This document discusses operating systems that Apple has created in the past, and many that it tried to create. Through this discussion, we will come across several technologies the confluence of which eventually led to Mac OS X. An important goal of the discussion is to better understand the reasons, and if possible, the rationale behind Mac OS X and its important components. This, in turn, will be helpful in understanding and appreciating the system as it is today.
A History of Apple’s Operating Systems
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Very well written.
It’s too bad that it took Apple this long
to recreate something like the AUX.
$500 million for Be OS? refusing $125 million offer from Apple, and accepting $12 million from Palm?
Is JLG nuts?
the greatest enemy of BeInc was JLG. — he should have taken the $100 million offer and walked.
apple would have beos and we would be in a better world today.
Times were different back then. When Apple was giving JLG $125M it was worth as much and Be HAD money in the bank, so they snobbed Apple’s offer.
When JLG accepted only $12M from Palm, Be didn’t have anything left, even that mere $12M was “good enough”.
But yeah, JLG DID wrong when he asked for $500M from Apple plus the CEO chair.
MacOS X is a very manyfaceted OS. I don’t think that if they had adopted BeOS as its base, it would have come as a result a cleaner and more coeherent OS.
If you browse the headers of Dano, you can see that many aspects of the Interface Kit were under revision: Be tried with BeOS 6 to implement the Model-View-Controller paradigm, a layout manager, a decent table widget…
Cocoa was fairly more mature.
apple would have beos and we would be in a better world today
Hmm..Better world? I guess. A PPC-only BeOS is better than no BeOS at all, like it is now. x86 users would have been left in the dust, so it wouldn’t have been a better world for them at least.
I see nothing wrong with Apple’s move. Whether anyone wants to argue with whether BeOS was superior or not, NeXT, as a company, offered a helluva lot more.
BeOS was much furthur ahead of it’s time than NeXT, no offense folks. The beauty of Be was that they cut out a lot of the fat in UNIX by stripping it down. Combine that with a full C++ api (which like it or not is more accepted than obj-c) and a heavily multithreaded kernel, and database file system is just staggering compared to NeXT’s offering.
Be wasn’t perfect though. It didn’t have DisplayPS, nor was it multiuser. But my god. When I installed Be on my 333mhz celeron w/ 128mb of ram, i crapped my pants. It’s like magically turning a 333 into a 700mhz. And what was the #1 thing i noticed? responsiveness. OS X still has this problem. I can have 20 apps running at once in os x and the entire interface starts to slow down to the point of frustration. And that’s with an 867 g4 and 640 ram. That much power shouldn’t ever have issues keeping up with a simple GUI event loop.
it’s my understanding that apple bought next for $420 million, and spent as much as another $500 million R&D to make mac os x what it is today.
let’s suppose that apple bought be inc for $120 million.
that means apple could have spent another $420-120=$300 million on R&D improving BeOS, and Sheepshaver allowed MacOS compatibility. plus another $500 million and several years?
apple would be untouchable in its chosen niche.
Where the best place to download it?
Apple seems to have always had some difficulty with R&D, fortunately they always seem to able to turn it right again at the end. For a humerous illustration of this check out ‘Quick hide in the closet’ ( http://www.folklore.org/StoryView.py?project=Macintosh&story=Hide_U… ) and other great Mac-stories on Folklore.org
Eugenia,
your husband is an ex-be engineer right?
Why didn’t Be/JLG consider the possibility that Apple would acquire rival OS NEXT?
Did it occur to JLG/Be that if they ask for too much money, Apple would go with NEXTSTEP?
Did it occur to JLG/Be that they are better off with $125 million, then being rejected by apple and apple acquiring NEXT?
Did it ever occur to Be/JLG that apple might go with another OS, so they should accept Apple’s offer?
personally, i would initiate a shareholder lawsuit against Be for this oversight!
“Why didn’t Be/JLG consider the possibility that Apple would acquire rival OS NEXT?
Did it occur to JLG/Be that if they ask for too much money, Apple would go with NEXTSTEP?
Did it occur to JLG/Be that they are better off with $125 million, then being rejected by apple and apple acquiring NEXT?
Did it ever occur to Be/JLG that apple might go with another OS, so they should accept Apple’s offer?
personally, i would initiate a shareholder lawsuit against Be for this oversight!”
They probably didn’t think Jobs would ever go back to Apple after what happened before.
What did Gasse ever do right?
>your husband is an ex-be engineer right?
Yes, so? JBQ was not working at Be when Apple did the offer.
People, stop harassing JLG. Without him, there would be NO BeOS. In 1996 when the Apple talks took place, it seemed that BeOS was poised to grow bigger than Apple becuase it had a much better OS. Even in early 1999, BeOS was the rising star, and after the IPO its share price went from $6.00 launch to over $25 dollars for a while, even topping $39 at one stage in late 1999. BeInc at this point in time was worth around $600M. At this point, noone accoused JLG of being a bad CEO, on the contrary.
The investors of BeInc forced the focus shift. JLG did the most he could do, and found a way to give BeOS away for free (R5). BeOS would have been dead long ago if it weren’t for R5 PE which extended its life by 2-4 years. So in effect, BeOS went out with a bang, given away for free.
Blame investors and owners of BeInc for the demise. Not JLG.
no newton?
Zenja
was it within JLG’s discretion to sell Be Inc to apple for $125 million? i do not personally know, but if the answer is yes, then JLG should be blamed.
let’s analyze this within Game Theory.
2 choices – choice 1 Be sells to Apple for $125 million,
choice 2 Be does not sell to Apple for $125 million.
Choice 1 is self-explanatory, Be pockets $125 million.
Choice 2, if Be does not sell to Apple for $125 million,
we can ask ourselves 2 questions
question 1 WHO would be willing and able to pay $125 million for Apple?
Not NEXT. Not M$. Not intel. Not IBM. Not Oracle, Sun or SGI. Not Sony. Not when you have free Linux.
Apple was willing to pay $125 million b/c of its unique problem -its copland project failed.
Question 2 Could Be generate enough revenue from OS sales?
if Be goes it alone, could it generate enough revenue? Unlikely with competition from free linux, m$, and Apple-Next.
If Apple goes with Next, and does not buy Be, could be develop for Powerpc? unlikely due to steve jobs. For Intel? Unlikely due to m$ Next OS was not commericially successful, nor Amiga.
From the standpoint of Gametheory, any competent CEO should see that Choice 1 is the clear Nash equilibrium.
BeInc stock in 2nd half of 1999 put BeInc at over $600M. BeInc investors made a mint. Are you trying to say that $125M is better than $600M?
BeInc stock in 2nd half of 1999 put BeInc at over $600M. BeInc investors made a mint. Are you trying to say that $125M is better than $600M?
Let’s just say history has proven that $125M was, in fact, better than $600M…Odd as it seems.
If JLG had realized that Be’s future was rather limited in options, he would have seen that. Hell, even I saw that, and I’m nobody. But what did he do? He took risks and failed, but I won’t blame him — He loved the product.
But every 3 months it seemed like he’d change strategy. And in the end, he realized Open Source had gobbled up most of the free space his commercial “alternative” could move around in…Linux killed Be. Not Microsoft.
but how many investors cashed out when be was worth $600 million?
if i recall correctly, the only reason why it was valued that much was due to internet/feverish speculation that be was going to be bought out by redhat
Be wasn’t perfect though. It didn’t have DisplayPS, nor was it….
I’ve noticed just a couple of windowing systems like that. There was also NeWS, OS X does PDF, and you say NeXT. (Perhaps(?) relatedly, Microsoft’s going to XAML.)
Anyone know what the advantage is in this? Some Mac sites said the PDF intergration was good because you could “save your desktop/document as a PDF” and wysiwyg.
That hardly sounds like the real reason. Anyone feel like educating me?
“If you browse the headers of Dano, you can see that many aspects of the Interface Kit were under revision: Be tried with BeOS 6 to implement the Model-View-Controller paradigm, a layout manager, a decent table widget…”
How about you not talk about things you know nothing about? The code you saw was very early work on a next-generation “internet appliance” OS; it had nothing to do with creating a new version of BeOS. And the features you cite are not at all the interesting parts of what was being done — much of which was ultimately integrated into Palm OS Cobalt.
Dianne wrote:
How about you not talk about things you know nothing about? The code you saw was very early work
I was not referring to the leaked source code, but only to the headers which have been distributed on Dano.
Like Jobs said to Apple, many components of BeOS were still in a prototypal state while OpenSTEP’s APIs were more advanced and already praised by the enterprise developers. It wasn’t even to compare the available visual IDEs: BeOS officially lacked simply one and you had to indicate manually the coordinates of the widgets due to missing layout managers and design tools (AppSketcher by BeatWare to be adopted by Be was dropped just a few week after the BeIS focus shift.
The 3D Kit was a false step, so the printing subsystem, the buggy Media Kit…
Anyone know what the advantage is in this? Some Mac sites said the PDF intergration was good because you could “save your desktop/document as a PDF” and wysiwyg.
That hardly sounds like the real reason. Anyone feel like educating me?
It’s more than being able to save as PDF. And WYSIWYG (or: what you see is what you print) is very important, considering Apple’s target audience.
It’s basically the same as what NeXTStep had, but Adobe wanted Apple to switch to PDF instead of PS (for various reasons I never found out about).
Postscript is only meant to decribe a page. PDF does that, but goes a little further and can hold more types of information (whether that be looks, fonts, images, hyperlinks, movies, etc. — one reason for being more suited for a system level display technology).
Besides that, we’re talking “beauty” here. So take a look at what’s possible with imaging in, say, Adobe Illustrator, whether that be text rendering or graphics. One thing you’ll notice (out of many) that’s worth mentioning is scalability: Which isn’t possible with bitmaps. All of that functionality, usually reserved for apps like Illustrator, is built-in the OS itself. Aqua isn’t even possible without it.
Anyways, kind of basic explanation, but hope it helps.
Just a little clarification:
[i]Which isn’t possible with bitmaps.[i]
“bitmapping” being what most windowing systems use, as opposed to scalable vector imaging, a la Illustrator/PDF.
That was a well written article on Apple history. It was quite thorough in covering the influences contributing to the Apple OS evolution.
This fellow knows how to write.
More likely is that Apple wanted Jobs back. The price was paying way too much for NeXT and rejecting BeOS regardless of technology.
personally, i would initiate a shareholder lawsuit against Be for this oversight!”
Nice, PantherPPC. Way to show the world that you like commenting on things about which you’re obviously ignorant. Be was a private company back in the “be bought by Apple days”, and remained that way for long after. It only went public a couple of years from its demise. The “shareholders” had no say in the deal, because there weren’t any.
As for all the people saying that Apple should have gone with Be, consider the following:
1. Steve Jobs is without a doubt the most valuable asset Apple got from NeXT. Many people hate Steve and that’s fine, but anyone who knows anything about Apple in the mid-nineties knows that they were above all a company without leadership. There certainly was no shortage of brilliant people there, but infighting, lack of direction, mismanagement, and the lack of focus was killing them. Steve streamlined the company to a huge degree and brought back one incredibly important thing: focus. Love Apple of hate them, they are now an incredibly focused company. Remember Apple’s product lines back then? Look at them now. JLG had many great qualities, but I doubt he could have done what Steve did: turn a huge organization completely around. That’s leadership, and Apple needed it more than any OS technology.
2. This deal was back when the BeOS was what, DR8 or DR9? It was not a mature OS at that time. Hell, it still wasn’t, years later, when Be went under. NeXT, by contrast, was mature and better yet, it was proven.
3. The BeOS was a damned nice little OS, but it wouldn’t have done a ton to gain Macs marketshare. Being Unix-based gives Apple many, many new opportunities that previously weren’t there. People actually take the operating system seriously for important tasks, and many legacy (yet powerful) systems can be easily migrated to it. The BeOS Mac would have been more of the same type of Mac, still alienated from the world, except now more stable and with a more modern foundation. That’s just not enough. Proven and compatible are better.
4. It’s not like every piece of the BeOS was impeccably designed, either. Because of OS X’s linking/binding, Apple has been able to migrate to three different versions of GCC now (that I count; maybe even more). Because of the BeOS’s C++ ABI, moving away from gcc 2.95 — which, by today’s standards, is a rather subpar compiler — meant huge binary compatibility issues.
5. Sure, the BeOS had a very nice and fast drawing model. The problem, of course, is that it was extremely simplistic. OS X’s model is slower, but way more advanced. Remember the way Netpositive and Soundplay would show redraw errors while resizing? You don’t get things like that with Quartz. Moreover, Apple is a hardware company. If having pretty graphics means enough to you that you’ll buy a new machine for them, all the better for Apple. It’s in their best interest to produce value that’s worth your money, not just “old technology that’s really fast!!!”
Be bet the farm because they thought Apple had no other choice. I don’t blame them for that. But Apple is in a much, much better place now because they bought NeXT (both the technology, which is excellent, but moreover, Steve) than they could possibly have been in had they purchased Be. Sure, “OS 10 built on Be technology” probably would run faster, but would it be a superior OS? Not a chance. OS X has more vision, more focus, more compatibility, more proven software, and more advancement. It’s not just “the same old alienated Mac, but now with a better kernel and memory architecture”; it’s a somewhat radical departure that keeps most of the Mac charm and ease of use while being beautiful, and while giving Apple opportunities they never would have had otherwise.
More likely is that Apple wanted Jobs back. The price was paying way too much for NeXT and rejecting BeOS regardless of technology.
Regardless of technology? BeOS was impressive, but it’s not like NeXTStep was DOS or something.
Jobs did not have much control until later on, and neither did he want it, so I doubt they paid so much money just to get him. And besides, I bet if they knew what they know now, and what was going to happen, they *definitely* would have chosen Be — Lol, the moment Jobs was given power he ousted all of their asses.
Apple was getting more than just an Operating System in the deal — OpenStep API, WebObjects, a much larger user/developer community, an entry way into the Enterprise market (of course, they haven’t really taken advantage of that aspect just yet).
Mach work was already being done with MkLinux, so it fit there as well (and a little sidenote: Once NeXT was purchased, Jobs even asked Linus to join the fray and help with Mach/Linux/etc..Of course Linus rejected the offer — I suppose if he had joined, most likely they would have scrapped BSD and Mac OS X would now be running on Mach/Linux).
What did Be have? A great OS, albeit incomplete, and without all the perks. And by “perks”, I mean it had none.
This article is truly a gem.
If only people (me included) would have enough time and effort to put ideas together to write pieces like these – uh, nobody wouldn’t have boring Sundays at least
Even in early 1999, BeOS was the rising star, and after the IPO its share price went from $6.00 launch to over $25 dollars for a while, even topping $39 at one stage in late 1999. BeInc at this point in time was worth around $600M. At this point, noone accoused JLG of being a bad CEO, on the contrary.
Notice the rise of Be – they were riding the dotcom wave. Nothing great about JLG or about Be Inc.’s business model and plans. Getting that amount of financing around that time, plus the fact JLG’s history in Apple – it wasn’t a sheer business genius that cause this rise.
Besides, JLG started a company. That company created one great product after another that in theory, should sell. What happen? Nobody bought it because there’s no use of a super-responsive desktop if most applications that can be found on Macintosh can’t be found on BeOS. Why buy a stylish, subsidized BeBox when it is as useful as a car without a steering wheel?
JLG could have prevented this, especially with the amount of connections he should have (considering his past with Apple). A lot of developers were unhappy with Apple in the early 90s because the promised Mac OS-replacement was shelfted (Copland). A lot of money was invested by third-party developers to porting their key applications to it. In fact, this contributed a lot to the reluctance of developers to go full steam ahead with development for OS X until after it is released.
JLG could have used that to his advantage and market BeOS to these third party developers as a Mac OS replacement.
The investors of BeInc forced the focus shift. JLG did the most he could do, and found a way to give BeOS away for free (R5).
The investors was right in this regard. Had the focus shift focused on another market instead of the overhyped underperforming IA market, things probably would have worked out. I don’t know whose fault is it, but choosing the IA market (even if it boomed instead of crashing and burning) was very risky considering that this is Be’s last chance.
Free BeOS couldn’t have saved Be Inc – Be wasn’t making money and no third-party developers were taking the bait.
I didn’t know about some of those projects. Good article. Thanks!
I’d still love to see a demo of the OS that would have been (Copland?)… Apple is improving OSX for sure, but it still isn’t elegant beneath the Desktop…
“TalOS”
Now there’s something I would like to have seen more of. Same with OpenDoc. Apple does some nice tech. Getting it to market is another matter.
I’d still love to see a demo of the OS that would have been (Copland?)…
Imagine a Mac OS 8 that crashed a lot..Wait, that’s not stretching the imagination. I mean crashed *even more*.
Apple is improving OSX for sure, but it still isn’t elegant beneath the Desktop…
Care to explain?
POSIX/UNIX, microkernel..There are few ideas more elegant. Even BeOS followed the same model. UNIX has outlived every idea out there, because it is elegant, and will probably outlive all of us too.
Be bet the farm because they thought Apple had no other choice.
uh did JLG seriously think Apple did not have a choice?
What about NEXT?
how could something be unixy and not be multi-user? isn’t that missing the point? weren’t both Unix and Multics both designed from the bottom up to be multi-user environments?
While looking up Multics I came across
We went to lunch afterward, and I remarked to Dennis that easily half the code I was writing in Multics was error recovery code. He said, “We left all that stuff out. If there’s an error, we have this routine called panic, and when it is called, the machine crashes, and you holler down the hall, ‘Hey, reboot it.'”
back in the day they called it unix… now they call it windows 95! http://www.multicians.org/unix.html
how could something be unixy and not be multi-user? isn’t that missing the point? weren’t both Unix and Multics both designed from the bottom up to be multi-user environments?
BeOS just didn’t use logins. You logged in as “guest”. Multi-user wasn’t implemented only because it wasn’t marketed that way, it was nothing in the design.
I’ve been trying to access this article for two days now and I keep getting a timed out error : (
Re mac os 8 crashing alot. 8.6 never crashed on me. ( I DO mean never). As crazy as it sounds it was more solid than XP, X.x, 9, redhat, etc.
Thats not to say apps didnt crash though….
Thats not to say apps didnt crash though….
Heh…OK, that was funny.
One of the best articles i’ve read in monts. Extremely informative, delightfully well written.
“Nice, PantherPPC. Way to show the world that you like commenting on things about which you’re obviously ignorant. Be was a private company back in the “be bought by Apple days”, and remained that way for long after. It only went public a couple of years from its demise. The “shareholders” had no say in the deal, because there weren’t any.”
The comment you seem to think I made was the one I was responding to. I have no complaint about your response, only it was directed at me instead of Red Pillar. My comment was the Be Inc. probably didn’t think Jobs would want to go back to Apple, so they didn’t think there was much of a change of Apple buying NeXT.
Why did apple not try to buy Be again when it was worth $12 mill?
Even that late in the day Be had a fair few features that could/should be integrated into Apple’s OS.
“Even that late in the day Be had a fair few features that could/should be integrated into Apple’s OS.”
Why do that when you can pick up some of the employee’s for a song?
“Apple does some nice tech. Getting it to market is another matter.”
Some Taligent developments were licensed to Sun and can be found in Java 2D and i18n (ResourceBundle) source, IIRC.
Not TalOS, but at least some of their work made it to market. Probably a fair amount was absorbed into other IBM products, too.
It has timed out for me once and again since it came out. Is there a mirror or something? Google has only cached the first page…
Be inc. was able to win the lawsuit against MS for the way MS bullied Toshiba into not loading Be’s software on their hard drives? But what’s happened has happened and another potentially great OS for x86 platform has bitten the dust. What a shame…