Ben Herrenshmidt (I think I got the name right) wrote the kernel for the new motherboards, winning instant fame and popularity...
He got it all working a while ago, so it should be included in the friendly GUI install included in this new LinuxPPC distro. Speaking of which, I just ordered it for my G4 Sawtooth!
Where is my mind? mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
High-end IBM servers, ancient BeBoxes, Mac clones, embedded systems:v) .
But most importantly, POP. The PowerPC Open Platform is allowing many companies like Prophet Systems (they were the first, I can't remember the others...) to develop their own motherboards based on the CHRP standard originally designed to support a wide variety of processors. These are not available yet, but when they are, they will be dirt cheap (especially considering that G4's are less expensive than Pentiums).
Where is my mind? mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
FWB does all that's physically possible. You say [in a post below] that it can only shrink, not grow, partitions. Growing partitions to the point that the allocation block size changes is impossible. You would need to reorganize all the files on the disk. If yer gonna install another OS, why would you be expanding the partition anyway?
If you wanna get down+dirty, and you're using a SCSI drive, I know how to rewrite the partition map to do just this (using the MacOS API). Hint: you can't do it through the disk driver.
Where is my mind? mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
I use Metrowerks. Not free, but OK. Optimizations are alright, but I really like RISC asm programming. You don't have to worry about all kinds of addressing modes, etc., or what gets translated to what microcode. Really besides the point, I guess, but asm is not and will never be a dead language.
Where is my mind? mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
For all you curious Mac people out there, there is a command line built into every PCI mac. On later models (G3's, I think anything older needs a nasty 2-computer setup) you can simply access the command line by holding down cmd-opt-o-f as the computer starts up. Instructions for using the command line can be found in Apple's dev technotes, 1061 and 1062.
If you're just bored, playing with this can be fun. Forth is a twisted language.
Where is my mind? mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
I've been watching the Prophet Systems site sporadically, and it looks like the development is just slow. The publicly posted POP design is just a 4-year-old CHRP board, so the companies have to do a lot of development on their own (if they want better than 1x AGP, ATA/33, etc.).
Where is my mind? mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
[I'm sorry about these HTML tags here, but Slashdot's Extrans posting mode seems to be broken]: <I>This page looks very unproffesional.</I>
That's because it wasn't put up by a pro. Apple has a huge number of sites/pages on the Apple.com domain that are maintained by people inside Apple who want the content to be there. Many of their dev pages aren't candy-coated.
I think the main point of this page was to show support of Linux by hosting Ben Herrenschmidt's kernel page off Apple.com. (Ben ported and maintains the kernels for Apple's latest machines.) Aside from the kernels page, this site has a total lack of content. Everyone's noted the spelling errors, and even the lack of decent links. I'd think Apple would try harder to get actual info on doing installations to the public.
Where is my mind? mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
If you look at Be's explanation of the no-G3-support matter (on their website, sorry no link), you'll see that they were dead afraid of reverse-engineering what Apple wouldn't tell them. I can't imagine it would be too hard for them to figure out - they were just afraid of a big angry Apple. Apple chose to support Linux, whether just to keep up or because the DTS people there wanted it so bad:v) , and Be was purposefully scared off.
Where is my mind? mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
I'm sorry this is offtopic, but if I posted it to the proper thread, that would undo some moderation:v(. I just have to contact the author.
The perl code you posted to the "obfuscated C contest" article is useful, but the &nbsp;'s it generates take a *lot* of space, bandwidth-wise. Don't most monospaced fonts have monospaced spaces?
I know on Windows it sucks, and on the Mac it's better, but how good is Linux at regulating what's turned on when, and processor cycling and whatever? Are the necessary features included w/ the major distros or what? An efficient preemptively multitasking OS should be best at this, so if people want to work for it, Linux laptops could potentially run a lot longer on a charge...
I think they're not open-sourcing it because there would be nothing to gain. It's at most a few K of assembler code. Their dev team can handle that, it would be more work to coordinate a public effort.
So long as people can fiddle around with the layer (replacing it, etc.) to write their own emulators or whatever, I don't think anything's lost.
Reading about this "vector-based" UI made me think about an experience I had with an original Asteroids arcade game. Back then, there were no bitmapped graphics, and the drawing was done on the screen by moving the electron beam around the screen in the pattern you wanted the shapes to take. (As opposed to scanning across repeatedly and switching the beam on and off.) The effect was no flicker and infinite resolution. Incredibly beautiful to look at. But it was monochrome.
So I started thinkin' - how could we make a vector based monitor that is color? If you bent the beam between the mask and the phosphor, could you cause it to selectively hit r/g/b phosphors in phase? The mask could be indefinitely fine, allowing for virtual infinite resolution. Bitmaps, if you needed them (I'm thinking of a super-ergonomic work computer), could be "emulated"?
If anybody who knows CRT's can help me, I'd be interested to know. It'd just be so easy on the eyes to have a decent sharp, smooth display...
If CD's decayed that quickly, a lot of environmentalists would be very happy. Plastic just doesn't decay that quickly in air, unless pollution has become a lot worse lately.
Welll, this is turning into a problem... The link is here.
The version I got pissed off about was the Mac version, and what pissed me off was the amount of time it wasn't even spending inside the cruncher. Simply doing random breaks with a low-level debugger showed that it's spending way too much time in graphics routines, and the disassembler shows that the graphics routines were likely inefficiently cross-compiled from x86. IMHO, if resorting to cross-compiling, they should be looking for help.
It's always better to get at the root of, to solve, the problem. Space exploration could be a way to let us continue to decline and reproduce without ill effects, but will we be any better off? We should solve the problems we have here before we spread them through the universe.
Any catastophe we bring upon ourselves is more likely to teach us a lesson than to kill us off completely.
Slashdot Mirror
Couldn't you do serial ATA with FireWire and a software intermediate layer?
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
VGA and S-Video ports for dual display and video mirroring
That's it. I'm getting one of these.
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
Ben Herrenshmidt (I think I got the name right) wrote the kernel for the new motherboards, winning instant fame and popularity...
He got it all working a while ago, so it should be included in the friendly GUI install included in this new LinuxPPC distro. Speaking of which, I just ordered it for my G4 Sawtooth!
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
High-end IBM servers, ancient BeBoxes, Mac clones, embedded systems :v) .
But most importantly, POP. The PowerPC Open Platform is allowing many companies like Prophet Systems (they were the first, I can't remember the others...) to develop their own motherboards based on the CHRP standard originally designed to support a wide variety of processors. These are not available yet, but when they are, they will be dirt cheap (especially considering that G4's are less expensive than Pentiums).
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
FWB does all that's physically possible. You say [in a post below] that it can only shrink, not grow, partitions. Growing partitions to the point that the allocation block size changes is impossible. You would need to reorganize all the files on the disk. If yer gonna install another OS, why would you be expanding the partition anyway?
If you wanna get down+dirty, and you're using a SCSI drive, I know how to rewrite the partition map to do just this (using the MacOS API). Hint: you can't do it through the disk driver.
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
I use Metrowerks. Not free, but OK. Optimizations are alright, but I really like RISC asm programming. You don't have to worry about all kinds of addressing modes, etc., or what gets translated to what microcode.
Really besides the point, I guess, but asm is not and will never be a dead language.
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
For all you curious Mac people out there, there is a command line built into every PCI mac. On later models (G3's, I think anything older needs a nasty 2-computer setup) you can simply access the command line by holding down cmd-opt-o-f as the computer starts up. Instructions for using the command line can be found in Apple's dev technotes, 1061 and 1062.
If you're just bored, playing with this can be fun. Forth is a twisted language.
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
I've been watching the Prophet Systems site sporadically, and it looks like the development is just slow. The publicly posted POP design is just a 4-year-old CHRP board, so the companies have to do a lot of development on their own (if they want better than 1x AGP, ATA/33, etc.).
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
I'm sorry this is offtopic, but I've gotta say this is the most poetic post I've ever seen.
Thanks for writing Jon,
The trolls say we,
Don't speak for me.
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
I wanna see Microsoft DOS!
Where's Bill when you need him?
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
OK, now it's REALLY pissing me off what gets moderated as offtopic. C'mon, moderate this one down, I dare you!
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
[I'm sorry about these HTML tags here, but Slashdot's Extrans posting mode seems to be broken]:
<I>This page looks very unproffesional.</I>
That's because it wasn't put up by a pro. Apple has a huge number of sites/pages on the Apple.com domain that are maintained by people inside Apple who want the content to be there. Many of their dev pages aren't candy-coated.
I think the main point of this page was to show support of Linux by hosting Ben Herrenschmidt's kernel page off Apple.com. (Ben ported and maintains the kernels for Apple's latest machines.)
Aside from the kernels page, this site has a total lack of content. Everyone's noted the spelling errors, and even the lack of decent links. I'd think Apple would try harder to get actual info on doing installations to the public.
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
You replied to an AC.
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
If you look at Be's explanation of the no-G3-support matter (on their website, sorry no link), you'll see that they were dead afraid of reverse-engineering what Apple wouldn't tell them. I can't imagine it would be too hard for them to figure out - they were just afraid of a big angry Apple. :v) , and Be was purposefully scared off.
Apple chose to support Linux, whether just to keep up or because the DTS people there wanted it so bad
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
I'm sorry. I'm not usually a "whiner". But that is a crappy post.
An astronomer says that life manipulates quantum physics? What are we gonna see next, "vampire girl discovered"?
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
Do you have nothing better to moderate?
Where is my mind?
I'm sorry this is offtopic, but if I posted it to the proper thread, that would undo some moderation :v(. I just have to contact the author.
The perl code you posted to the "obfuscated C contest" article is useful, but the &nbsp;'s it generates take a *lot* of space, bandwidth-wise. Don't most monospaced fonts have monospaced spaces?
Where is my mind?
I know on Windows it sucks, and on the Mac it's better, but how good is Linux at regulating what's turned on when, and processor cycling and whatever? Are the necessary features included w/ the major distros or what? An efficient preemptively multitasking OS should be best at this, so if people want to work for it, Linux laptops could potentially run a lot longer on a charge...
Where is my mind?
What happens when you turn it off? Does all the light that was trapped come back out and burn everything around it? Sounds nasty to me.
Where is my mind?
I think they're not open-sourcing it because there would be nothing to gain. It's at most a few K of assembler code. Their dev team can handle that, it would be more work to coordinate a public effort.
So long as people can fiddle around with the layer (replacing it, etc.) to write their own emulators or whatever, I don't think anything's lost.
Where is my mind?
Reading about this "vector-based" UI made me think about an experience I had with an original Asteroids arcade game. Back then, there were no bitmapped graphics, and the drawing was done on the screen by moving the electron beam around the screen in the pattern you wanted the shapes to take. (As opposed to scanning across repeatedly and switching the beam on and off.)
The effect was no flicker and infinite resolution. Incredibly beautiful to look at. But it was monochrome.
So I started thinkin' - how could we make a vector based monitor that is color? If you bent the beam between the mask and the phosphor, could you cause it to selectively hit r/g/b phosphors in phase? The mask could be indefinitely fine, allowing for virtual infinite resolution. Bitmaps, if you needed them (I'm thinking of a super-ergonomic work computer), could be "emulated"?
If anybody who knows CRT's can help me, I'd be interested to know. It'd just be so easy on the eyes to have a decent sharp, smooth display...
Where is my mind?
If CD's decayed that quickly, a lot of environmentalists would be very happy. Plastic just doesn't decay that quickly in air, unless pollution has become a lot worse lately.
Where is my mind?
Welll, this is turning into a problem...
The link is here.
The version I got pissed off about was the Mac version, and what pissed me off was the amount of time it wasn't even spending inside the cruncher. Simply doing random breaks with a low-level debugger showed that it's spending way too much time in graphics routines, and the disassembler shows that the graphics routines were likely inefficiently cross-compiled from x86. IMHO, if resorting to cross-compiling, they should be looking for help.
Where is my mind?
It's always better to get at the root of, to solve, the problem. Space exploration could be a way to let us continue to decline and reproduce without ill effects, but will we be any better off? We should solve the problems we have here before we spread them through the universe.
Any catastophe we bring upon ourselves is more likely to teach us a lesson than to kill us off completely.
Where is my mind?
The link in the message body, to the comment I wrote earlier.
Where is my mind?