A lot of the concepts behind Next are in OS X but to say it's OpenStep with a Mac interface shows how blissfully ignorant you are of both operating systems.
Having used MacOS and OpenStep, and having switched from my SE/30 to a NeXT Cube in 1993 (without looking back), I'd have to say OS X is a heck of a lot more like OpenStep than that stunted, underpowered runt OS X replaced.
It's really nice to use an operating system that wasn't designed to run on hardware a little better than an Apple II gs.
Anyone who says OS X is more MacOS than OpenStep has severe delusions of the old Mac OS's grandeur.
If you take a look at O'Reilley's History of Programming Languages chart [oreilly.com] you'll see that Objective-C was just one of many parents to Java.
Java was not inspired by any one language, but rather the desire to see some of the nicer features of a variety of languages brought together.
This is true, but that chart doesn't indicate the relative influence of each parent language.
There's an old Usenet Post by Patrick Naughton, where he talks about how much Objective-C influenced Java. He like ObjC so much he was on the verge of moving over to NeXT, when he was persuaded to stay and work on this new project that turned out to be Java. And a number of ex-NeXTers worked on Java, early on.
Ok, dimwit, go find some OS X apps with source available, even the circa first release source code, and recompile them for the latest version of OpenStep.
TextEdit's source code hasn't changed much since the OpenStep days. It dates back to 1995.
The major change between OpenStep and OS X, as far as Cocoa apps go, is the change from Display Postscript to Quartz. And that's mostly an issue only if the application does much drawing.
"In Quartz 2D, you deal with a graphics context; in QuickDraw, you deal with a GrafPort. Same thing."
Guess what? Display Postscript had a graphics context, too.
This small similarity is kinda overshadowed by the fact that Quartz has a simplified set of drawing primitives, compared to QuickDraw, and Quickdraw uses regions, while Quartz uses bezier-based paths, which are rather different.
Check Google Groups, you'll find people claiming to be running such a system.
One such person, in 2002, suggests using an ABit KT7A motherboard, which works up to 1.2 GHz. Also, don't use more than 512 MB RAM.
I'm sure it requires some older hardware, and careful selection of parts, but it seems to work. Or, rather, have worked. I don't know if it would work with a vintage 2005 Athlon motherboard.
Quartz uses a PDF imaging model. Display Postscript uses a Postscript imaging model. PDF's imaging model is not terribly different from Postscript's imaging model.
Quickdraw's imaging model is like neither.
Quartz is architected quite differently from Quickdraw, and is rather more complex, because it has more to do.
Quartz does alpha compositing. Quickdraw does not.
I remember one cool University of Michigan software project that required a pseudotty for each remote user
Forgive me, but I seriously doubt that was a drop in the bucket compared to the other competitive challenges NeXT had in the 90s.
And really, the focus on high-dollar customers like financial traders was also part of what killed them. In the mid-90s I could have written and sold a ton of great solutions built on NeXT technology, but only financial traders could afford to license the NeXT OS or runtime.
Um, sure. By 1995, I don't think pricing would have helped much, if at all. NeXT was already facing Win95 and Java.
Not true, we had a cube, it came with an optical drive, and didn't have a floppy.
I specifically said "Turbo Cube". A 1992 revision, with a 33Mhz 68040 and no optical.
I believe the first 040 Cubes, which were 25Mhz, did support the optical. Their motherboards were sold as upgrades for the 030 Cubes, and I know they worked with the optical, because mine did.
Your going to need some strong, immediate, and obvious reasons to get a CTO to buy all new boxes with 'new' OS and interface.
Startups are buying new boxes, so can choose the platform they feel gives them an advantage, if one exists.
NeXT's biggest hardware customers were companies like investment banks and energy traders. It is not unheard-of for a company to take a risk on an unusual hardware platform. It's just far more rare nowadays. It was easier in 1991 when PCs and their operating systems sucked so badly.
The menus on the NeXT were a vertical stack of tiles, by default at the top left of the screen.
Sub-menus could be torn off and positioned wherever you wanted.
You could also obtain a full, temporary duplicate top-level menu by right-clicking. The duplicate would appear where you clicked, and disappear when you were done.
I imagine that could work a lot better on a multi-headed system than having to throw the cursor back and forth across thousands of pixels of real-estate. Fitts' law or no.
"Actually, the problem they had, is that nearly nobody in the industry was used to OOP."
The real killer was that everyone in the industry got religion when Java came out. And it sucked the air out of the space.
Just prior to Java's debut, NeXT and Sun had been working on a version of the OpenStep development environment (which used Objective-C, naturally) that ran on Solaris. That went bye-bye soon enough.
"Kits" (proprietary software--collections of ObjC objects and classes--one was encouraged to build dependencies upon) were obsoleted quite quickly, frustrating developers.
Yes, I can certainly see why developers would be upset that NeXT gave them frameworks to build upon, which let them build their highly profitable trading systems very, very quickly. No, what they really wanted was a primitive system which required them to start from scratch.
Unfortunately for your thesis, those "kits" were what NeXT customers really wanted, and what kept the company going so long.
The mono monitor was ribbed, or flanged. I have two in the room with me. The monitor in the video is not. It also looks too big to be the mono monitor, which only came in 17".
Also, the mono monitor had fat rubber rollers at the front of the base. It actually looked a lot like the old Apple IIc greenscreen monitor, which was designed by the same company (frogdesign). The monitor in the picture lacks the rollers.
(There was a differently-designed mono monitor towards the very end of the black hardware era (introduced in October of 1992). I don't recall if it had the fins, but it surely wasn't that big.)
Really, why would Steve Jobs be sitting in front of a low-end slab when he could sit in front of the most tricked-out color box they had available? That would involve their top-end monitor, the Hitachi 21".
"(wasn't a Next workstation something like $20 grand?)"
From a 1992 Usenet post of the Winter 1992 price list
NeXTstation 8-1MB SIMMS, 105MB HD $3775
NeXTstation Turbo 2-4MB SIMMS, 250MB HD 4775 NeXTstation Turbo 2-8MB SIMMS, 250MB HD 5775 NeXTstation Turbo 2-8MB SIMMS, 400MB HD 6775 NeXTstation Turbo 4-8MB SIMMS, 250MB HD 7775 NeXTstation Turbo 4-8MB SIMMS, 400MB HD 8775
NeXTstation Color 4-4MB SIMMS, 105MB HD 5650
NeXTstation Turbo Color 2-8MB SIMMS, 250MB HD 6650 NeXTstation Turbo Color 2-8MB SIMMS, 400MB HD 7650 NeXTstation Turbo Color 4-8MB SIMMS, 250MB HD 8650 NeXTstation Turbo Color 4-8MB SIMMS, 400MB HD 9650
These prices are in the ballpark of comparable machines from Sun and Apple.
but he did say he was going to port to 486. I can't help but wonder if a 486 could do this kind of stuff (a dx 100 could, but I think the dx33s where current when this was being done). All I can say is, what the heck happened?
It was ported to Intel in the 486 era, but it didn't really become practical to run until the Pentium 2. Ran pretty well on my AMD K6-350, if I recall correctly. Supposed to scream on Athlons.
In addition to Intel, it was ported, and sold, to run on Sun Sparc workstations and HP PA-RISC workstations.
I've read a bit of the history (I hear those MO drives they Next Stations ran off of were kinda buggy), but this is big enough stuff that they should have been able to get through a few lean years and sell the technology....
It wasn't the stations that had the Optical drive, it was the cube. That was the machine that got really expensive, when loaded up with a NeXTDimension color graphics card, big hard disks, and lots of RAM. The Optical was dropped before very long, and the Cube just shipped with a floppy drive. I think the Turbo Cube (33 MHz) couldn't even connect to the optical drive.
What happend to NeXT is (roughly) this:
First, customers realized they didn't so much want the hardware, they wanted the operating system. So NeXT dropped hardware and started doing their OS for other peoples' hardware.
Second, customers realized it wasn't so much the operating system they wanted, it was the development tools. So NeXT came up with a way to run the development tools on NT. And they had their WebObjects product, which let people use NeXT development tools to do web apps. So they de-emphasized the OS.
Then Apple bought them. The dev tools for NT were de-emphasized, except as a way to do WebObjects development. The OS was refreshed and updated, a process which continues.
and the whole Amidala giving birth to Luke and Leia thing
And the world waits to find out if the caesarian is done with a lightsaber. Or the episiotomy, if they go that way.
You know what'd rock? If the babies pop out and they're half-Gungan.
That'd rock. You'd hear fanboy heads explode all over the world. "Noooo! That's not... Can't be!... Nooo! Luke and Leia aren't Jar-jar's children!! Nooo!"
Slashdot Mirror
A lot of the concepts behind Next are in OS X but to say it's OpenStep with a Mac interface shows how blissfully ignorant you are of both operating systems.
Having used MacOS and OpenStep, and having switched from my SE/30 to a NeXT Cube in 1993 (without looking back), I'd have to say OS X is a heck of a lot more like OpenStep than that stunted, underpowered runt OS X replaced.
It's really nice to use an operating system that wasn't designed to run on hardware a little better than an Apple II gs.
Anyone who says OS X is more MacOS than OpenStep has severe delusions of the old Mac OS's grandeur.
Webster.app had a funny thing it would do if you looked up 'gullible'.
I hope it makes a reappearance.
If you take a look at O'Reilley's History of Programming Languages chart [oreilly.com] you'll see that Objective-C was just one of many parents to Java.
Java was not inspired by any one language, but rather the desire to see some of the nicer features of a variety of languages brought together.
This is true, but that chart doesn't indicate the relative influence of each parent language.
There's an old Usenet Post by Patrick Naughton, where he talks about how much Objective-C influenced Java. He like ObjC so much he was on the verge of moving over to NeXT, when he was persuaded to stay and work on this new project that turned out to be Java. And a number of ex-NeXTers worked on Java, early on.
Wasnt OpenStep first priced at $10,000; very few takers.
For the most part the OS cost about $795.
The developer tools, however, were separate, and cost something like $4995.
There was an academic package which included OS and developer tools, and cost $295.
I believe there were some limited-time discounted price promotions, particularly around the time that the Intel port appeared.
Ok, dimwit, go find some OS X apps with source available, even the circa first release source code, and recompile them for the latest version of OpenStep.
TextEdit's source code hasn't changed much since the OpenStep days. It dates back to 1995.
The major change between OpenStep and OS X, as far as Cocoa apps go, is the change from Display Postscript to Quartz. And that's mostly an issue only if the application does much drawing.
"In Quartz 2D, you deal with a graphics context; in QuickDraw, you deal with a GrafPort. Same thing."
Guess what? Display Postscript had a graphics context, too.
This small similarity is kinda overshadowed by the fact that Quartz has a simplified set of drawing primitives, compared to QuickDraw, and Quickdraw uses regions, while Quartz uses bezier-based paths, which are rather different.
Show me an Athlon running Openstep :)
Check Google Groups, you'll find people claiming to be running such a system.
One such person, in 2002, suggests using an ABit KT7A motherboard, which works up to 1.2 GHz. Also, don't use more than 512 MB RAM.
I'm sure it requires some older hardware, and careful selection of parts, but it seems to work. Or, rather, have worked. I don't know if it would work with a vintage 2005 Athlon motherboard.
I'd suggest a virtual reality 3D interface, but it's pretty clear that the White House and Congress are operating in a virtual reality already .
I just hope their next level isn't called "Knee Deep In The Dead".
Quartz uses a PDF imaging model. Display Postscript uses a Postscript imaging model. PDF's imaging model is not terribly different from Postscript's imaging model.
Quickdraw's imaging model is like neither.
Quartz is architected quite differently from Quickdraw, and is rather more complex, because it has more to do.
Quartz does alpha compositing. Quickdraw does not.
I remember one cool University of Michigan software project that required a pseudotty for each remote user
Forgive me, but I seriously doubt that was a drop in the bucket compared to the other competitive challenges NeXT had in the 90s.
And really, the focus on high-dollar customers like financial traders was also part of what killed them. In the mid-90s I could have written and sold a ton of great solutions built on NeXT technology, but only financial traders could afford to license the NeXT OS or runtime.
Um, sure. By 1995, I don't think pricing would have helped much, if at all. NeXT was already facing Win95 and Java.
Not true, we had a cube, it came with an optical drive, and didn't have a floppy.
I specifically said "Turbo Cube". A 1992 revision, with a 33Mhz 68040 and no optical.
I believe the first 040 Cubes, which were 25Mhz, did support the optical. Their motherboards were sold as upgrades for the 030 Cubes, and I know they worked with the optical, because mine did.
Your going to need some strong, immediate, and obvious reasons to get a CTO to buy all new boxes with 'new' OS and interface.
Startups are buying new boxes, so can choose the platform they feel gives them an advantage, if one exists.
NeXT's biggest hardware customers were companies like investment banks and energy traders. It is not unheard-of for a company to take a risk on an unusual hardware platform. It's just far more rare nowadays. It was easier in 1991 when PCs and their operating systems sucked so badly.
The menus on the NeXT were a vertical stack of tiles, by default at the top left of the screen.
Sub-menus could be torn off and positioned wherever you wanted.
You could also obtain a full, temporary duplicate top-level menu by right-clicking. The duplicate would appear where you clicked, and disappear when you were done.
I imagine that could work a lot better on a multi-headed system than having to throw the cursor back and forth across thousands of pixels of real-estate. Fitts' law or no.
the color picker (except for the fact that it was a grayscale monitor)
On OS/X, try going into Universal Access, and turning on greyscale display mode.
(Then adjust the contrast slider a little to make things stand out a bit more.)
IIRC, Knowledge Navigator was John Sculley's baby, from after Jobs was ousted.
I think Jobs would sooner give Michael Dell a blumpkin than do anything related to Knowledge Navigator.
Or Mortor, for that matter.
Too bad most of the major apps on OS X are sticking with the Carbon route to avoid rewriting their codebases.
Well, I'd say it's the "legacy" apps. They're they Mac world's equivalent of those 30 year old Cobol apps that your bank depends on.
But the Mac apps that get the big buzz, probably tend to be Cocoa. Delicious Library, NetNewsWire, etc.
"Actually, the problem they had, is that nearly nobody in the industry was used to OOP."
The real killer was that everyone in the industry got religion when Java came out. And it sucked the air out of the space.
Just prior to Java's debut, NeXT and Sun had been working on a version of the OpenStep development environment (which used Objective-C, naturally) that ran on Solaris. That went bye-bye soon enough.
"Kits" (proprietary software--collections of ObjC objects and classes--one was encouraged to build dependencies upon) were obsoleted quite quickly, frustrating developers.
Yes, I can certainly see why developers would be upset that NeXT gave them frameworks to build upon, which let them build their highly profitable trading systems very, very quickly. No, what they really wanted was a primitive system which required them to start from scratch.
Unfortunately for your thesis, those "kits" were what NeXT customers really wanted, and what kept the company going so long.
The mono monitor was ribbed, or flanged. I have two in the room with me. The monitor in the video is not. It also looks too big to be the mono monitor, which only came in 17".
Also, the mono monitor had fat rubber rollers at the front of the base. It actually looked a lot like the old Apple IIc greenscreen monitor, which was designed by the same company (frogdesign). The monitor in the picture lacks the rollers.
(There was a differently-designed mono monitor towards the very end of the black hardware era (introduced in October of 1992). I don't recall if it had the fins, but it surely wasn't that big.)
Really, why would Steve Jobs be sitting in front of a low-end slab when he could sit in front of the most tricked-out color box they had available? That would involve their top-end monitor, the Hitachi 21".
And it's not like Jobs is bad at doing demos...
Jonathan Hendry
From a 1992 Usenet post of the Winter 1992 price listThese prices are in the ballpark of comparable machines from Sun and Apple.
but he did say he was going to port to 486. I can't help but wonder if a 486 could do this kind of stuff (a dx 100 could, but I think the dx33s where current when this was being done). All I can say is, what the heck happened?
It was ported to Intel in the 486 era, but it didn't really become practical to run until the Pentium 2. Ran pretty well on my AMD K6-350, if I recall correctly. Supposed to scream on Athlons.
In addition to Intel, it was ported, and sold, to run on Sun Sparc workstations and HP PA-RISC workstations.
I've read a bit of the history (I hear those MO drives they Next Stations ran off of were kinda buggy), but this is big enough stuff that they should have been able to get through a few lean years and sell the technology....
It wasn't the stations that had the Optical drive, it was the cube. That was the machine that got really expensive, when loaded up with a NeXTDimension color graphics card, big hard disks, and lots of RAM. The Optical was dropped before very long, and the Cube just shipped with a floppy drive. I think the Turbo Cube (33 MHz) couldn't even connect to the optical drive.
What happend to NeXT is (roughly) this:
First, customers realized they didn't so much want the hardware, they wanted the operating system. So NeXT dropped hardware and started doing their OS for other peoples' hardware.
Second, customers realized it wasn't so much the operating system they wanted, it was the development tools. So NeXT came up with a way to run the development tools on NT. And they had their WebObjects product, which let people use NeXT development tools to do web apps. So they de-emphasized the OS.
Then Apple bought them. The dev tools for NT were de-emphasized, except as a way to do WebObjects development. The OS was refreshed and updated, a process which continues.
Jonathan Hendry
" NeXT, for example, had nothing like Quartz. Quartz is largely informed by the word Bill Atkinson did on QuickDraw in the early 1980s."
Um, no. NeXT had Display Postscript. Quartz is much closer to that than to QuickDraw.
If you wait until Tiger is released, and until new Macs are shipping with it, then you won't have to pay the $129 for the new OS.
Plus, if you buy a Mac before Tiger comes out, you won't quite know how well it will take advantage of Tiger features like CoreImage.
If Tiger comes out first, you can check the hardware requirements for various features before buying, so that you won't be disappointed.
He's been locked into this series, in some directorial equivalent of typecasting.
So in a passive-aggressive attempt at getting even with the damn fanboys who won't let him be, he's self-sabotaging the movies.
and the whole Amidala giving birth to Luke and Leia thing
... Can't be!... Nooo! Luke and Leia aren't Jar-jar's children!! Nooo!"
And the world waits to find out if the caesarian is done with a lightsaber. Or the episiotomy, if they go that way.
You know what'd rock? If the babies pop out and they're half-Gungan.
That'd rock. You'd hear fanboy heads explode all over the world. "Noooo! That's not