Any of the 8 registers can be used as the source or destination of most arithmetic operations arithmetic operation or as an index. Some operations only worked on specific registers, and the encoding was weird. I don't believe compilers typically generated things like string instructions directly... they were in the runtime library.
Most computers back then had special modes or instructions that only worked on specific GPRs. For example the PDP-11 used R6 and R7 for the stack pointer and program counter. The 8086 was more screwed up than most in all kinds of ways, of course, and the special register-specific operations were just the beginning.
Getting back to the topic, the fact that the AMD64 instruction set has shed so much of that crap is a much better reason to use it than the fact that it allows you to use 64-bit addresses.
I have to nitpick your first sentence about people feeling the need to upgrade frequently.
If you want to run typical third-party software on Mac OS X, you need to upgrade. Why? because developers don't bother to use the tools apple provides to check compatibility problems and avoid them. So if you're running 10.1, very little software (commercial or not) is tracking it. 10.2, you can still use 10.2 and have a decent chance of finding software you need. But you have to look.
A multicast network does not exist solely for the purpose of streaming media, and streaming media does not require the use of a multicast network.
IDGI, what's your point? If people were doing streaming media over the mbone in 1993 and 1994, then Real didn't invent it and it's more than 10 years old. The mbone was designed to stream stuff in general, and one of the things that was streamed over it was audio. Lots of it.
As an aside, as the Atari didn't address more than 16 bits of memory [I think you mean 24 bits] you could use the top (unused) half [byte] of the ax registers to store other stuff sometimes.
How dare you copy Microsoft!
Microsoft took years to learn that this was a bad idea. Some of their software broke when the Mac went over 512k because ofthis, then it broke on the Amiga when it went over 1M, then it broke on both platforms when the 68020 came out... which was apparently why Apple stuck with 24-bit addressing up through the SE/30.
Plus with each new version, the development time is getting longer
The release interval is getting longer because people resent the frequent need for upgrades... Apple does provide people with the tools needed to write software that's portable to previous versions, but there's an awful lot of developers who don't bother to try. Or even to understand the API: there's one program I ran into that depended on Panther only because the developer called out to PHP to fetch a web page instead of using the existing Cocoa frameworks.
Those of us not born yesterday remember Bill Gates vaporware announcement of "Windows" soon after the original Mac came out.
Windows was announced in 1983, the Mac came out in 1984. How did that happen? Easy... Microsoft was already developing software for the Mac before it was released, and Bill knew Microsoft had to do at least as well if they were going to stay in business.
"To create a new standard, it takes something that's not just a little bit different, it takes something that's really new and really captures people's imagination and the Macintosh, of all the machines I've ever seen, is the only one that meets that standard." -- Bill Gates, 1984
"the Mac is the only microcomputer beside the IBM PC worth writing software for." -- Bill Gates, 1984
Windows on Itanium was pre-doomed like Windows on Alpha and Windows on MIPS and Windows on Power. People run operating systems to run applications, and people run Windows to run IA32 applications. Unless your 64-bit, RISC, EPIC, or what have you version of Windows runs IA32 applications as fast as comparably priced IA32 version of Windows on IA32 hardware, what's the point?
Sure, IA64 has pretty good IA32 emulation, but Alpha had pretty good IA32 emulation as well thanks to the Freeport Express JIT recompiler. And that just wasn't good enough... It has to be cost-competitive with the best Intel 32-bit chips if it's going to succeed as a Windows platform long term.
So, really, this is the first commercially viable 64-bit windows. What came before this, the never-released 64-bit Alpha version and the Itanium version, well... they have to be considered betas.
The big reason for going to 64-bit Windows has nothing to do with the word size. The main reason is that AMD64 has shed another chunk of the 8086 instruction set legacy. The IA32 has 8 32-bit general purpose registers, about the same total register storage as the Cosmac 1802... a 4/8 bit microprocessor from the '70s. AMD64 gives you 16 64-bit registers, which is pretty small for a 64-bit machine (Alpha and Power have 32) but big enough to give the compiler room to work in, especially since it's also doubling the number of SSE registers.
Here's some other computers for comparison:
PDP-11, late '60s... 8 16-bit general purpose registers. VAX, '70s... 16 32-bit GPRs. 68000, ~'80... 8 32-bit GPRs, 8 32-bit index registers. z8000, ~'80... 16 16-bit registers. 8086, late '70s, 8 16-bit GPRs. MIPS, '80s, 32 32-bit registers. SPARC, ~'90... 32 32-bit GPRs, but only 8 were really usable as GPRs for the optimiser. Thus has hurt the Sparc's performance. Power PC, '90s, 32 32 or 64-bit GPRs Alpha, '90s, 32 64-bit registers
I would say the 4x register-file space increase is going to be far more important than the larger virtual memory.
There is only one update they need to do to IE, and they will never do it.
They need to abandon zones, put the application in charge of the security of a window, and NEVER let a window open, launch, link to, or reference a "more trusted" object than the one the link, embedded object, what have you is referenced from.
That means IE would be a hard sandbox. If you want to use ActiveX components that aren't sandboxed, you need to run a separate program.
Yes, that means that Windows Update would need to be a separate application shell around the HTML control. That's a teeny tiny problem compared to these sneaky damn zones.
I've only recently started using GUI mail clients, instead of command line ones that you go into, check your mail, then log out... and once you're out you're out. When I run any command in the shell, it checks for mail, so when I'm in the editor working on a program I don't get any mail alerts until I'm back at the command line and that's usually an OK place for a context switch.
I've noticed that the GUI app that's always running is MUCH more intrusive. I've turned off the "new mail" sound but still the count of messages in the icon is a distraction. It annoys me, so I feel like I have to go check it. I want to turn it off, but there's no equivalent of "now you're back in the shell, you have new mail, eh" then.
I'm with you on that. It can't be THAT hard for the OS to scan a keyboard, notice extra non-Standard (all the normal QUERTY bits that are SUPPOSED to be there) keys and offer them up for remapping.
Well, the problem is that the extra keys on a USB keyboard aren't keys,they're system control endpoints. But, yes, there are standards for all of them and they all have descriptors that describe them. It should still be able to put up a list of system control buttons and let me assign events to them.
And what I'm talking about is something even deeper and central to the system, and that's hotkey management in general, whether they're standard keys or not. There's no central place that lets an application say "I'm prepared to accept the following named events, and here's my preferred key bindingd for them" and then let the user say "when this keystroke comes in, send this event to this application".
There's PART of that, for the CMD-foo keypresses, you can bind them to menus and things when you're in an app. But the usual hotkey API is an undocumented set of Carbon calls that probably seeped in from classic Mac OS. That's the part that should be centrally managed, so you could bring up your hotkey manager and see Key=F11 - Shift=none - App=Expose - Cmd=Show Desktop and go down to Key=Unknown (%F6) - Shift=CMD - App=Expose - Cmd=Show Desktop and get Expose action from the otherwise unused "Menu" key on your Windows style keyboard.
Connecting that to your post, you should be able to go down further and select Key=Unknown (System Control 6) - Shift=any - App=iTunes - Cmd='set playlist to "Party Shuffle"'.
And of course you should be able to click "enter key" and hit the "search" key on your USB keyboard and find yourself looking at Key=Unknown (System Control 3) - Shift=none - App=Safari - Cmd=open http://www.google.com.
There's been a patch for E16 for a while now to have window transparency during moving.
Not if the rest of the system that's actually doing the heavy lifting for transparency is there... or is it faking transparency by snapshotting the window, the way Expose (which shows all the applications in their minimised forms updating in real time) doesn't?
FWIW, Enlightenment 17 + X11 looks like more of what I want out of a modern GUI than either Longhorn or OSX.
I just don't get Enlightenment.
It's a really complex window manager that does about the same as any other window manager except you can make the edges of windows look like a Giger painting.
And don't come back with Mac OS X and Expose and the lickable interface. Enlightenment doesn't do ANYTHING to let application partake of the cool window border action, and it doesn't give you window transparency, and it doesn't give you really nice consistent text input widgets and Services and contextual menu plugins... because it's not a toolkit or framework, it's just a window manager.
It just lets you do cool shit to the edges of windows.
Why is that so damn enticing? I honestly don't get it.
So you mean if I, say, post an ad that I'm looking for a cheap car [...]
Stop right there.
See the problem with your logic? Let me try that again, I'll hilight it.
So you mean if I, say, post an ad that I'm looking for a cheap car [...]
Ten grand for a magazine that probably cost a quarter when it came out? A specific issue? That's like offering a million bucks for a '55 Chevy Ambulance.
I'm sure they didn't expect people to do the equivalent of breaking into small town hospitals and driving off with their ambulances, but that's allegedly what happened. It's not anyone's fault, but it *is* Intel's responsibility. It would be a nice gesture (and give them good PR) if they'd come up with some kind of compensation.
No, the OP is right. Someone at Intel screwed up. I'm sure they had the best of intentions, but they did inadvertently encourage criminal behaviour. I'm sure they can afford some kind of compensation to the library that had their copy plundered, the good PR that would result would more than outbalance the cost.
Dud, if you're all raring to go out and start innovating all over Mac OS X, how about filling in some of the missing tools that can be done using supported APIs?
Why would anyone pay for a newsreader when Google Groups exists?
Um, because Google could arbitrarily change Google Groups at any moment and remove features that you depend on?
Oh, wait, Doctor Evil, that already happened.
(yes, it's their code, their hardware, they have a right to do it, the point is they can and do, and that's one good reason to do Usenet yourself instead of depending on the kindness of strangers)
Slashdot Mirror
I didn't think that the 8086 had any GPRs.
Any of the 8 registers can be used as the source or destination of most arithmetic operations arithmetic operation or as an index. Some operations only worked on specific registers, and the encoding was weird. I don't believe compilers typically generated things like string instructions directly... they were in the runtime library.
Most computers back then had special modes or instructions that only worked on specific GPRs. For example the PDP-11 used R6 and R7 for the stack pointer and program counter. The 8086 was more screwed up than most in all kinds of ways, of course, and the special register-specific operations were just the beginning.
Getting back to the topic, the fact that the AMD64 instruction set has shed so much of that crap is a much better reason to use it than the fact that it allows you to use 64-bit addresses.
I have to nitpick your first sentence about people feeling the need to upgrade frequently.
If you want to run typical third-party software on Mac OS X, you need to upgrade. Why? because developers don't bother to use the tools apple provides to check compatibility problems and avoid them. So if you're running 10.1, very little software (commercial or not) is tracking it. 10.2, you can still use 10.2 and have a decent chance of finding software you need. But you have to look.
A multicast network does not exist solely for the purpose of streaming media, and streaming media does not require the use of a multicast network.
IDGI, what's your point? If people were doing streaming media over the mbone in 1993 and 1994, then Real didn't invent it and it's more than 10 years old. The mbone was designed to stream stuff in general, and one of the things that was streamed over it was audio. Lots of it.
Or something that content providers want you to listen to, but not own.
If you can hear it, you can record it. Streaming media is no more secure than any other DRM-protected audio... that is, not protected at all.
No, just because something is multicast does not mean it is streaming media.
Do the Rolling Stones count as "media" then?
They were the second band on the Internet, back in 1994. Live on the mbone...
"Back in June 1993 when HTML was more common in alphabet soup and the MBone, or the Multicast Backbone, was another technical novelty, STD was the first band to perform live on the Internet."
As an aside, as the Atari didn't address more than 16 bits of memory [I think you mean 24 bits] you could use the top (unused) half [byte] of the ax registers to store other stuff sometimes.
How dare you copy Microsoft!
Microsoft took years to learn that this was a bad idea. Some of their software broke when the Mac went over 512k because ofthis, then it broke on the Amiga when it went over 1M, then it broke on both platforms when the 68020 came out... which was apparently why Apple stuck with 24-bit addressing up through the SE/30.
Plus with each new version, the development time is getting longer
The release interval is getting longer because people resent the frequent need for upgrades... Apple does provide people with the tools needed to write software that's portable to previous versions, but there's an awful lot of developers who don't bother to try. Or even to understand the API: there's one program I ran into that depended on Panther only because the developer called out to PHP to fetch a web page instead of using the existing Cocoa frameworks.
Windows was announced in 1983, the Mac came out in 1984. How did that happen? Easy... Microsoft was already developing software for the Mac before it was released, and Bill knew Microsoft had to do at least as well if they were going to stay in business.
didn't apple steal the whole idea of the graphical interface and the mouse from xerox?
No, Apple licensed it from Xerox. So did Microsoft, for that matter.
If M$ had a customer base as small as Apple's, I'm sure they'd be able to put out new releases every six months as well.
Apple's putting out new major versions about every 18 months these days.
Windows on Itanium was pre-doomed like Windows on Alpha and Windows on MIPS and Windows on Power. People run operating systems to run applications, and people run Windows to run IA32 applications. Unless your 64-bit, RISC, EPIC, or what have you version of Windows runs IA32 applications as fast as comparably priced IA32 version of Windows on IA32 hardware, what's the point?
Sure, IA64 has pretty good IA32 emulation, but Alpha had pretty good IA32 emulation as well thanks to the Freeport Express JIT recompiler. And that just wasn't good enough... It has to be cost-competitive with the best Intel 32-bit chips if it's going to succeed as a Windows platform long term.
So, really, this is the first commercially viable 64-bit windows. What came before this, the never-released 64-bit Alpha version and the Itanium version, well... they have to be considered betas.
The big reason for going to 64-bit Windows has nothing to do with the word size. The main reason is that AMD64 has shed another chunk of the 8086 instruction set legacy. The IA32 has 8 32-bit general purpose registers, about the same total register storage as the Cosmac 1802... a 4/8 bit microprocessor from the '70s. AMD64 gives you 16 64-bit registers, which is pretty small for a 64-bit machine (Alpha and Power have 32) but big enough to give the compiler room to work in, especially since it's also doubling the number of SSE registers.
Here's some other computers for comparison:
PDP-11, late '60s... 8 16-bit general purpose registers.
VAX, '70s... 16 32-bit GPRs.
68000, ~'80... 8 32-bit GPRs, 8 32-bit index registers.
z8000, ~'80... 16 16-bit registers.
8086, late '70s, 8 16-bit GPRs.
MIPS, '80s, 32 32-bit registers.
SPARC, ~'90... 32 32-bit GPRs, but only 8 were really usable as GPRs for the optimiser. Thus has hurt the Sparc's performance.
Power PC, '90s, 32 32 or 64-bit GPRs
Alpha, '90s, 32 64-bit registers
I would say the 4x register-file space increase is going to be far more important than the larger virtual memory.
It's notthe new email showing up, it's the ongoing increae in the mail counter... after it gets to double digits it really annoys me.
It's not as bad as people calling me on the damned telephone instead of sending me email though.
There is only one update they need to do to IE, and they will never do it.
They need to abandon zones, put the application in charge of the security of a window, and NEVER let a window open, launch, link to, or reference a "more trusted" object than the one the link, embedded object, what have you is referenced from.
That means IE would be a hard sandbox. If you want to use ActiveX components that aren't sandboxed, you need to run a separate program.
Yes, that means that Windows Update would need to be a separate application shell around the HTML control. That's a teeny tiny problem compared to these sneaky damn zones.
I've only recently started using GUI mail clients, instead of command line ones that you go into, check your mail, then log out... and once you're out you're out. When I run any command in the shell, it checks for mail, so when I'm in the editor working on a program I don't get any mail alerts until I'm back at the command line and that's usually an OK place for a context switch.
I've noticed that the GUI app that's always running is MUCH more intrusive. I've turned off the "new mail" sound but still the count of messages in the icon is a distraction. It annoys me, so I feel like I have to go check it. I want to turn it off, but there's no equivalent of "now you're back in the shell, you have new mail, eh" then.
Any ideas?
I'm with you on that. It can't be THAT hard for the OS to scan a keyboard, notice extra non-Standard (all the normal QUERTY bits that are SUPPOSED to be there) keys and offer them up for remapping.
Well, the problem is that the extra keys on a USB keyboard aren't keys,they're system control endpoints. But, yes, there are standards for all of them and they all have descriptors that describe them. It should still be able to put up a list of system control buttons and let me assign events to them.
And what I'm talking about is something even deeper and central to the system, and that's hotkey management in general, whether they're standard keys or not. There's no central place that lets an application say "I'm prepared to accept the following named events, and here's my preferred key bindingd for them" and then let the user say "when this keystroke comes in, send this event to this application".
There's PART of that, for the CMD-foo keypresses, you can bind them to menus and things when you're in an app. But the usual hotkey API is an undocumented set of Carbon calls that probably seeped in from classic Mac OS. That's the part that should be centrally managed, so you could bring up your hotkey manager and see Key=F11 - Shift=none - App=Expose - Cmd=Show Desktop and go down to Key=Unknown (%F6) - Shift=CMD - App=Expose - Cmd=Show Desktop and get Expose action from the otherwise unused "Menu" key on your Windows style keyboard.
Connecting that to your post, you should be able to go down further and select Key=Unknown (System Control 6) - Shift=any - App=iTunes - Cmd='set playlist to "Party Shuffle"'.
And of course you should be able to click "enter key" and hit the "search" key on your USB keyboard and find yourself looking at Key=Unknown (System Control 3) - Shift=none - App=Safari - Cmd=open http://www.google.com.
There's been a patch for E16 for a while now to have window transparency during moving.
Not if the rest of the system that's actually doing the heavy lifting for transparency is there... or is it faking transparency by snapshotting the window, the way Expose (which shows all the applications in their minimised forms updating in real time) doesn't?
FWIW, Enlightenment 17 + X11 looks like more of what I want out of a modern GUI than either Longhorn or OSX.
I just don't get Enlightenment.
It's a really complex window manager that does about the same as any other window manager except you can make the edges of windows look like a Giger painting.
And don't come back with Mac OS X and Expose and the lickable interface. Enlightenment doesn't do ANYTHING to let application partake of the cool window border action, and it doesn't give you window transparency, and it doesn't give you really nice consistent text input widgets and Services and contextual menu plugins... because it's not a toolkit or framework, it's just a window manager.
It just lets you do cool shit to the edges of windows.
Why is that so damn enticing? I honestly don't get it.
there's far too much hero-worship of Linus
Odd, I was thinking more of ESR.
I still can't imagine how you can equate that to being Intel's responsibility.
That's OK, I'm not going to feel guilty about your lack of imagination. I've gotten over that kind of thing. Mostly.
So you mean if I, say, post an ad that I'm looking for a cheap car [...]
Stop right there.
See the problem with your logic? Let me try that again, I'll hilight it.
So you mean if I, say, post an ad that I'm looking for a cheap car [...]
Ten grand for a magazine that probably cost a quarter when it came out? A specific issue? That's like offering a million bucks for a '55 Chevy Ambulance.
I'm sure they didn't expect people to do the equivalent of breaking into small town hospitals and driving off with their ambulances, but that's allegedly what happened. It's not anyone's fault, but it *is* Intel's responsibility. It would be a nice gesture (and give them good PR) if they'd come up with some kind of compensation.
No, the OP is right. Someone at Intel screwed up. I'm sure they had the best of intentions, but they did inadvertently encourage criminal behaviour. I'm sure they can afford some kind of compensation to the library that had their copy plundered, the good PR that would result would more than outbalance the cost.
Dud, if you're all raring to go out and start innovating all over Mac OS X, how about filling in some of the missing tools that can be done using supported APIs?
Why would anyone pay for a newsreader when Google Groups exists?
Um, because Google could arbitrarily change Google Groups at any moment and remove features that you depend on?
Oh, wait, Doctor Evil, that already happened.
(yes, it's their code, their hardware, they have a right to do it, the point is they can and do, and that's one good reason to do Usenet yourself instead of depending on the kindness of strangers)