mirror (which would as you said cause compression waves in the air and send pretty substantial vibrations through the machine (which could then travel to the leds throwing their focus off, but I digress). Why not have a box with mirrors on the outside and rotate it . Because you'd have four mirrors you'd only have to rotate it at 7.5 rps, you'd counteract any vibrations (assuming you balance the box right) and you'd get a smoother motion to boot (no slow down at the extremeties of the pictures, its hard to vibrate something in a manner thats not sinusoidal (though sinusoidal motion in that mirror could be counteracted by a cylindrical mirror afterwards but again I digress)).
The problem with using a row of 640 leds is that you once again have the fixed resolution problem (at least horizontally). I don't know if it would be feasable to rotate a miror box at 3925 rps (to do the horizontal scan with one led) but it would solve that problem.
IBM drives are on the whole very good. Yes the 75GXPs are crap. On the other hand the 40GXPs and 60GXPs are some of the best consumer level drives on the market right now. If you're trying to avoid a faulty drive don't go to a slightly less faulty alternative like Maxtor (I have seen too many Maxtors (of varying generations) fail, I will never trust them again) rather than one of the products that gave IBM their previous reputation for reliability.
P.S. No I don't work for or are in any way affiliated with IBM. I'm just a sysadmin thats run a lot of hardware into the ground over the years, and am a happy owner of a 60GXP.
I can at least vouch for the 60GXPs. From my own experience and everyone I've talked to about them they are one of the more reliable models IBM has come out with. I've put them in some pretty heavy use fileservers and never had any problems
>Think about that for a second. I have a *Lisp* >system that takes less disk space than 12M. That is >a huge, huge, amount of code. Sure, it is less than >the drivers, but considering that X does very >little, it is positively *enormous*.
Ok, but first of all this is stuff used by X clients, not the X server. These are things like libX11 and libXt, essential apis for X clients. Then you have the fact that alot of these apis have been pretty much depreciated for modern X programs. Gnome doesn't use Xintrinsics or the athena widgets. If you're running mostly gnome programs libXt and libXaw almost never loaded. Then there are things sitting around in there like libPEX. Don't even try to say you've used a PEX based program in the past 4 years. All in all, you're probably only going to use libX11, libICE, libSM, libXm and if you're using pretty antialiased fonts libXft and libXrender. Thats about four megs.
Is it just me or does this sound strangely similar to Plan 9.
I just find it funny that not only is Microsoft Research doing something that they think is new but was really done by bell labs back in the 80s, but back in the 80s every one looked at it and said "who cares".
Yes, he was forced to trademark it a while back because the publisher to some linux books tried to trademark it and then sue all the distribution makers for usage rights. Linus proved prior usage and trademarked it (with explicit permission given to anyone for free use) to keep someone else from doing that.
No he wasn't talking about removing usefull stuff. Specifically, he's talking about removing anything not usefull. All the driver version and "written by Joe Shmoe 2001" messages. Also the "blah.c driver loaded OK" are being considered for removal because, really, they serve no purpose. Error messages and actually usefull status updates (like the ide driver spitting out the device ids of the drives attached) will be left in. And for all those of you who really value the ego boost of having your name in the kernel boot, he also specifically said to leave this stuff in module load outputs. Just get it out of compiled in drivers.
They DIDN'T use a 2.4 test kernel. They used 2.2.16. They did however make sure all the libraries and support programs are up to date enough that a 2.4 kernel will drop in with an rpm.
Ok, even if he was trying to emulate MacOS and not the PowerPC archetecture in general, and even if it was MacOS X and not Darwin (the kernel...no more) that booted on x86 you'd still have the problem of getting a PowerPC binary to run on Intel hardware. Which is what he did.
But MS can't kill this by not running on it. They already do run on it. Its completely IA32 compatible. It just has the capability to do more with some 64bit instructions. If this chip just catches on thanks to the fact that its all around fast and people realize "hey....on linux its even better" then things could happen.
Slashdot Mirror
The kernel already halts processes to free up resources for more pressing functions. Thats kind of the defenition of multitasking. What this allows is for the kernel to halt itself if something in userspace is more deserving of processor time.
I know a lot of other people have corrected you already but the one thing no one has mentioned yet is that because this version of PalmOS runs on ARM (a 32 bit processor), then the ARM version will most likely have 32 bit addressing.
Actually if I remember correctly it came from a sort of fight between the Win9x and WinNT teams. You're right that OpenGL was touted as a feature of NT. As a result Microsoft originally decided not to put it in 9x, as they believed at the time that by the time opengl was needed NT would be the standard. So the 9x team started making direct3d, targeting it at games, and because no game developer expected to be run on NT it started catching on. By the time NT had d3d and 9x had opengl (which wasn't too long afterwards) ms decided to use d3d as a tactical technology and proceeded to ram it down our throats.
No it means the preservation of loki's already open source code (SDL, OpenAL). Loki can't open their game code. They licensed the source from the original Windows versions under proprietary terms so they could port them. Unless Loki managed to convince every company they ported a game for to open the Windows versions as well then its not going to happen.
Actually the Power4 and the G4 are two completely different beasts. If the Mac were using Power4s I'd buy one in a second, but Motorola has completely screwed up the PowerPC line since diverging from IBM (the G3 is the same as the Power 750, the last time IBM and Motorola made the same chip).
And Darwin is just the kernel. That's all that's been ported to X86. The Quarts graphic engine is highly optimized for the PowerPC architechture, and only really performs decently when Altivec is present. That's one of the reasons OSX on a G3 is so dismal.
OK true windows isn't the greatest at that. At best windows has some constants defined that application developers are supposed to use. MacOS on the other hand is very different in this respect. At least in previous versions of MacOS everything regarding a keybinding is pulled from some system resource fork somewhere (not sure which one...haven't used macs in a while) and I'm pretty damn sure OSX is the same (or probably a bit more powerfull with an XML file somewhere doing that job). With linux however, yes you and I both love our emacs keys, but I can guarantee that there's someone you know lurking in the shadows that wants to push escape and use h j k and l to move around. And many KDE and GNOME apps do copy those keybindings. But staroffice isn't a KDE or GNOME app, it uses its own widget set. Netscape 4.x isn't (yes I know you should be using mozilla but the point is still valid), its motif based (as are a lot of older apps that people rely on). Xpdf is an Xaw3d app, so not only does it have realy spartan keybindings but its scrollbars don't even work in remotely the same way as modern GUI users have come to expect.
And keybindings aren't even the point, they're just an example. A gtk 1.0 app might look like a gnome app, but when the user realizes that in gnome he can click one menu to drop it down and then move the mouse over another menu to have it open and the old one close automatically, he'll expect the gtk 1.0 app to work the same way. But it won't. And don't say that he should have all his apps up to date because that's not a responsibility that should be placed on the user just to have a functioning environment. If anything that shouldn't be a responsibilty simply because it wasn't with windows. When a user upgraded from win3.1 to win95 they got those nice new mouse following windows on every application, not just new win95 designed apps. Hopefully we'll have this with api's being a little more standardized in gnome2.0 and KDE 3.0. But even then you have a user who's only used gnome accidentaly install a kde program (or an Xaw or strait Xlib program) and not understand why it doesn't look and act like his others. Knowledge of which api a programmer used should not be a prerequisit for using a program!
And again, I am a linux user and I do appreciate the choice between KDE and GNOME and others. I however would really like it if at least someone came up with a common display library that could be wrapped by qt and gtk, just providing them with the same look and feel regardless of whether the developer liked DCOP or bonobo better. We have made some progress in this respect. We've got DnD interoperability. We have KDE loading pixmap based gnome themes. We have Gnome making motif apps adopt the same color schemes and at this point most window managers support both environments. We still have a ton of apps out there that just don't fit with anything (much as I love it xemacs is one of these), but those are getting replaced or updated over time. Until then I have to make excuses for things when I show people my linux machines.
It means that the OS or the Toolkit or the window manager or desktop environment doesn't force this on the applications. In MacOS (classic and OSX) and windows the api you pretty much have to use makes the programmer get keyboard shortcut bindings from some centralized part of the system. Linux on the other hand doesn't exactly have that. Just now are environments like KDE and GNOME getting capabilities like that. But, because there's no one forcing anyone to use just one of those environments (or any at all) this consistency breaks down. Now this does have its benefits in some people's eyes in that it allows applications developers to experiment with new types of interfaces without rewriting the entire system, this is only beneficial from an experementer's point of view, not a user's. Now I tend to be a person that likes to experiment with my computer so this doesn't bother me. But the vast majority of people out there don't care.
The windows quicktime works flawlessly in wine now and Codeweavers makes a wine enabled netscape plugin for linux (basically a linux netscape plugin that's sole purpose is to run windows netscape plugins plugged into a program running under wine which is embedded by the linux netscape plugin....a pretty cool setup) who's main use right now is getting the quicktime plugin to work on linux. The codeweavers plugin is $19.95 but its well worth it (with realplayer for linux and mplayer I can see anything windows people can).
If you could be persuaded into getting another console however, Smash Brothers Melee (for the game cube....due to be released tommorow) is an exceptionally cool game. Just picture the four player chaos of the original Smash Brothers with graphics that really show off what the game cube can do and about ten times as much hidden crap to unlock. They've added a new "Adventure" mode for single player where you get to play through a prettied up version of a level from each characters original game before you actually fight them (and going through the Mushroom Kingdom as Samus (from Metroid) is a very interesting experience). Lots more characters and items and stages, and all around just a hell of a fun game.
I got better......
Actually iPlanet's calendar server is pretty nice. From what I've seen (my company uses it) the server itself will do anything that Exchange does, only with a more open protocal. Unfortunately the only client right now for that protocal is the web application they bundle with it. Its that web app thats not so impressive, the underlying server is much nicer. Thing is, there has been talk on the evolution mailing list about giving evolution iPlanet support. If they ever get around to finishing this I will be so unbelievably happy. Besides the support for a real calendaring server (right now it does calendaring through email attachments) I would use evolution over outlook any day.
I hate to tell you this but SGI didn't "make the Cray". SGI bought Cray along with their product line and then proceeded to slowly kill off their product line (while gradually trying to transition their customers to Origins). Thing is conventional systems like Origins (conventional relative to Crays at least) just really can't do what a true vector supercomputer can. And anyways, making a multiprocessor system is nothing like making a vector machine.
What about instead of vibrating a mirror (which would as you said cause compression waves in the air and send pretty substantial vibrations through the machine (which could then travel to the leds throwing their focus off, but I digress). Why not have a box with mirrors on the outside and rotate it . Because you'd have four mirrors you'd only have to rotate it at 7.5 rps, you'd counteract any vibrations (assuming you balance the box right) and you'd get a smoother motion to boot (no slow down at the extremeties of the pictures, its hard to vibrate something in a manner thats not sinusoidal (though sinusoidal motion in that mirror could be counteracted by a cylindrical mirror afterwards but again I digress)).
The problem with using a row of 640 leds is that you once again have the fixed resolution problem (at least horizontally). I don't know if it would be feasable to rotate a miror box at 3925 rps (to do the horizontal scan with one led) but it would solve that problem.
IBM drives are on the whole very good. Yes the 75GXPs are crap. On the other hand the 40GXPs and 60GXPs are some of the best consumer level drives on the market right now. If you're trying to avoid a faulty drive don't go to a slightly less faulty alternative like Maxtor (I have seen too many Maxtors (of varying generations) fail, I will never trust them again) rather than one of the products that gave IBM their previous reputation for reliability.
P.S. No I don't work for or are in any way affiliated with IBM. I'm just a sysadmin thats run a lot of hardware into the ground over the years, and am a happy owner of a 60GXP.
I can at least vouch for the 60GXPs. From my own experience and everyone I've talked to about them they are one of the more reliable models IBM has come out with. I've put them in some pretty heavy use fileservers and never had any problems
>Think about that for a second. I have a *Lisp* >system that takes less disk space than 12M. That is >a huge, huge, amount of code. Sure, it is less than >the drivers, but considering that X does very >little, it is positively *enormous*.
Ok, but first of all this is stuff used by X clients, not the X server. These are things like libX11 and libXt, essential apis for X clients. Then you have the fact that alot of these apis have been pretty much depreciated for modern X programs. Gnome doesn't use Xintrinsics or the athena widgets. If you're running mostly gnome programs libXt and libXaw almost never loaded. Then there are things sitting around in there like libPEX. Don't even try to say you've used a PEX based program in the past 4 years. All in all, you're probably only going to use libX11, libICE, libSM, libXm and if you're using pretty antialiased fonts libXft and libXrender. Thats about four megs.
From a pretty standard XFree86 4.1.0 installation.
drivers/modules/extensions (the contents of the X11R6/lib/modules directory): 21M
essential libraries (X clients can't function without them, toplevel X11R6/lib): 12M
X server binary itself: 1.5M
I'd say the drivers are a fair chunk of that.
Is it just me or does this sound strangely similar to Plan 9.
I just find it funny that not only is Microsoft Research doing something that they think is new but was really done by bell labs back in the 80s, but back in the 80s every one looked at it and said "who cares".
Yes, he was forced to trademark it a while back because the publisher to some linux books tried to trademark it and then sue all the distribution makers for usage rights. Linus proved prior usage and trademarked it (with explicit permission given to anyone for free use) to keep someone else from doing that.
No he wasn't talking about removing usefull stuff. Specifically, he's talking about removing anything not usefull. All the driver version and "written by Joe Shmoe 2001" messages. Also the "blah.c driver loaded OK" are being considered for removal because, really, they serve no purpose. Error messages and actually usefull status updates (like the ide driver spitting out the device ids of the drives attached) will be left in. And for all those of you who really value the ego boost of having your name in the kernel boot, he also specifically said to leave this stuff in module load outputs. Just get it out of compiled in drivers.
Actually if you're listening to mp3s with something that uses EsounD (like xmms) just launch mozilla with esddsp. That fixes it.
They DIDN'T use a 2.4 test kernel. They used 2.2.16. They did however make sure all the libraries and support programs are up to date enough that a 2.4 kernel will drop in with an rpm.
Ok, even if he was trying to emulate MacOS and not the PowerPC archetecture in general, and even if it was MacOS X and not Darwin (the kernel...no more) that booted on x86 you'd still have the problem of getting a PowerPC binary to run on Intel hardware. Which is what he did.
Actually it is integrated into gnome now. Get the newest helix-gnome and look in the control center under "legacy applications".
But MS can't kill this by not running on it. They already do run on it. Its completely IA32 compatible. It just has the capability to do more with some 64bit instructions. If this chip just catches on thanks to the fact that its all around fast and people realize "hey....on linux its even better" then things could happen.