Apple has to fight off three of the biggest PC companies, sick of the security bugs that plague Windows, wanting to license the Mac operating system. My take: Embrace it my child, don’t fight it off.
Steve Jobs: PC makers want to license Mac OS X
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
196 Comments
2005-02-15 1:31 am
Anonymous
Not really.
By what standard ? What’s your measure of “impressive” ? Was the migration from 68k to PPC for MacOS “impressive” ? How about the level of compatibility in OS X’s Classic Environment ?
This is an ugly hack at the very least, not a particular feature to be proud of. Uh gross. This isn’t impressing me actually it is making my stomach turn.
If you call trying to fix app bugs in the OS as features sorry we speak different languages and have different meanings of impressive .
When it stops me (as a customer) have to spend hundreds (if not thousands) of dollars on new software and/or hardware, you better bloody well believe it’s a great feature.
When that result is achieved without having a significant negative impact on the system functioning well otherwise, you better bloody well believe it’s impressive.
Reasoning like yours is why Linux and much other OSS software is having such trouble actually getting into widespread use. To much attention paid to academic trivia, not enough attention paid to customers.
Conversely, it’s the attention Microsoft pays to these sort of fundamental customer concerns that keeps them a popular choice *despite* their technical flaws.
Memory protection and multitasking are not paticularly modern concepts.
They were to Windows and MacOS in the early ’90s (and still in the late ’90s, for MacOS).
Exposing internal implementation as a part of the API is bad design to begin with.
The example wasn’t “exposing internal implementation”, it was merely retaining compatibility with it.
Interesting you feel that something OSS explicitly encourages is “bad design”, I’ll have to remember that 
.
Adding a preemptitive scheduler or using the hardware mmu for prtotection is orthogonal to upperlevel apis at least on most well designed OSes.
The legacy support Windows and MacOS have/had to provide goes a hell of a lot lower than “upperlevel APIs”. It goes all the way down to supporting direct hardware access.
Additionally, the reason they’re needed to provide that level of legacy support has mostly to do – surprise, surprise – with application problems rather than OS problems.
Pointing out a hack to make one paritcular App work is not helping your cause in making me think that Microsoft actually designed windows well.
I have never tried to make the case that Windows 9x was well designed. All I’ve said is that it was an impressive achievement. So was the 68k emulator in the first PPC versions of MacOS, for similar reasons (legacy support).
2005-02-15 2:10 am
Anonymous
By what standard ? What’s your measure of “impressive” ? Was the migration from 68k to PPC for MacOS “impressive” ? How about the level of compatibility in OS X’s Classic Environment ?
Whoa…. Are we talking about windows 95’s feature set or O S migrations over the years???
You need to stay on topic, or atleast the sub topic of if MacOS classic ever had preemptitive multitasking, the answer is yes. Everything else is irrelevant.
When it stops me (as a customer) have to spend hundreds (if not thousands) of dollars on new software and/or hardware, you better bloody well believe it’s a great feature.
I don’t give a rats ass what it was supposed to do but what it is, is ahorrible hack.
When that result is achieved without having a significant negative impact on the system functioning well otherwise, you better bloody well believe it’s impressive.
It’s still a hack. Hacks are unimpressive, by virtue of being inelgant solutions. Sorry still not impressed
Reasoning like yours is why Linux and much other OSS software is having such trouble actually getting into widespread use. To much attention paid to academic trivia, not enough attention paid to customers.
I thought billions of dollars and an already exisitng monopoly was preventing this. And things like special contracts with vendors barring them from selling bare machines or alternative Oses did it as well. Before you say
that isn’t true read the findings of facts and also realize that things today are different becuase of it.
Conversely, it’s the attention Microsoft pays to these sort of fundamental customer concerns that keeps them a popular choice *despite* their technical flaws.
No ubiquity and history makes them the popular choice.
They were to Windows and MacOS in the early ’90s (and still in the late ’90s, for MacOS).
No sorry it still isn’t a mordern concept. The fact that two consumer grade oses didn’t have it doesn’t make it any more mordern.
The example wasn’t “exposing internal implementation”, it was merely retaining compatibility with it.
Hunh…. you lost me here. Which example? I am talking about implementing the two features we were talking about.,
Interesting you feel that something OSS explicitly encourages is “bad design”, I’ll have to remember that .
What does this have to do with OpenSource Software? this is basic software design and is applicable reagrdless of access to source code. In fact, it is more applciable to enforce that APIs not expose implmentation details in OpenSource software because developers tend to take shortcuts( read hacks) if they do. Like the one MS did in the example.
The legacy support Windows and MacOS have/had to provide goes a hell of a lot lower than “upperlevel APIs”. It goes all the way down to supporting direct hardware access.
WTF are you talking about. Give me some examples of said APIs.
Additionally, the reason they’re needed to provide that level of legacy support has mostly to do – surprise, surprise – with application problems rather than OS problems.
Here we go again….
I have never tried to make the case that Windows 9x was well designed. All I’ve said is that it was an impressive achievement. So was the 68k emulator in the first PPC versions of MacOS, for similar reasons (legacy support).
Sorry still not impressed.
2005-02-15 3:27 am
2005-02-15 7:16 am
2005-02-15 7:40 am
2005-02-15 7:49 am
2005-02-15 9:19 am
2005-02-15 9:30 am
2005-02-15 4:25 pm
2005-02-15 4:34 pm
2005-02-15 9:28 pm
Fact of the matter is that questioning is the first step to breaking MS hold on the PC market. HP has already got some pretty good deals with Apple and iPod. For big companies like HP, they are big enough to pretty much do what they want…they don’t HAVE to be afraid of MS anymore… Fact of the matter is that MS hasn’t delivered anything “WOWING” for the big PC makers like HP in a long time… Things like XBox only prove that MS is out to “eat their lunch” by creating technologies they completely control then sell to the lowest bidder… Look at WMA for another example… they tried letting Dell and HP “compete” in the music player market…but MS is more than willing to sell out their bigger partners for newcommers that do flashier marketing [i.e. selling out musicmatch for napster]
Apple is proving that their name can sell stuff again. They’ve got the moxy to keep control and choose good partners too. The tipping point is almost here…. Linux has been a good “icebreaker” but Apple is the only other “real” competitor to MS in the Public’s mind. With Apple’s rise to popularity, attaching their name to more stuff could really break the MS hold on the PC industry.
There are a few distinct routes Apple could take with regard to x86 and licensing.
1) Sell Apple x86 machines.
I don’t really see that Apple has too much to gain here. Not much would really change and they would have to deal with the problem of binary incompatibility. The only real effect of this is that they would no longer be so dependent on IBM and Motorolla. However, the G5 is fast enough now that I don’t think that people avoid Apples because they are slow. People don’t buy Apple because of compatibility (Games etc.).
2) License OSX for PowerPC to other vendors.
This is exactly what Apple tried back in the 90s. It failed. To a degree, this could commoditize the PowerPC hardware and drive down prices, but I think the supply constraints on PowerPC processors would limit this effect. Again, not terribly interesting.
3) License OSX for x86.
Apple becomes an OS vendor and a direct competitor to Microsoft. This would basically kill their computer sales as they couldn’t compete on cost/volume with PC hardware vendors. Apple could retain a high end niche in computer sales, but even that would be hard to maintain. Some of the Apple magic (beautifully designed hardware) would definitely be lost.
– Apple could actually do really well money wise. They are already making a killing on higher volume iPod products. The margins in software are a lot better than hardware, so Apple would really clean up if this move significantly increased OSX sales volume (admittedly a big if).
– The biggest obstacle is Office. Microsoft probably wouldn’t be very quick to produce a version that runs on x86 OSX (insert antitrust wildcard here).
– Games are the second biggest obstacle. This is the only reason I currently don’t use Linux on the desktop, and it would apply to OSX as well. Getting some high profile games released wouldn’t be too hard. To get good game coverage, Apple would need to significantly increase OSX sales volume over their current OSX sales. “Chicken and Egg” problem here.
Now, I don’t think any of this will happen. Things will almost certainly continue as they are, but it is fun to speculate.
> How much needs to be ported? BSD and Darwin run on x86.
> Obviously Aqua, but the truth is that many core pieces of
> OS X (obviously KHTML, Darwin, Xorg…) are not tied
> directly to the architecture, right?
> I’m sure that there’s a ton of work to be done with
> hardware support, but I’d imagine that a lot of stuff
> wouldn’t be that hard using Linux drivers as a base.
see
http://www.softpear.org
Quote:
The SoftPear Project aims to create compatibility software between the IBM PC and the Apple Macintosh architecture.
With Mac OS X, an excellent operating system is available, unfortunately it runs only on Macintosh hardware. Therefore this project seeks to create compatibility layers to run Mac OS X on IBM PC hardware.
We do not develop an emulator for PowerPC Macintosh machines (like “PearPC”, or like “Basilisk” or “UAE” for other platforms), though; the project has more similarities to Digital’s “FX!32”, FreeBSD’s “Linux Binary Compatbility” and “WINE”.
Instead, Darwin/x86 or GNU/Linux will run on the PC, and the Mac OS X user interface, its libraries and all applications running on top of it will run on Darwin or GNU/Linux, using SoftPear’s compatibility layer.
This compatibility layer consists of an emulator/dynamic recompiler of user mode PowerPC code and a layer between PowerPC code and native x86 code that handles endianness issues.
Another layer can make x86 applications run on top of the PowerPC Mac OS X libraries, which in turn run on x86 hardware, so Mac OS X applications that have been developed for the PowerPC can be recompiled for the x86 CPU. In addition, all that knowlegde that will be gained by making Mac OS X run on the IBM PC architecture might help creating a reimplementation of parts and eventually all of Mac OS X under the GPL, providing a fully open source Mac OS X clone for other processors than just the PowerPC.
Do you honestly think this project will really ever get off the ground? Cmon, now… this will take YEARS and YEARS to complete even if they do manage to pull it off. And by that time, Apple should be way ahead of the game.
Realy hope we’ll get MacOS X for x86 soon..
Apple leads the market in directions and always has. They are not an industry leader and that is apparent by their lack of licensing the technologies they create.
Marketing leaders create and show other companies the possibilities.
Industry leaders provide the technology to enter new markets.
I doubt we’ll see a clone market or Apple licensing OS X to other vendors.
To do so would be against what apple is.
Apple should license to X86 before someone takes BSD and builds one from scratch, it is not like Apple can do things others can not do since Apple computer is run by the same breed of humanity as anyone else walking around on planet earth…
I can’t really see what offering OS X on x86 would buy Apple. They’d then be stuck with supporting 2 hardware platforms, and one of them (x86) comes with a lot of legacy baggage. Not only that, it puts 3rd party developers in the position of having to support two different platforms as well (which they probably wouldn’t do).
If they were going license, they’d be better off licensing the PPC version. This would give them the advantage of a.) making a larger variety of hardware available for the Mac platform and b.) they’d be able to leverage the current body of software available for Mac OS PPC.
While I recognize that Apple tried the cloning route before, it’s important to remember that now they’re a lot less dependent on Mac hardware sales for revenue now than they were then. Also, they’ve become quite a formidable software house, having considerably broadened their portfolio of application software. This is also another revenue stream for them.
Consider, as well, that margins are quickly getting squeezed out the hardware market, making this a less attractive revenue source going forward.
If they were going to transition themselves from a hardware shop to a software house, they’re probably in a much better position to do that now than they were with their first attempt at licensing clones. At this point they have enough alternate revenue streams available to mitigate the pain of the hit from the temporary loss of revenue from losing hardware sales while they develop their software business.
Apple is a much healthier company now than when they first attempted to license clones, with a lot more options. That makes all the difference. I’m betting they eventually go for it.
“Apple should license to X86 before someone takes BSD and builds one from scratch”
That’s an interesting idea…Actually, I wonder how Apple would react if that really did happen…A MacOSX clone for the x86. No Apple purchases required.
It may reduce their hardware sales, but I don’t believe it will weaken their revenue. Here’s my reasoning:
– iPod & iTunes seem to be their current cash-cow. An x86-OSX would not negatively impact this source of revenue (unless in a positive manner).
– if OSX is licensed as a Windows alternative, and these big OEMs get their way… they will be buying volumes upon volumes of OSX-x86 licenses ($$$). This should *at the very least* compensate Apple/Mac for lost hardware sales… but, more likely, increase revenue (see below).
What I’m saying is that Software sales seem to be the most direct way to make $$$, and Microsoft has clearly demonstrated this. They (MS) may not have a superior product, but their control over the OS market has head-butted their way to millions.
If Apple took this OSX-x86 step, they’d enjoy an instantaneous gain of the OS/pie/marketshare. This would make the platform more attractive for developers, adding a greater number of software titles (an arguable weakness of the Mac platform). The increased OSX accessibility and higher # of software titles would also stimulate greater competition b/w software vendors. Which would, in turn, translate to more aggressive pricing strategies and less $$$ for the consumer to spend.
OSX-x86 could also be seen as a prosocial movement. Diversifying the OS arena by an influx of OSX boxes would generally enhance data security worldwide. (See Viral/Worm/Exploit ecology for details.)
Apple would still have $$$ to sell Macs, should they choose to. And, with the exception of a handful, Apple would probably keep a reasonable portion of their userbase devotees for hardware sales. But, even if they didn’t, the revenue from OSX-x86 licenses would more than make up for it.
I suppose it’s just a matter of opening your mind to a different business model. Just because you’ve done things one way for years and years and years, doesn’t mean it’s the most effective/correct way of doing this. Especially when the PC market has changed so much since when you first entered the scene. (Even w/ MacMini, hardware price differences tend to shake most consumers from considering non-x86 machines.)
…just my take on things. I could easily be incorrect.
I think on average Apple has some of the fattest margins in the industry compared to other PC manufacturers.
IBM sold a lot more systems than Apple per quarter and their PC division was still not profitable.
The notions that people have on the PC side of the fence that Apple makes the same amount of money or less on hardware is not true. The iPod is all hardware and apparently there is a lot of money there so the hardware business for Apple is seems to be very profitable.
“Err, no they didn’t. They bought iTunes, and it’d already undergone a lot of development as SoundJam before then.”
Correction. They didn’t buy iTunes; they purchased an basic audio player application known as SoundJam. I’m guessing that it was meant to be a basic guide in the building of iTunes. If you were to compair the two today, the only thing they have in commen is that the can both play mp3s. And even that has changed thanks to AAC.
So Apple makes an X86 version of MacOSX so a company like a Adobe is going to put out…just for the Mac platform:
X86 Photoshop MacOSX
PPC Photoshop MacOSX
doesn’t make sense even from a developer point of view and an X86 version pretty much screws the current Mac installed base which is very loyal to Apple.
Everyone keeps taking what Jobs said as others wanting Apple to license Mac OS X and then other manufacturers build the computers. But what if this isn’t what is going on.
As we have seen before, this method of licensing didn’t work out so well. It really hurt Apple, why go through this again. Also, Steve Jobs loves hardware. He probably would never like to see Mac OS X on anything other than a machine that Apple makes.
So where am I going with this?
Well, what if other computer manufacturers want to strike up a deal with Apple similar to the one they did with the HP branded iPod. What if other companies want Apple to still create the computers and OS, but want to rebrand the machines and sell them.
This worked extremely well for Apple and the iPod. It really increased sales and got the iPod sold in more places than Apple could have done alone.
I personally think this is a much better strategy. Think of it…an HP branded iMac that is not white, but maybe a pale blue or something similar.
Mike
>>>So Apple makes an X86 version of MacOSX so a company like a Adobe is going to put out…just for the Mac platform:
X86 Photoshop MacOSX
PPC Photoshop MacOSX
doesn’t make sense even from a developer point of view and an X86 version pretty much screws the current Mac installed base which is very loyal to Apple.<<<
The same argument could be used against supporting x86-Linux versions of Windows software titles… but, it really all comes down to marketshare, doesn’t it?
It makes perfect sense to write the aforementioned titles if the distribution of Operating Systems change. If 25%+ of the market is using x86-OSX, a software company would be foolish not to jump onto that platform.
Mac Windows Mac Windows…yeesh how about we kidnap Billy Gates and make him write a better OS from scratch eh? Seems like that would be a lot easier to do than have Stevey “Control Issues” Jobs to have an X86 port. Sure x86 is not as good an architecture as the PPC but no one can deny the insane speed that AMD processors have clock for clock. Ther POWER chips that are used in servers are awesomely powerful http://www.aceshardware.com but seriously, I dont think they are gonna be out for us regular users. And Mac is seriously not that cool at least in terms of hardware. I would no sooner buy a pc from Voodoo and stick 2 30 inch Mac monitors on there than to buy boxes from Apple. I think kidnapping is in order then rofl
Only license it to people that produce PowerPC-based computers. This would force some people to introduce the chip into the PC market.
That would keep the price issue with apple for awhile.
AND—Charge top dollar for Mac OS X. Charge much more than windows.. this will also help apple balance the pricing.
ALSO: When you have the name “Apple” on a competitor’s PC box.. ask yourself… would you rather buy a computer from Apple directly or from someone that just licenses this stuff?
Marketing help too. I think the 2 big things–only PowerPC and high prices will help apple do it. OH yeah, sell directly 😉
if it ran on x86 hardware, then it would run windows at native speeds.
The every software vendor would just make an x86 windows version, because the Mac people could run it too.
a-la OS/2…everyone made a windows verison, because OS/2 could run it too.
There have been a lot of alternative OS’s on x86…buy one of them , if you must, or use linux.
I think I can understand a pc manufacturer wanting apple to release an OS X version for their hardware, so they could sell a few more boxes.
BUt from the consumer part of this…its a false argument. You want x86 compatibility, because you want a cheaper machine. So, you have this false goal, but the real goal is cheaper machine.
Just demand a cheaper machine then. Mac mini is a start, but they could do more in this area…like a full sized-low cost tower.
>>>>Only license it to people that produce PowerPC-based computers. This would force some people to introduce the chip into the PC market.<<<<
They already did that… about, ermmm, 8 or 9 years ago. As far as I know, it’s still legal to produce Mac clones today.
My old High School purchased a slew of mac clones in 1996 or 97’… the company was called Power Computing, if memory serves. The boxes performed just as well as traditional apple machines and were competitively priced against Apple’s units.
In the end, Apple Inc. bought-out and shut down most of the clone manufacturers.
“Why do you think apple osx is so great over anything else that’s based on frebsd or linux?, you guys are talking about IBM-Sony and stuff like that, Why wouldn’t you consider IBM-Novel (remember big blue from the first personal computer?, and what about netware?”
So you think we should look to two companies with a track record of dismal OS failure for the next big OS?
I’m surprised you’ve time to share your comments here – with brilliant insight like that you should be off doing high paid consulting work.
Because People tend to learn from their mistakes, and because normal people and SMART people don’t go out on their first post to insult some one else because they think they are special. You see you just showed how much insight you can bring to a conversation…
>>>if it ran on x86 hardware, then it would run windows at native speeds. The every software vendor would just make an x86 windows version, because the Mac people could run it too. a-la OS/2…everyone made a windows verison, because OS/2 could run it too.<<<
Having an x86 CPU doesn’t mean an OS will automatically be able to run Windows applications. If this were the case, Linux would be eating up marketshare like Delta Burke eats cake. You would require some sort of Windows-compatability abstraction layer/interpreter (like WINE in Linux).
The only reason OS/2 ran Windows programs was because MS convinced IBM that they believed OS/2 was the future and that MS-Windows was only a stepping stone to that future. IBM and MS cooperated in this end, as a supposed means to “transition” from windows to OS/2. IBM is still probably kicking itself for falling for this.
>>>>>I think I can understand a pc manufacturer wanting apple to release an OS X version for their hardware, so they could sell a few more boxes. BUt from the consumer part of this…its a false argument. You want x86 compatibility, because you want a cheaper machine. So, you have this false goal, but the real goal is cheaper machine.<<<<<
Vendors want OSX to reduce support costs (e.g., people calling in b/c their machine “doesn’t work”, when in fact it’s only botched up with Windows spyware/viruses/exploits).
The reason they want OSX to run on x86 hardware is because that is what their engineers, designers, and facilities integrate, support, and are intimately familiar with. Adding an entirely different platform to their mix of offerings would muddy up the water. They’d need to make significant investments in training, R&D, and overhaul their support system. (It would take Apple less energy to produce an x86 version of their software… especially if “MARKLAR”, the alleged x86 OSX port being maintained by Apple for security sakes exists.)
Cheaper OSX (via x86) machines is a mere coincidental, but positive side-effect for the consumer and OEM manufacturer.
Let me remind you guys that a great deal of the code inside Windows XP came from the code from OS/2 and Free Bsd. The Fact that they made a bad marketing campaign does not mean the software was bad at all.
I remember when IBM told Microsot that they could have the software because the money was in the hardware part. (well they where partly wrong). I think that they can step back up and fix their mistakes. And only Money, and a very strong will (I think they have a great motivation when they think about money, pride and revange). To do it right, and for most of the things that deal with consumerism, the people with the most money and power tend to win. (think Microsoft here).
A good pool of resources might help to shake them a bit.
Supporting OS X on non-Apple hardware makes no sense whatsoever. Apple is a hardware company, as were all computer manufacturers in the 80s. Software is but something they needed to develop in order for their product to be usable.
It’s plain to see that their software being runnable in hardware other than their own can only hurt their business. The **only** way it could happen would be if it ran worse on non-Apple hardware, so that it could allure people to the platform. One way for that to happen would be to leave all x86 support to the third party, with Apple supplying only the base OS. Eventually you could buy a Dell with OS X, but you’d be limited in parts choice, and you’d come to yearn for the ‘real thing’. The Dell could be cheaper enough for you to prefer it over the Mac, but not so much that you wouldn’t buy a real Mac the following year. It would be sort of a stepping stone.
There’s also an issue nobody seems to understand. Apple is a small company, with a small market share. It’s meant to be that way*. If demand for its products were to increase, it would have to become bigger. That would seriously hurt its corporate operation and efficiency, and one grown it would be hell to shrink. All those resources they’d have to acquire would eventually have to sit idle at times, costing millions and bringing nothing. Microsoft is a good example of a company that’s desperate to find sources of revenue, because its structure is just too big to maintain on an OS. After they grew enormously to supply the world with Windows, the world got supplied and has little more use for Microsoft**. Apple doesn’t want that to happen to them***.
(*) Everyone could care less what plans of world domination Apple reportedly may have [had in the past].
(**) All those side businesses ‘[formerly] bleeding red ink’ are just that, an attempt to find something for their workforce to do. If it weren’t for them, the workforce would be allocated to Windows only, and Windows too might risk bleeding red in the future.
(***) Apple may have only a fraction of the bank account or stock value of Microsoft, but they’re much healthier fractions.
>>>>It’s plain to see that their software being runnable in hardware other than their own can only hurt their business.<<<<
Not if their software sales compensated for lost hardware sales…. which is a likely consequence.
When a user notices that their *brand new* PC slows down to a crawl, they think/say “this PC sucks; it’s broken; I’m never buying another <<<insert vendor name here>>> PC again!” Some OEMs are looking to get away from Windows because they see the over-targeted/exploited WindowsOS as bad for their business.
When your average joe buys a PC, they expect it to work consistently over time. Unfortunately, this is rarely the case with Windows. Over the past decade, Windows maintenance has gone from a simple defrag/virus scan…. to a convoluted, hodgepodge of maintenance —defrag/scandisk, registry scanning, virus scanning, adware/malware scanning, active protection, virus shields, adblocking, firewall/internet security monitors, etc.
This problem does not *currently* exist with OSX, and it’s likely to stay that way for a while (until marketshare levels out b/w the bigshot OS makers).
Apple’s original business practices may be defunct this day and age. If they’d like to remain a niche provider, they should stick to their current business model. But, if they’d like to spin more directly against Microsoft… they should seriously consider this move.
Right you are, but I feel that Apple did not manage or do this correctly. They have a much better chance to make some headway.
As for the Mac Clones, I’ve seen companies attempt this–more so buy refurbished macs these days and sell them with mac OS, but apple shut those down thru the legal system.
i just think they really hurt themselves in the past and did it the wrong way. When something dosen’t work you should change it to make it work rather than abandoning it.
>>>>Right you are, but I feel that Apple did not manage or do this correctly. They have a much better chance to make some headway. As for the Mac Clones, I’ve seen companies attempt this–more so buy refurbished macs these days and sell them with mac OS, but apple shut those down thru the legal system. i just think they really hurt themselves in the past and did it the wrong way.<<<<
Agreed.
>>>>>When something dosen’t work you should change it to make it work rather than abandoning it.<<<<
*EXACTLY* …Changing with the times is important to the continued success of any business/institution.
Computing examples aside, look at Madonna. She’s reinvented herself about a bazillion times to stay fresh.
ok, that was sorta gay… but highly relevant. =)~
Perhaps to “keep all options open” they will continue to support and expand hardware compatibility for Darwin x86, but as has been pointed out, they shoot themselves in the foot if they allow their OS to compete with their hardware sales. One of the purposes of the Mac mini was to expand sales of Apple *hardware* by making the “introduction to the Apple way” more attractive, price-wise. The mini was a smart move. I’m sure PC makers would love to have an OS X x86 version – they would be stupid not to – and I’m sure they have approached Jobs. But Jobs knows that their greatest profit potential comes from maintaining the software/hardware tandem. He is simply throwing this information out to continue to generate interest – build the general public perpection – that OS X is so superior that the even the PC makers would dump Windows if they could, to drive Apple sales. Pretty simple. BTW, I’m an Apple guy. 🙂 It IS better. I’ve used Windows since 3.1 and I still service Windows machines. I got a Mac two weeks ago to better familiarize myself with the hardware/OS (because of Mac mini sales), and I’m not going back.
>>>I personally think this is a much better strategy. Think of it…an HP branded iMac that is not white, but maybe a pale blue or something similar. <<<
I’ve been thinking along the same lines. Let HP or Sony or another big brandname build the case, and stick Apple hardware and OSX inside, maybe with a little custom slant for that brand.
It could work!
If cell ships on time it will be a fast PPC processort regardless of the funky fpu units. I could see a mac witll 4.6 GHZ cells and a library that uses the fpu units like altvec. If nobody but apple and linux have and os which would you want to run for the home user? Clearly an hp/gateway/ibm OSX cell box would be less work then a linux version for the builder. Either way they have to pay somebody to setup there software for their desktops. So it is either RedHat, Novel, or Apple. Home users know who Apple is…
Of course Apple imagines it can win that business. Be interresting to see if they can win that business.
Apple has been so good in its marketing since OSX and the first eMac that it gets PC makers thinking they’d like a piece of that…. But none of them will drop windows, unless they are made to.
If linux vendors come up (in common and in the long run) with such a sexy communication, PC vendors will start offering it preloaded. But Linux vendors still need to understand that they are all too small to achieve that on their own… Shrug.
I don’t think there would be much of a danger in a x86 port for the Mac hardware business though. It is superb stuff for those who like this kind of things (I do). If someone did a clone of the iMac on a x86 arch, I’d get one as my Linux workstation. Feel free to point me to someone in the UK btw 😉
Let me remind you guys that a great deal of the code inside Windows XP came from the code from OS/2 and Free Bsd.
Prove it! Not some consipracy theory crap, hard evidence.
Much of the original 32-bit TCP/IP stack was BSD code. I don’t know how much of it is still in there. Of course you could always grep through the Win2k code that is out there.
And right you are! -I see it slightly different, though, when it comes to the type of alternative I would like to see emerge….
I have for years been wondering what IBM has done about its OS/2 development efforts. Why not realise that killing off a (in many aspects) great product might not be the way to go??
Personally I think that many of the strenghts of OS/2 and Linux/*BSD are shared, that being stability, ability to be deployed for mission critical tasks, (relatively) small memoty footprint and great scalability. -All of these points being points touted by Apple as being reasons for deploying Mac OS X in the enterprise.
I find, though, that a very interesting alternative to Mac OS X on x86/x86-64 would be an IBM branded OS built on top of FreeBSD or OpenBSD, adding a GUI built on OS/2’s Work Place Shell, with its very unique and powerful customization possibilities and object oriented design, and Apple’s ‘Yellow Box’ Cocoa API’s to provide a ‘Mac OS X like operating system’ that is not in fact Mac OS X but would allow applications to run – virtually unchanged – on both PPC and x86/x86-64 like in the old days with OpenStep and NeXTStep.
A system built on *BSD with the immense stability offered by this platform and ‘acceptable’ hardware support, an attractive GUI with all of the OO-strengths of OS/2 and the ability to run ‘slightly modified’ Mac OS X apps would be great for Apple — Greater exponation of key technologies, like e.g. QuickTime, Cocoa, Rendevouz, Spotlight and iTunes without it being an exact ‘sam-same-but-different’ approach could be great for Apple.
Right now everyone wants Windows Media Player 10 to come to their platform, since they’re dependent on the .wma format and DRM support (yet to be found in Mac OS X releases of Windows Media Player), but with Apple’s QuickTime technology being spread to other OS’s than Mac OS X and Windows the alternative would strengthen and with the ability for Apple to show off a superior product on two different platforms, and with a strong supporter like IBM in on the idea, plus another strong, well featured, easy to use and powerful OS to help gaining ground in the server market through possibly market leading technologies shared between Mac OS X and ‘IBM bsdOS/2’ both companies would be able to benefit greatly from this.
Look at Sun Microsystems. They’d love to have a desktop OS like Mac OS X to acompany their tchnicaly brilliant Sun Solaris 10 OS…… IBM might want the same thing for their AIX or Linux solutions, however, this solution described might bring the best of both demands – modern, desirable end user OS strenghts combined with platform independence application-wise and the ability to secure development efforts on both platforms due to the ‘Yellow Box’ ‘write once run anywhere’ philosofy. An Oracle DBMS could be easily ported to both platforms, giving people the opportinity to choose OS and hardware suiting their needs without sacrificing stability, security and ease of use.
The key is to have Mac OS X compatibility, the ability to run Cocoaa-apps, on x86/x86-64 without it ‘taking over’ from Apple’s PPC business. I for one would buy it (‘IBM bsdOS/2’) as I’ve bought Mac OS X….
Lumbergh: “Let me remind you guys that a great deal of the code inside Windows XP came from the code from OS/2 and Free Bsd.”
Prove it! Not some consipracy theory crap, hard evidence.
IIRC, it was only the Windows 9x series that had borrowed some code from BSD, and with that being said, does it matter? BSD code is free for all and sundry to use – how is it wrong for Microsoft to use it? no more wrong than SUN using BSD as a base for its own operating system.
WinXP does not have an appreciable amount of BSD code in it. MS bought a BSD-based TCP stack from another company for NT as a temporary measure, but an MS-written TCP stack replaced it in NT 3.5. Some user utilities, however, were never rewritten, and that’s why programs like the XP ftp client are based on the BSD code.
Three big “PC” companies?
Let’s see… IBM-Sony-Toshiba, are big uh?
They have Cell but don’t have a viable OS for it…
Who said they want OS-X for x86? Doesn’t even exist!
question: do those companies have something apple wants?
answer: yes, a range of cpus that competes with intel
question: what’s in it to make it worth apples while?
answer: commitment to the line of cpus apple needs to stay in business.
ok, there’s still a problem with these chip makers competing on apples computers, but perhaps that could be worked on by having a set of rules on what they can and cannot do. os x on a pda anyone, or a phone, or a set-top box or other areas that apple doesn’t want or cannot go after.
just a guess …..
why? too much risk, too much work.
the cell processor speculation is much convincing.
os x for cell not only to apple but also ibm, sony, toshiba.
i might buy my first mac if this is the case.
somehow i sense it just after cell is announced…
Coding is not a trivial matter if you are 1 or 2 man operation. But before I gaive my rationale, please excuse me cos I am no programmer and I would not know the difference between C#, ++ or Pascall for that matter. I am giving my view on economics.
Back to the 2 man developer. Imagine that they have quit their job, invested whatever money for that killer app they have in mind. Released it for Mac OS X on G4 or G5 and the app flew. Everybody love it. Then out of the blue Apple says we are porting to 86, no if and buts. But here lies the the developers dilemma, do you port for Mac 86 or might as well port to Windows where the market is undeniably bigger. It is not you would like to repeat your months of instant noodle (put your staple poor programmers choice here) routine all over again. What if Apples Mac 86 bombed? There is real life to take care of.
Even major developers are not supporting Apple in a way that is conducive for Apple’s well being; Adobe, Quark. Macromedia, MS – everyone of those apps is lacking somewhat on the Mac platform. Remember these are big corporations who would leave Apple at a snap of the finger. Would they want invest more time and money to entertain Job’s “eccentricities”
So if these big developers leave Apple, what do you think would the small timer do?
In conclusion if Apple were to jump into 86, the should well have a huge trampoline and safety net as backup!!!
Apple would go directly to x86-64
Apple still doesn’t have an operating system that works with 64 bit. Even Tiger will be compiled in 32 bit binary. So much for the “first ever 64 bit personal computer” As far as I can tell, the only 64 bit feature that Apple uses is the ability of having more than 4GB of RAM.
Refer to http://arstechnica.com/articles/paedia/cpu/cell-1.ars
Haven’t read all the replys so forgive if I say the obvious, but Macos is based on unix yes? So how much more difficult would it be to create a macos clone os? Also I have just changed back to Windows from MacosX because I found it getting a mite annoying being unable to find the software I wanted. Not that there isn’t any for mac, its just there is all the stuff I want on windows pc, not mac, pity really, it wasn’t a bad machine to use.
>The other thing is that the X86 architecture is hitting >its theoretical limit.
On the contrary, the current AMD64 ISA has theoretical limit of 256 processor cores “per package”. Refer to http://www.theinquirer.net/?article=21111
Secondly, modern X86 processors doesn’t natively execute X86 ISA.
What exactly is the benefit for Apple moving to X86?
Faster processors?
Apple is off to a great start with the G5 and the perceptions that a G4 is useless jut doesn’t hold up.
Lower cost?
I’ve been buying Macs for years, they sold for a premium then and they still do now but not as much. I really don’t see Apple selling X86 Macs for under 500, they already are in that market with PowerPC and I don’t see any kind of X86 in the Mini’s form factor with the same computing power, functionality and software bundle.
The value proposition for the Mini has been debated and judging by its popularity people do perceive it as a value. Apple has been steadily dropping its prices across the board. They did it again with the recently announced PowerBooks.
Selection?
How many different types of NICs does someone really need? ATI cards? Macs got em. NVidia? OEM on Apple and you can find them. SATA controller? RAID controllers? Fibre Channel? PCI sound cards? Macs have them.
Games?
Still has to be coded for MacOSX X86, although it may be easier. Not really a compelling arguement for Apple. Yeah lets port to X86 so we can have more games.
Hardware business unprofitable?
Not at all, exactly the opposite. Apple has fat margins, they are not in same pickle as other PC manufacturers putting out X86 boxes. Microsoft is in the hardware business too and I do not see them pulling out of that anytime soon. I also doubt that MS is in the hardware business for charitable reasons.
What is so compelling about moving to X86 anyway? The legions of people that already have X86? Is it so that someone with a Packard Bell buys MacOSX86 off the shelf and thinks it a peice of crap because it runs so slow or someone with a gaming rig buys MacOSX thinking that they can get higher framerates but realizes that the drivers are not as advanced and the games are not as optimized and writes off MacOSX. No thanks, thats a lot of wishful thinking and poor business sense.
Its not really a technical issue but more related to the business side of computing. As geeks we like to think that MacOSX on X86 is neat but the bottomline of it being profitable doesn’t seem to matter.
Then take into account the Apple stores. Why even have them if you are just going to sell PCs? Ask Gateway this question too. Also with a Mac box once you buy it you are in a sense stuck with MacOSX if you don’t like it. On the PC side of the fence you reformat the drive and install Linux or Windows. Its not exactly a great way for Apple to retain its customers. As it is right now I know of very few people that are totally dissatisfied with MacOSX, I encounter quite the opposite actually and the biggest Mac fans actually are the PC converts.
Does moving to X86 make Apple money? I don’t think so. Makes no business sense any way you look at it. I think what Apple needs to do actually is move people off of X86.
Apple still doesn’t have an operating system that works with 64 bit. Even Tiger will be compiled in 32 bit binary. So much for the “first ever 64 bit personal computer” As far as I can tell, the only 64 bit feature that Apple uses is the ability of having more than 4GB of RAM.
64-bit is NOT the silver-bullet of computing that everyone makes it out to be.
Most programs simply won’t benefit by any substantial amount just by converting everything to 64-bits. If a loop counter only needs to count to 100, then making it a 64-bit integer isn’t going to make a lick of difference, except possibly occupying more memory, which is wasteful.
Just because OS X itself itsn’t fully 64-bit code, doesn’t mean any of your own apps can’t be 64-bit. A G5 will still run all your 64-bit code (like Oracle’s latest OS X version), so what does it matter what the OS is?
Yes, Steve Jobs lied again. The last version of Soundjam MP and the first version of iTunes were nearly identical. People did resource by resource comparisons at the time. One can likely find them by searching Usenet. Although I rarely take any Wikipedia article as authoritative, the one on iTunes does state that “The first release of iTunes was very similar to SoundJam MP. . .”
http://en.wikipedia.org/wiki/Apple_iTunes
[i]Actually, when I wrote this I was mostly thinking a very nice episode of “Star Trek: The Next Generation”. In that episode a “computer game” makes the player a drone. Wesley Crusher and his girlfriend are the only ones in the ship who haven’t try it and near the end they are forced to play the game to become drones too. His mother says these words I think at some point. Commander Data saves the day at the end.</>
IMHO it was one of the worst TNG episodes I’ve ever seen. The only good thing was wesley’s girlfriend (ashley judd)
🙂
The thing is that you’re assuming that porting the code to a new processor is a huge task. For properly written code, it shouldn’t be. In your example, porting to Windows is a much bigger task than recompiling for OS X/x86 should be. It makes your product support more complicated, but that should be the extent of the extra cost, and that might be offset by your larger customer base.
It’s certainly possible to write portable code, especially code that is portable between CPUs. Most code doesn’t do any low-level access, so really the only thing to worry about is word size and byte order. Handling those properly isn’t more effort, it’s just a matter of being careful. Especially when we’re talking about desktop apps written in a high-level language like Objective-C, porting to a new processor shouldn’t be as enormous a pain as everybody is making it out to be.
FYI:
“Windows 95 doesn’t do multitasking, just multithreading…
Windows 95 handles multitasking in the kernel and threads out to programs, but they have to specifically be written for multithreading.”
WinNT was the Multitasking multithreading success story. And it did not compete with MacOS Classic.
it’s main competition was Unix.
OS7.1 which was released in 1992 was the first MacOS with solid backgrounding of applications, with the introduction of the Thread manager. (I remember him).
“older programs do not have to be rewritten to take advantage of threading, they are automatically threaded by the OS”
this competes with Win3.1.1, which does not.
OS 7.5 improved on the Backgrounding.
“runs multiple programs simultaneously and can run background threads which do not have to be programmed into the application, and are optionally controlled either through the application or with defaults in MacOS itself.”
This is released in 94.
This competes with Win95: “can background program threads, but backgrounding is controlled in the OS, not the program.
there isn’t a priority chart. Programs must be written to support the multithreading in Win95, as it was a new feature when 95 was released.”
finally os8 “runs multiple programs simultaneously on top of the kernel, and can run background threads from multiple programs which do not have to be programmed into the application, and are managed entirely by MacOS with an automatically adjusted priority chart determined by application need and load…”
This is released in 97 and competes still with Win95 until Win98 came out later.
check the references man.
http://kb.indiana.edu/data/abmc.html?cust=630800.96989.30
http://www.lowendmac.com/tech/macos_v_win.shtml
http://applemuseum.bott.org/
@Hammer
post1– what are you saying? (that article mostly supports my argument)
post2– That AMD can do a million cores per cpu does not make the x86 IP anyless outdated.
Read the rest of the Ars Processor Articles Jon ‘Hannibal’ Stokes and others have writen, the overall picture of the future of processors point away from x86, time to move on.
( http://arstechnica.com/articles/paedia/cpu.ars )
The cool thing to me about AMD is that technically they have the skill and flexibility to move away from x86 with the IP/engineering base they have. (I do think they should buy transmeta though for added software flexibility).
@DoubleTap
I completely agree, it makes no sense, either in business or Technically. why would Apple do well where Linux/BeOS/Solaris are already struggling with.
@itunes/soundjam.
I own soundjam, and looking trough my stuff I found a back up cd that has both soundjam MP plus, and iTunes1 launched classic and checked it out, and I can tell you that they are like twins, except that one has dark hair and blue eyes and the other Brown hair and brown eyes.
I do remember the whole thing about the code also being the same.
Steve has to be referring to the OSX port, or he was smoking some of the fancy grass from Humbolt County.
Also someone mentioned that PPC744X processors are not as bad as they are made out to be. I have to agree, the folks that rag on G4 should really get their facts checked, G4 never was a bad design, although Motorola dev/production processes have sucked hard in the past and indered the G4 a lot; but now with the steady increases in clock and bus speeds the doubling of cores, they are fine processors.
On the flip side though PPC750 AKA G3 is such a better design, and I wish IBM hadn’t messed up its timing and just gotten the altivec license much earlier.
The funny thing is just like in the x86 world where the PIII design is getting a revival in the Pentium M which is so much nicer then the P4; the G3 might get a revival as well now that IBM has licensed Altivec; it would be a great thing if we had 2Ghz PPC750 with altivec on a new super thin and light Apple Notebook.
ok why would anyone assume that pc vendors want to put Os X on x86?
To do so would be utterly useless due to a lack of applications and the pc makers know that.
Don’t you think its more likely that the PC makers want to build an apple clone using a powerpc?
PC makers might lack the courage to stand up to MS but that does not make them dumb.
If Apple were to license an x86 MacOS X to mainstream PC manufacturers–and consumers actually bought these new MacOS-running PCs in large numbers–then adware/spyware/malware authors would just take their wares to the new platform. It’s basically trying to fix a social problem by resetting the clock.
To Apple, this wouldn’t seem any better than if Dell or HP were requesting them to restart the Mac clone market. They might as well ask Apple to sell off its elitist brand and close up shop. As an added bonus, it would encourage people to come and make their platform suck more through the production of junk software. It doesn’t seem likely that PC makers would be willing to pay the sort of cost that would justify this effort and potential brand depreciation.
But I really must wonder if the claims made are even true. It would be as much of a pain in the ass for PC manufacturers to ship machines running MacOS as it would be pretty much any fringe operating system. When it doesn’t run the programs that their customers want, they’d go somewhere else and get the Windows PC they really want.
“Windows 95 doesn’t do multitasking, just multithreading…
Windows 95 handles multitasking in the kernel and threads out to programs, but they have to specifically be written for multithreading.”
Windows 95 (assuming you don’t have any legacy 16 bit Windows or DOS drivers or software lurking around) pre-emptively multitasks Win32 applications. They do not have to be “specially written” for this to happen, they just have to be Win32.
No version of MacOS Classic pre-emptively multitasks anything. You can freeze up the *entire OS* just by holding the mouse button down on a menu. MacOS Classic software *does* have to be specially written to “play nice” with other applications and individual applications can easily soak up all available CPU resources or crash the machine in regular day to day operations.
I’m also rather sceptical of “older programs do not have to be rewritten to take advantage of threading, they are automatically threaded by the OS”. I don’t quite see how a program that wasn’t written to be multithreaded can magically become multithreaded just from an OS upgrade.
Windows 95’s multitasking and memory protection is certainly not as good as NT’s or unix’s, and makes substantial reliability and performance sacrifices in the name of legacy support, but it was streets ahead of anything MacOS Classic ever did.
The quickest and easiest way to demo this is to decompress a file in the background under both OSes while performing some typical actions in the foreground (browsing the web, email, word processing, etc). Even under ideal conditions, on MacOS the (backgrounded) decompression will absolutely crawl, but under Windows 95 is will hardly slow down at all. Another good example is backgrounded file copies (particularly over the network).
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/wi…
http://www.microsoft.com/technet/archive/win95/rk31_arc.mspx
People who are interested should pick up and read a copy of “Inside Windows 95”. It really is amazing what Microsoft did with Windows 95, and you don’t fully appreciate it until you see the nitty-gritty details.
“Just to remind mister Jobs that:
.When Windows 95 was released MacOS wasn’t (preemptive) multitasking
.Mac users had to wait for Apple to TAKE and FOUND their OS on an Unix kernel to discover multitasking… 5 years after Win95, 15 years after AmigaOS, …
But he’s right: they did not copy anything, no… only “borrowed” a kernel maybe.”
– Windows 95 wasn’t preemptable, and neither was AmigaOS. You don’t know what you’re talking about.
– Mac users didn’t have to wait 15 years for multitasking, it was available in a slightly more crappy form than Windows 95 since system 7.
– Apple didn’t steal a kernel, OS X is based on the Mach kernel developed at Carnegie Mellon University by Avie Tevanian, who was VP of engineering at Next, and current head of software at Apple.
– The BSD portion of OS X is in the form a wrapper around the Mach kernel that is based on FreeBSD, again though, Apple hired the technical lead of the FreeBSD project to lead the team working on the BSD subsystem.
But otherwise I completely agree with you.
> Windows 95 wasn’t preemptable, and neither was AmigaOS
Pre-X MacOS and Windows 3.x offered only cooperative Multitasking.
MacOS X, Windows 95 and AmigaOS (since 1985) offer preemptive multitasking.
A video from the 80s demonstrating some of the classic Amiga’s capabilities:
http://www.archive.org/movies/details-db.php?collection=computerchr…
I agree with Ryan, why would the top pc makers be able to support Apple when they depend on MS for the bulk of their sales? the only pc makers that are “independent” of MS would be IBM, Sony and Toshiba who must have revenues beside their PC business. and these 2 are collaborators on a power-PC based low-cost, super powerful chip (Cell) which is going to kick the crap out of any x86 box. the Playstation 3 is going to be $299 – what do you think the next cell-based Mac Mini is going to cost?
“- Windows 95 wasn’t preemptable, and neither was AmigaOS. You don’t know what you’re talking about”
AmigaOS had preemtive multitasking since it born (1985) so you are what don´t know what you´re talking about
What’s the use, really? I go out, by a x86 PC, it comes with Mac OS X. Then I go out, by Office for use on x86 PC w/ Mac OS X, then I buy Adobe this ‘n that, then I buy…then I buy…. Point is, just because the HW is different does not save any money. You still have to buy OS X programs, etc. If you want to run OS X, buy Apple HW. It would be the same…just the color of the box would be different.
BullShit.
Jobs said that just to sell more iPod. am not sure they wanna to maintain the macintosh, they like to transform into the next Sony, not into the next IBM, SUN or whatever..am not surprise when they introduce their own game console box= Box-X!
Apple is a hardware company, they make money by selling hardware. OS X is a deal sweatener. Despite industry pundits saying that they have been at deaths door for over twenty years the only time when this was really the case was when they allowed the clones. Doing this would be very dangerous.
Some reasons why they will not be going x86:
Apple will not release for x86, there has been a lot of optimisation work done for the PPC, especially getting the most out of the Altivec unit. Switching to x86 would lose all of these optimisations with OS X being back at responding like 10.0.
Not matter what special hardware they include to prevent piracy ways will be found around it. Having scrapped their (high margin) hardware for (low margin) software, suddenly nobody is willing to pay for it anymore as they can just download a bittorrent. Result no revenue, and soon after no company.
Where they to release for x86 ALL current software would not run, it would all have to be re-written and optimised for x86. Not all the current developers would consider this to be worth the cost of maintaining support for OS X. Result, the developer flight that x86 people have been claiming for ages suddenly gets some substance.
Microsoft, as has been shown by the example of BeOS, will use whatever it needs to destroy any rival (or potential rival) that it sees breaking into its monopoly. On PPC apple isn’t a threat to Wintel, on x86 it would be and the knives would come out.
Lastly the PPC is a better chip, especially in heat consumption. One of Apples main markets is the music industry. To compete here you need very quiet, but very powerful, workstations. If your machines suddenly start needing cooling fans that sound like a jet turbine people are not going to be best pleased, as Apple aready found to their cost with the ‘wind tunnel’ G4.
http://www.computercloset.org/BandaiPippin.htm
I wonder if Apple will try again going down that road…
Apple should license to X86 before someone takes BSD and builds one from scratch, it is not like Apple can do things others can not do since Apple computer is run by the same breed of humanity as anyone else walking around on planet earth…
if Apple sells the license then Apple gets to make something off of it, if they don’t sell the license and somebody else builds a slick BSD based desktop then Apple makes nothing off of it…
That’s a nice one. No, Apple can’t do anything others can’t. Only, the others are not doing it. For years it’s been “Apple will be dead in two quarters”. Now the PC makers are sick of Microsoft’s dictating them what they can and cannot do and they want to get rid of them. So now Apple has to licence their OS to them. It ain’t happenin’, cousin. Besides, your “they can’t do what others can’t” is missing the whole point of the experience you buy when you buy an Apple. There is a reason why it works. It’s not some funky coinkidink.
If people want to run Mac OS X that badly, as someone already mentioned, have them walk into an Apple Store. The people there will be more than happy to sell them a computer, no worries.
As far as running Mac OS X on linux is concerned: no doubt this works, but I’m not at all convinced your emulator takes care of the interoperability of the software and the hardware.
I installed my iSight this weekend. Actually, that’s far too big a word for it: I plugged it into the FireWire socket. It worked without installing any software, changing any settings. Zero configuration.
Is your emulator going to offer the same functionality? Across the board?
I’ll believe it when I see it. And frankly, I don’t need to see it. Don’t whine about installing OS X on your dishwasher running Debian. Get yourself a Mac and get over it.
Next thing you know, they’re going to sue Apple for being a monopoly.
blah, blah, blah, blah, blah
Apple sells hardware, not OSs
Commander Data die, Why!!?? Why!!??
People who are interested should pick up and read a copy of “Inside Windows 95”. It really is amazing what Microsoft did with Windows 95, and you don’t fully appreciate it until you see the nitty-gritty details.
Nice plug there. Microsoft didn’t do anything “amazing” with windows 95. What NeXT did with NeXT Step was amazing in 1992, leagues ahead of win95, 98 and ME or NT 4.0.
People who are interested should pick up and read a copy of “Inside Windows 95”. It really is amazing what Microsoft did with Windows 95, and you don’t fully appreciate it until you see the nitty-gritty details.
To be more complete. They finally wrote a scheduler that uses a clock interrupt to pick the next runnable process off the run queue after an expired quantum, woohoo. I did that in my college OS class as a 1 week sub project. We did develop the OS with VM, devices drivers and a Microkernel on x86 hardware.
Truly amazing indeed my college project that is, not win95.
No version of MacOS Classic pre-emptively multitasks anything.
Wrong again…..
From apple
Mac OS 9 Threading This section describes threading on Mac OS 9. Mac OS 9has two threading APIs. Thread Manager provides cooperatively scheduledthreads within a process. MP tasks are preemptively scheduled by thenanokernel.
………..
While the “MP” in “MP tasks” stands for”multiprocessor,” MP tasks are available andscheduled preemptively even on single processorsystems. You don’t need a multiprocessor system totake advantage of preemptive threading with MPtasks.
……………..
The entire cooperative environment runs within a singleMP task. This task is known as the . All Process Manager processes and all ThreadManager threads are executed by the blue task. Other MPtasks, created either by the system or by an application,are executed as separate entities. The nanokernel schedulestasks to run on the processor (or processors) in apreemptive fashion. bluetask
I do agree that the way it is done is not elegant at all but your statement is wrong.
“Apple should license to X86 before someone takes BSD and builds one from scratch, it is not like Apple can do things others can not do since Apple computer is run by the same breed of humanity as anyone else walking around on planet earth…
if Apple sells the license then Apple gets to make something off of it, if they don’t sell the license and somebody else builds a slick BSD based desktop then Apple makes nothing off of it…”
Uh sure OK. Apple is going on its 5th revision of MacOSX. Apple probably doesn’t have the programming resources of companies like MS, IBM or Oracle and their are hundreds maybe thousands of Linux programmers and contributors so who exactly is going to come out with this slick BSD based desktop?
The BSD crowd? yeah right like anyone who runs BSD needs a GUI…
stop spreading FUD. Mac OS had multitasking built in in os7 . It could be done on earlier versions like os6 using multifinder. Get a clue.
Nice plug there. Microsoft didn’t do anything “amazing” with windows 95. What NeXT did with NeXT Step was amazing in 1992, leagues ahead of win95, 98 and ME or NT 4.0.
As usual, completely lacking in context. There’s more to an OS than the buzzwords and bullet points.
NeXT certainly was impressive – so were the early PPC versions of MacOS with 68k emulation (and most of the OS running in it) – but for different reasons.
To be more complete. They finally wrote a scheduler that uses a clock interrupt to pick the next runnable process off the run queue after an expired quantum, woohoo. I did that in my college OS class as a 1 week sub project. We did develop the OS with VM, devices drivers and a Microkernel on x86 hardware.
Did you manage to implement complete win32, win16 and DOS compatibility – including low level driver support and bug equivalence – as well ?
The impressive thing about Windows 95 was the legacy support it offered *in addition* to its buzzword compliance and “modern” features. As you obviously know, just whacking together a basic OS isn’t particularly hard – it’s all the other stuff you have to do to make it useful.
Wrong again…..
Sorry, I forgot about your difficulties with context.
No version of MacOS Classic pre-emptively multitasks anything _usefully_. The end user sees no benefit unless the software developer has specially written their application to allow it (and the impression I”ve gotten is that it isn’t/wasn’t trivial). Very few developers actually did this (probably because by the time the feature was available OS X was visible on the horizon).
Not to mention MacOS 8.6 (first version with the nanokernel) was released in 1999. 4 years after Windows 95, 3 years after NT4, 1 year after Windows 98 and about 9 months before Windows 2000.
Apple were very, very late to the whole “modernising” party.
Some interesting commentary here:
http://lists.apple.com/archives/mt-smp/2001/May/msg00007.html
“Apple should license to X86 before someone takes BSD and builds one from scratch, it is not like Apple can do things others can not do since Apple computer is run by the same breed of humanity as anyone else walking around on planet earth… ”
actually, i would put apple designers as a different breed, part of the species of demi-god that walks the halls of blizzard, that make the world a better place by their miracules creations. 😉
no seriously, apple has something that is *extremely* rare in the industry. good designers, and management who listens to them. cant think of any other operating system out there today with the elegence of the mac, probably the last one was be.
They already did that… about, ermmm, 8 or 9 years ago. As far as I know, it’s still legal to produce Mac clones today.
The problem is it’s not legal to run many versions of MacOS (in particular OS X) on non-Apple computers.
My old High School purchased a slew of mac clones in 1996 or 97’… the company was called Power Computing, if memory serves. The boxes performed just as well as traditional apple machines and were competitively priced against Apple’s units.
Most of the Mac clones were faster and cheaper than the Apple machines. That’s why they nearly killed Apple.
In the end, Apple Inc. bought-out and shut down most of the clone manufacturers.
Apple simply didn’t relicense the necessary software and hardware information the cloners needed to make their machines (or, more accurately, asked for a much, much higher price that the cloners were prepared to pay).
The thing that always baffles me about arguments with people like drsmithy is a statement like this:
“No version of MacOS Classic pre-emptively multitasks anything _usefully_. The end user sees no benefit unless the software developer has specially written their application to allow it (and the impression I”ve gotten is that it isn’t/wasn’t trivial).”
What is this.
Did you use Classic?
Did you use 95 at the same time?
Can you support your claims with facts or even a little bit of real personal experience?
Do you believe that using authoritative statements based on hearsay are enough to make you right?
I don’t know what your background is or how old you are but right now you sound like a pretty ignorant 16y old MonoOS user.
As a Mac user from OS1, an Irix user since vs5, an NT user since 3.5, and a power user (serious graphics and video work) since Mac OS7 in 93, I can say that when cooperative Multi-tasking came to age in OS7.5, it usefully improved my workflow and usefully improved the Mac user experience.
I started using NT with version 3.5.1, with 3d applications Lightwave and Softimage.
It could not replace the high end machines in some of the studios I worked for, but definitely killed the SGI Indy line as well as the O2 which replaced it.
NT was significantly more robust OS then the Mac OS at the time, but I could still do whole projects using Electric Image and Cosa After Effect and Avid Media Composer without having to really worry about crashing as much as people assume.
Just like any other workstation environment once locked down the classic Mac OS was as functional as NT; and where the stability and memory protection of NT shone, the user interface, lack of plug and play, and the lack of multiple monitor support (this was a part of Mac OS since system 6) really hindered it.
Win95 was not a viable replacement for the Mac because it was not significantly better in any way.
It’s multitasking was a kludge just as that of the MacOS and suffered the same faults, due to legacy constraints, and it was still a DOS based windowing system.
I knew of not one active professional who would use Win95 when they could use NT or Mac OS classic.
Part of the reason was that 95 was not significantly more unstable, but because of millions of registry conflicts, DLL corruptions legacy 16 bit code in DOS, etc; it was a support nightmare.
With NT I could get significantly faster hardware then on either what was available for the Mac or Win95, (ALPHA), and it made a real difference in my workflow.
So if we are going to use subjective statements about these operating systems, at least lets talk about them from our experience and not hearsay.
I will tell you this though, when I did first use Win95 I was very impressed with it in relationship to what it replaced, but within the same year os8 came about and erased any real advantage.
95 was a vast improvement over Win3.1.1, but definitely not an improvement over the Mac OS experience; not even by a long shot.
(it would be like saying that a MCD’s burger is a really good burger because it is bought and eaten more than any other burger in the world.)
What Apple was loosing to and competing against was WinNT, and this was definitely a scary scary time for Apple back then. Especially when so many ISV’s and Vertical Market apps where moving to NT.
When Adobe and Quark made NT versions of their programs on par with the Mac Versions I almost died.
I mean the versions of PS and AI available for Win before that where 2 years behind.
PS3.0 had been out on the mac almost a year, and all of a sudden PS3.0 and AI 4.1 on NT at the same time.
Now that was scary, but that was then.
NOW look at the situation. Steve Jobs can blatantly lie, yet since he is finally supported by the best OS available (by far) and really good hardware, he is safe from Pie in his face (which Bill G Has experienced a couple of times).
Did you manage to implement complete win32, win16 and DOS compatibility – including low level driver support and bug equivalence – as well ?
The impressive thing about Windows 95 was the legacy support it offered *in addition* to its buzzword compliance and “modern” features.
That’s not particularly amazing. Maintaining legacy APIs is a requirement more than an amazing feat. Lots of Oses have maintained legacy and backward compiliance over the years.
“Apple simply didn’t relicense the necessary software and hardware information the cloners needed to make their machines (or, more accurately, asked for a much, much higher price that the cloners were prepared to pay).”
Actually the other poster was right and you are wrong.
Apple Bought PowerComputings Operations and Licensing rights.
BOUGHT back from IBM and Motorolla their rights to the OS, and the right both parteners had to re-license the OS to other manufacturers.
(for example UMAX got it’s license from Motorolla not Apple)
the original poster is wrong in that it is not leagal to make machines that can run osX presently, as Apple does not sell or license it’s ROM to anyone which is what allows a computer to run the MACOS classic or not.
New Machines don’t use an actual ROM chip anymore, but they still need an Apple ROM-IMAGE to run.
The CHRP platform which was to supplent PREP and the old apple hardware, was to abolish the need for the ROM and would have allowed cloners to make machines completely independently from Apple.
These reference designs are what make the Pegasos PPC design Motorolla licenses (CHRP), and the OpenServer designs which IBM licenses (PREP).
This is when Mr. SJ, decided to close down the cloning biz, as he realized that he would be IBMed by IBM.
At the time people where mad about this, but looking back he saved Apple for REALS.
ah ah.
NeXT certainly was impressive – so were the early PPC versions of MacOS with 68k emulation (and most of the OS running in it) – but for different reasons.
You were responding to seabasstin about multitasking, yes. What are the different reasons? NextStep was great at multitasking.
Not to mention MacOS 8.6 (first version with the nanokernel) was released in 1999. 4 years after Windows 95, 3 years after NT4, 1 year after Windows 98 and about 9 months before Windows 2000.
Apple were very, very late to the whole “modernising” party.
I don’t know what “modernising” has to do with preemptive multitasking. UNIX has been doing it about 20 years longer than anything microsoft has put out.
I think you are the one taking things out of context there. You claimed no MacOS classic had pre-emptive multitasks anything. Then later you changed it to “anything useful”. Anything pre OS X is Mac OS classic, regadless on when it was released. Release date is irrelevant to the statement you made.
If you meant to say that all the cooperative stuff runs as a single mptask so the end result is the same, I might give you credit. But you made a blanket statement as usual and acussed some one else of switching context, again just as usual.
Your original post which raptor replied to, started out talking about preemptitive multitasking features in windows 95 and ended with you saying microsoft did amazing things with win95.
Here is the prologue to your post.
Windows 95 (assuming you don’t have any legacy 16 bit Windows or DOS drivers or software lurking around) pre-emptively multitasks Win32 applications. They do not have to be “specially written” for this to happen, they just have to be Win32.
No version of MacOS Classic pre-emptively multitasks anything. …..
Here is the epilogue:
People who are interested should pick up and read a copy of “Inside Windows 95”. It really is amazing what Microsoft did with Windows 95, and you don’t fully appreciate it until you see the nitty-gritty details.
No where in the post did you mention win32, win16 or DOS legacy support as being the amazing features you were refering to. BTW maintaining legacy API support is not an amazing feat, it is a requirement for most customers. The reason MacOS X has a classic emulator and also the reason why many haven’t moved to OS X. Before you go out of context again by saying ” but win95 had better legacy support than OS X does”, that is not the point. Adding legacy support is not Amazing.
the original poster is wrong in that it is not leagal to make machines that can run osX presently, as Apple does not sell or license it’s ROM to anyone which is what allows a computer to run the MACOS classic or not.
New Machines don’t use an actual ROM chip anymore, but they still need an Apple ROM-IMAGE to run.
The CHRP platform which was to supplent PREP and the old apple hardware, was to abolish the need for the ROM and would have allowed cloners to make machines completely independently from Apple.
The problem with running OS X on non-Apple hardware isn’t the ROM, it’s the EULA. The OS X EULA (and, I’m pretty sure – although I don’t have one to check – later versions of MacOS CLassic like 8.6 and 9.x) forbids running it on hardware that isn’t “Apple branded”.
Dood what is wrong with you.
we are not talking about windows.
Of course the EULA forbits it, but try to go to the Apple store, buy a copy of OSX then go to this site ( https://www.pegasosppc.com/store.php?category=1 ) and buy one of the MotherBoards now build a machine, then get that $129 OSX Panther disk and put it in the optical drive and try to instal it. say yes to the EULA. Wait it’s not working, what is going on…
NO ROM.
dood.
if you need more background go here and read this site from front to back.
http://www.ppcnerds.org/
Then read this one too: http://www.ppczone.org/modules/news/
Now
what where you saying about the EULA again?
Did you use Classic?
Sometimes. It multitasked far too poorly and was far too unstable for my tastes, however (not to mention expensive, slow hardware). I’d feel pretty confident in saying I used it a lot more than the average computer user of the time, but I certainly didn’t use it day to day.
Did you use 95 at the same time?
Generally only for games and when I had to (eg: jobs, classes) – for much the same reason (although it wasn’t anywhere near as bad as MacOS). By that time (~1996 and onwards) I’d moved to NT. Immediately prior (~1993 – ~1996) I was using OS/2.
Can you support your claims with facts or even a little bit of real personal experience?
I gave some examples that anyone with a MacOS Classic capable Mac can check today. MacOS multitasked terribly and almost all backgrounded tasks suffered a *severe* performance impact, even if the CPU usage was extremely low.
The fundamental problem with co-operative multitasking is it requires applications to be well written and “play nice” to deliver a satisfactory multitasking experience. Most MacOS apps (as with, indeed, most apps on similar platforms) didn’t “play well”.
Do you believe that using authoritative statements based on hearsay are enough to make you right?
I believe making correct statements is enough to make me right.
As a Mac user from OS1, an Irix user since vs5, an NT user since 3.5, and a power user (serious graphics and video work) since Mac OS7 in 93, I can say that when cooperative Multi-tasking came to age in OS7.5, it usefully improved my workflow and usefully improved the Mac user experience.
Compared to single tasking, it probably did, you would have no argument from me about that – I still remember moving to Windows from DOS.
However, the comparison at hand is to a (mostly) pre-emptively multitasking environment, not a single tasking one.
Just like any other workstation environment once locked down the classic Mac OS was as functional as NT; and where the stability and memory protection of NT shone, the user interface, lack of plug and play, and the lack of multiple monitor support (this was a part of Mac OS since system 6) really hindered it.
While I’d disagree with the “as functional as” part (the instability aspect, mainly, but also the multitasking part), I agree with you. The UI and “overall experience” of MacOS were (and remain) its primary strengths.
Win95 was not a viable replacement for the Mac because it was not significantly better in any way.
It was *significantly* better at multitasking – particularly under ideal conditions (which weren’t particularly hard to achieve from mid-1996 onwards). IME its mostly-there memory protection also offered better stability than MacOS.
It’s multitasking was a kludge just as that of the MacOS and suffered the same faults, due to legacy constraints, and it was still a DOS based windowing system.
Firstly, it was really only a “DOS based windowing system” in that it was bootstrapped from DOS – Windows 95 used its own device drivers, CPU scheduler, memory allocation, etc. There wasn’t much DOS left once Windows 9x was running. In every meaingful sense, Windows 95 was an operating system.
Secondly, while it’s pre-emptive scheduling was certainly kludgey, it produced *substantially* better real life results than either the contemporary MacOS’s co-operative scheduling or later versions’ nanokernel-but-special-developer-attention-needed model. Any win32 apps were pre-emptively scheduled, without any special procedures being taken by the software developers.
I knew of not one active professional who would use Win95 when they could use NT or Mac OS classic.
I don’t know anyone who would use Windows 95 if they could use NT.
I don’t know any *Mac users* who would use Windows 95 in preference to MacOS (but that also applied to anything that wasn’t MacOS).
I know a hell of a lot of Windows 9x users who tried MacOS and preferred Windows 9x.
Part of the reason was that 95 was not significantly more unstable, but because of millions of registry conflicts, DLL corruptions legacy 16 bit code in DOS, etc; it was a support nightmare.
Windows 9x, running all win32 applications and 32 bit drivers, was a substantially different beast to the Windows 95 environment you are alluding to. I’ll grant, however, that the ideal situation didn’t start becoming fairly common until late ’96 – early ’97.
So if we are going to use subjective statements about these operating systems, at least lets talk about them from our experience and not hearsay.
I am speaking from experience, and my experience says compared to MacOS Classic, Windws 9x’s multitasking was clearly superior and its stability was notably better. That was the only comment I made. NT, Irix, and other OSes you have tried to bring into the equation were *not* a consideration when I made them.
I will tell you this though, when I did first use Win95 I was very impressed with it in relationship to what it replaced, but within the same year os8 came about and erased any real advantage.
OS8 (and 9) still had the same problems with regards to multitasking and stability that OS <8 did. That was my original point – “no version of MacOS classic pre-emptively multitasked anything _usefully_”.
Personally, given the choice between Windows 9x and MacOS Classic, I’d go with Windows every time – because the OS8 UI didn’t have any major advantages to me (indeed, there were many things about it I have a distinct dislike of) and the better multitasking and stability of Windows 9x is more important to me.
OTOH, given the pick of common OSes in, say, 1994/1995, I would (and did) go with OS/2.
95 was a vast improvement over Win3.1.1, but definitely not an improvement over the Mac OS experience; not even by a long shot.
(it would be like saying that a MCD’s burger is a really good burger because it is bought and eaten more than any other burger in the world.)
Not really. Whether or not it was a better UI would be largely a matter of taste. For example, I’ve *never* liked Finder as a file manager and I *hate* the way MacOS task switching switches between “applications” and not “windows”.
What Apple was loosing to and competing against was WinNT, and this was definitely a scary scary time for Apple back then. Especially when so many ISV’s and Vertical Market apps where moving to NT.
I’d strongly disagree MacOS Classic was – in any significant manner – in direct competition with NT. NT was competing with Novell, Unix and OS/2. Macs were only really common in markets NT didn’t have a strong presence (and vice versa). I think their market overlaps were quite small.
When Adobe and Quark made NT versions of their programs on par with the Mac Versions I almost died.
I mean the versions of PS and AI available for Win before that where 2 years behind.
PS3.0 had been out on the mac almost a year, and all of a sudden PS3.0 and AI 4.1 on NT at the same time.
Now that was scary, but that was then.
But most Mac people would still not migrate, regardless. Most Mac users IME use MacOS because of the UI, not the software or technical features of the OS. Hell, there was a fairly large contigent of the Mac community that wouldn’t even move to OS X because it “wasn’t MacOS”.
Indeed, you appear to express much the same sentiments yourself – being more concerned about impact on Apple/MacOS of application parity between NT and MacOS than being able to do your work on a better system.
You were responding to seabasstin about multitasking, yes.
I was using multitasking as an example of how Windows 95 was better than MacOS Classic.
What are the different reasons?
If you know about it, I’m sure you can think of a few.
NextStep was great at multitasking.
I never suggested otherwise.
I don’t know what “modernising” has to do with preemptive multitasking.
Because it’s a feature that was missing from some “old” OSes of the time like MacOS and Windows. Memory protection being another one.
If you meant to say that all the cooperative stuff runs as a single mptask so the end result is the same, I might give you credit. But you made a blanket statement as usual and acussed some one else of switching context, again just as usual.
I didn’t accuse anyone of switching context, I said they didn’t take it into account (and instead chose to pick on a minor semantic point rather than address the general discussion). It’s all well and good MacOS had that weird little nanokernel and the ability to pre-emptively multitask some things if the software developer went to the necessary lengths to make it possible. I’m even happy to admit I’d forgotten it ever existed.
My point was that *in the context of comparing general usage of Windows 95 and MacOS Classic*, MacOS “didn’t multitask anything” because, firstly, the nanokernel didn’t appear until 1999 and, secondly, SFA applications ever took any advantage of it (because, as I said, by that time OS X was at the “real soon now” stage).
Your original post which raptor replied to, started out talking about preemptitive multitasking features in windows 95 and ended with you saying microsoft did amazing things with win95.
Yes. Do you think you can handle the possibility of two different ideas being in a single post ?
No where in the post did you mention win32, win16 or DOS legacy support as being the amazing features you were refering to.
Nor should I need to. At that point – not being in the context of a specific example – I was talking about Windows 95 as an entire system, because it’s only as the entire system – and in the appropriate timeframe – that it becomes impressive.
BTW maintaining legacy API support is not an amazing feat, it is a requirement for most customers. The reason MacOS X has a classic emulator and also the reason why many haven’t moved to OS X. Before you go out of context again by saying ” but win95 had better legacy support than OS X does”, that is not the point. Adding legacy support is not Amazing.
The lengths Microsoft went to with Windows 95 were pretty impressive.
Few OSes and venfors have done as good a job as Windows, or had as many obstacles in the way of doing so.
Here’s one fairly well-known example:
http://www.joelonsoftware.com/articles/fog0000000054.html
“Windows 95? No problem. Nice new 32 bit API, but it still ran old 16 bit software perfectly. Microsoft obsessed about this, spending a big chunk of change testing every old program they could find with Windows 95. Jon Ross, who wrote the original version of SimCity for Windows 3.x, told me that he accidentally left a bug in SimCity where he read memory that he had just freed. Yep. It worked fine on Windows 3.x, because the memory never went anywhere. Here’s the amazing part: On beta versions of Windows 95, SimCity wasn’t working in testing. Microsoft tracked down the bug and added specific code to Windows 95 that looks for SimCity. If it finds SimCity running, it runs the memory allocator in a special mode that doesn’t free memory right away.”
The extent of Windows 95’s legacy support – while still offering most “modern” features like memory protection and pre-emptive multitasking – were quite impressive. Particularly taking into account they had a target platform of a 4Mb 386SX and that it was still just a consumer-level desktop OS, not an enterprise level OS.
Of course the EULA forbits it, but try to go to the Apple store, buy a copy of OSX then go to this site ( https://www.pegasosppc.com/store.php?category=1 ) and buy one of the MotherBoards now build a machine, then get that $129 OSX Panther disk and put it in the optical drive and try to instal it. say yes to the EULA. Wait it’s not working, what is going on…
NO ROM.
dood.
You sure it’s not just hardware drivers or the installer checking for “genuine” hardware ? AFAIK no current Mac has a ROM either – the ROM is distributed with the OS in a simple file image. Then of course there’s Darwin, which is all the low level parts of OS X (ie: the ones that have technical dependencies on the hardware). Not to mention PearPC, that manages to run OS X (albeit incredibly slowly).
“Dood”.
what where you saying about the EULA again?
That it’s the problem. Even when you do get around any hardware support limitations (which really shouldn’t be very hard, with Darwin being Open Source and all) you still can’t legally fire up your copy of OS X on that machine. I imagine even running OS X inside MOL on one of those machines isn’t legal either.
The lengths Microsoft went to with Windows 95 were pretty impressive.
Not really.
” On beta versions of Windows 95, SimCity wasn’t working in testing. Microsoft tracked down the bug and added specific code to Windows 95 that looks for SimCity. If it finds SimCity running, it runs the memory allocator in a special mode that doesn’t free memory right away.”
This is an ugly hack at the very least, not a particular feature to be proud of. Uh gross. This isn’t impressing me actually it is making my stomach turn.
If you call trying to fix app bugs in the OS as features sorry we speak different languages and have different meanings of impressive .
The extent of Windows 95’s legacy support – while still offering most “modern” features like memory protection and pre-emptive multitasking – were quite impressive. Particularly taking into account they had a target platform of a 4Mb 386SX and that it was still just a consumer-level desktop OS, not an enterprise level OS.
Memory protection and multitasking are not paticularly modern concepts. At the scheduler and memory management level there is no distinction between what userland APIs are in user, unless those APIs particularly mess with the internal implementation of said APIs. Exposing internal implementation as a part of the API is bad design to begin with.
Adding a preemptitive scheduler or using the hardware mmu for prtotection is orthogonal to upperlevel apis at least on most well designed OSes.
Pointing out a hack to make one paritcular App work is not helping your cause in making me think that Microsoft actually designed windows well.
Nor should I need to. At that point – not being in the context of a specific example – I was talking about Windows 95 as an entire system, because it’s only as the entire system – and in the appropriate timeframe – that it becomes impressive.
No I don’t find Windows95 impressive for it’s time frame. Next Step was far more impressive around the same time.
Yes. Do you think you can handle the possibility of two different ideas being in a single post ?
You spent the whole post talking about multitasking/ multithreading between Windows and MacOS. What was the other idea?
Here is the rest of your post for your reference. I also took the liberty to highlight your generalizations which raptor and later I took exception to, you turned what you said around later by talking about “how many years after win95 and time of realse”. The word “ever” has a very precise meaning.
MacOS Classic software *does* have to be specially written to “play nice” with other applications and individual applications can easily soak up all available CPU resources or crash the machine in regular day to day operations.
I’m also rather sceptical of “older programs do not have to be rewritten to take advantage of threading, they are automatically threaded by the OS”. I don’t quite see how a program that wasn’t written to be multithreaded can magically become multithreaded just from an OS upgrade.
Windows 95’s multitasking and memory protection is certainly not as good as NT’s or unix’s, and makes substantial reliability and performance sacrifices in the name of legacy support, but it was streets ahead of anything MacOS Classic ever did.
The quickest and easiest way to demo this is to decompress a file in the background under both OSes while performing some typical actions in the foreground (browsing the web, email, word processing, etc). Even under ideal conditions, on MacOS the (backgrounded) decompression will absolutely crawl, but under Windows 95 is will hardly slow down at all. Another good example is backgrounded file copies (particularly over the network).
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/wi…..