Home
About
Apple
Career
Experiences
General
Graphics
Hardware
History
Humor
Interface
Networking
OS
Opinion
Politics
Programming
Quotes
Reviews
Security
Software
Sound
Thought
Web
Cheap International
Airfare Online
Wachovia online banking
Get Free Coupons Online
Finding the perfect
discount hot tub
Payday Loans
Stock Trading Online
Stuffed Animals
Smart Investing Online
|
|
History of Taligent
Not everything is pretty in Pink
By: David K. Every
|
Kind:
Created:
Size: |
Article
2002-09-27 00:00:00
13 KB |
|
|
 |
 aligent became the OS that never was. This is sad because it had such a promising future once.
Apple had seen the Smalltalk demos at Xerox PARC, as well as all the User Interface demos, and was sold on the concepts. Smalltalk was a pioneer of OOD (Object Oriented Design). Many people working at Apple had come from Xerox (or other places where interface and OOD mattered) so they were sold too. It is because of this that the entire computing industry changed to OOD.
          
Apple was so inspired by OOD (in early 80's) that on the Lisa they created "Clascal" (Pascal with Classes) and their LisaWorkshop programming framework. This soon evolved, and they started working with Nicholas Wirth (the creator of Pascal) to create "Object Pascal", and they evolved the framework into what became MacApp, one of the first real Application Frameworks on a microcomputer.
It was around this time that Steve Jobs was leaving Apple ('86), where he was going to go create his own company that extended these OOD concepts (among other things) in order to create NeXT. Like Apple they needed to create their own language and framework in order to do OOD; and actually partner with someone else who had a partially complete solution and academic reputation. They chose Objective-C (since C++ was still just a dirty thought in someone's mind).
At first NeXT was going to be a hardware company (like Apple). They wrote their OS and started using their Object Oriented Design knowledge (from the Mac and elsewhere) to create an Object Oriented Operating System. But the hardware market wasn't working well for them and they saw that Microsoft and others were succeeding in software (without a hardware platform). Because of this NeXT later dropped the hardware and became a software only company by the early 90's.
Apple wasn't sitting on it's laurels either and they had evolved MacApp a lot over the next few iterations. The industry started catching on to OOD. Even Microsoft was using it and had created their copy of MacApp: MFC (Microsoft Foundation Classes). But the industry had headed towards a different language (C++), and Apple had to migrate MacApp to the new language. Part of the problem of being a pioneer is that sometimes you find out that you're too far out ahead, and you either have to wait for others to catch up or adjust to their course (if you want their support).
Nothing inside of Apple (or inside of any engineering organization) happens without politics or infighting. And the future of Mac OS itself was being hotly debated (roughly the late '80s). Some wanted to keep tweaking and revising the Mac OS (System 6) and bring it forward. Others (from the OOD or MacApp side of the fence) felt that they needed a new and more modern kernel, a more Object Oriented Operating System, and some radical advances elsewhere.
The Vice President in charge of OS's was tracking the two competing projects using Post-It notes: Pink for the OOD OS, and Blue for the incremental-improvement-based OS. Blue became System 7 and Pink became the code name for a new OOD-OS team. The teams (and projects) were split and management chose to staff both projects.
Since Pink was a long-term project it wasn't too heavily staffed but it was more than just a research project and it was creating some code. But an OOD-OS was also a huge job and was going to take many years. When NeXT broke away from Apple they said they would have one in a couple years but it took them closer to 5 years to deliver; creating an Operating System is not a small job.
       
Apple wasn't the only one looking at Object Oriented Operating Systems. NeXT was starting to gain a little success. But more than that, the hype about what they were going to be able to do was beginning to get the attention of the Big Boys. In at least a few ways NeXTSTEP (later OpenStep) was easier to write software for than other offerings at the time. Since Application frameworks and OOD was making it easier to write Applications, why not apply that to Operating Systems? That potential momentum scared the competition.
Plus there was a lot of hype surrounding OOD in general. It had become "vogue" to start talking about Classes and Objects (Apple had only been about 5 years too early on that one). A friend had joked during this time that Objects had become the new oat bran (another pop-culture trend of the time). Phrases like "Now with Objects" started appearing in descriptions of people's Apps, languages, or tools as if OOD alone somehow made their product better (which of course it doesn't); crappy software in OOD is just as crappy as crappy software everywhere else. But hype is hype.
Apple, Microsoft, and IBM all had to start following the Object-Oriented Operating System path as a "cover your butt" strategy. Even Sun was following it as people found out later with Java (OAK), though not in their primary market. Apple had staffed up Pink a lot more and was actually taking these concepts seriously. They had many other OOD projects they were doing as well; Dylan, Bedrock, MacApp, Newton, and others come to mind. IBM certainly wasn't going to be caught with their pants down and they had started their own OOD projects. But IBM was traditionally very slow and bureaucratic. They had blown most of their opportunities in Microcomputers and were starting to realize this.
      
Apple and IBM started talking. Some of it was about IBM buying Apple, which was a deal that had previously fallen through. But they realized that Microsoft and Intel were bigger threats to them than they were to each other. So they made nice and everything was brought to the table for discussion. They started AIM for the hardware (Apple-IBM-Motorola alliance). Motorola was brought into this agreement by Apple because Apple had such a long standing working relationship with them and because it short-cut some redevelopment. But IBM and Apple also started talking about software.
When IBM and Apple started talking software the topic of Pink came up along with IBM's own pilot projects and visions. Their goals were in alignment and since they were partnering in hardware, why not in software as well?
In order to facilitate that partnership they decided to create a whole new company (a few of them actually) to share work between the two organizations. One of those companies was Taligent.
Taligent (via Pink) already had a lot of frameworks working, and a lot of the OS as well. They'd borrowed a little from QuickDraw GX for imaging (conceptually), QuickTime for streaming (licensing), MacApp (for some background knowledge and people: little actual code was used), and so on. IBM did bring a lot of people and talent to the table; they just didn't bring much code. There were rumors that the basic OS borrowed from AIX (IBM's Unix) and possibly from OS/2. I know less about that part but I assume that the lower levels got a lot more from IBM. But Taligent took off with a bang and sadly died years later with hardly a whimper.
 
Microsoft had its eye on NeXT and was concerned; NeXT had partnered with Canon and Sun and was starting to look like more of threat. Then the Apple-IBM deal (Taligent) put them in a panic. It was IBM's name, money, and reputation that had made Microsoft in the first place, and borrowed ideas (from Apple) made up most of Microsoft's R&D. Putting the Apple's R&D and IBM's money and reputation together was Microsoft's worst nightmare.
Microsoft responded how they usually do: with hype and empty promises. They claimed they had been secretly working on an Object Oriented Operating System for years (and quickly threw a few coders in a back room to work on it). This Operating System would be fantastic and virtually replace Windows as people knew it. It was entirely OOD, built on NT technology (which was just coming out at the time), and only a couple of years from release (much closer than anything Taligent could deliver). It would run all Windows software but be completely OOD. It would run on everything and use less memory than anything else (defying the laws of physics and nature), so people should suspend any disbelief or mistrust in Microsoft.
Of course years later when Taligent died, so did Cairo (if not before). Eventually the truth about Cairo leaked out. It seems that there were some projects that were lightly OOD (like a new document model and some other parts and frameworks) but Cairo was really nothing but marketing vapor and a new name on all the same old Microsoft technologies - a common practice for them; if you rename it then it becomes something "new".
The problem with Taligent was not the technologies involved but rather the politics of big organizations. Once it was spun off it was basically dead. The company never got really focused and the internal politics of both IBM and Apple meant that it was doomed. Even though Taligent got a lot of code working it wouldn't ever be accepted by either company.
Apple wasn't committed to a technology that was not invented in-house and which Apple didn't control. Why would Apple managers bet on someone else's projects when they were fighting for their own projects and people? Once Taligent was spun off, Taligent was out of the internal loop (infinite loop, or the Apple Campus) and so was its ability to influence decisions inside of Apple. Taligent's people couldn't even defend themselves against bald-faced lies or misinformation about the project that are common in back-biting corporate politics. Roughly the same thing happened with IBM, if not worse. But IBM people were at least used to that (one hand not knowing what the other was doing and throwing money at it anyway). That was the normal operating procedure there. Funding could go on forever for a "produce nothing" kind of project at IBM. Apple started feeling pressure and had no such inclination towards paying for nothing. So, while it was a neat project with good technologies and concepts, and while Apple and IBM would throw money at it for a while (just as insurance), neither was committed to it.
Because Taligent had two different companies pulling in different directions from the outside and two very different corporate cultures to try to merge internally, they did not progress very fast at the start. There is a shakedown period while a new culture is made and people and groups all figure out the pecking orders and processes. Combine that with the usual issues with software engineering that always take much larger than anyone considers, and the result is slow progress. There were different directions and goals. In the end TalOS was basically designed by a committee, which is never a pretty sight.
By the time Taligent released something it wouldn't even run on MacOS, and IBM had taken over most of the control -- but only so IBM could kill it in their own special way (death by bureaucracy). IBM did throw some support behind TalOS, but not much -- Apple even less so. It was stillborn, even if it had some good technology. Later, IBM bought Apple out and gobbled up the remains of the company. Both got some code that they wouldn't really touch and they went their separate ways.
Sadly, even when Apple was having problems, they wouldn't seriously consider what they had in Taligent; design, people, or code. IBM put the remains into CommonPoint Frameworks and then into their VisualAge for C++ OpenClass frameworks. In a huge twist of fate, Microsoft then licensed some of those patents, which all belonged to Apple. IBM also used some of the imaging stuff (design) in their Java graphics engines, which I think made it back to Sun and into the Java that we know today. It's funny how some code lives on and on and on.
 
NeXT had been progressing but they did not have the muscle to really market what they had or distribute it on enough platforms to entice lots of development. Canon and Sun both dropped the ball on their part of the agreements with NeXT. To hear them tell it, though, at least some of that was because NeXT (and it's founder) wasn't always easy to work with. And not everything was rosy on the product side; there were some serious quality issues. But being a smaller company let NeXT stay focused, and being free(er) of in-fighting has its advantages as well.
In all honesty, Taligent was both better and worse than NeXT. TalOS had a better and more document-centric UI in many ways. It borrowed some concepts (actually parallel development) with OpenDoc and had an interesting "People, Places, and Things" interface which was far more innovative than big icons, display postscript, and a file browser for UNIX (NeXT). But changes have a cost and those changes confused many people; a fact which hurt adoption. TalOS tried to build on C++ as the foundation; but in order to fix C++'s FBC issue (Fragile Base classes -- which just means you can't change the framework over time), they had to hack in padded methods and some other things, very similar to what Be had to do (selling the future for the present). This hack is not a great fix long-term and was going to bite programs and programmers later on. It was not nearly as good a solution as ObjectiveC. So there were lots of good and bad things, and it doesn't matter because it is long gone.
Of course we all know that Apple bought NeXT and ended up with something that they would have had years before if they'd just followed through on their own plans. But sometimes that's the way things work out. In many ways NeXT is more mature and feature rich than Taligent ever got. Plus it wasn't designed by committee and so at least it had some vision. Apple ended up (through acquiring NeXT) with a better OS than they were shooting for in-house and now they have complete control - and it cost them less to buy NeXT than their own development of Taligent cost them. Live and learn; or in this case, die and learn. Either way, you learn.
 Format for Printing  Mail
|
|
|
|
