What other hardware/Technology is on the horizon?

[Humus]

Ah, using volatile works in MSVC too

The initial goal with gcc wasn't performance, rather just to have a working cross-platform compiler. Today it's available on pretty much every platform already so the goal is changing towards performance. I'm not surprised if commerical compilers can produce quicker code, but I don't think gcc deserves to be called "poor" in terms of optimisations.

 

[DemoCoder]

Yep and I agree with GCC's goals, I just wanted to point out that the overall optimization levels based on rather old techniques before GCC3.0, which includes SSA, basic block reordering with branch prediction, and many other new global iterative techniques.

GCC now needs major work to support SSE2, IA-64, and Opteron "optimally". SSE2 especially. Intel is supposed to be donating code from their compiler to GCC for Pentium4 optimizations.

 

[demalion]

How do you know who it was targeted at without a marketing effort to determine that?

Sometimes it is unproductive to ask such questions without quoting. The first time, the body of text was so large, it served clarity. This time, I gave you a brief reply and you would not be able to ask this question again if you quoted that as it would demonstrate that I've answered.

Maybe when Be's programers were writing the OS, they were targeting consumer (implementing the features they though consumes would want).

You think they were doing something else?

But if their sales never tried selling it to consumer, I would argue that it was never targeted at them by the company.

So in short, they only made it for developers to admire? Do you see the lack of logical coherency in this proposition?

Ultimately, its the marketing strategy rather then programer's intentions that determines where a product is targeted at.

Marketing strategy is part of the picture, not all of it. Without Microsoft's monopoly, they would have had reason for a different marketing strategy. I've proposed to you that the company's overall strategy was targetted at consumers in the circumstances allowed by Microsoft's monopoly. If you don't agree, please explain why, my reasoning is already provided. The reason I mention quoting at the beginning of this post is that all you've done instead is repeat the assertion that they did not without addressing the reasons I gave.

 

[demalion]

Well, this is my last response in this thread since it is going nowhere. You seem unable to grasp that there are several discussions going on at once in these messages:

I've addressed every single question you've posed directly. You routinely ignore the questions I've posed to you, said you couldn't find questions after I directed you them for a 3rd time, and then avoided them yet again when I provided a direct link. Amazing why things aren't getting anywhere, yes?

Discussion #1: You assert BeOS was killed by Microsoft monopoly

What I assert is that Microsoft's monopoly did not allow BeOS to compete on Quality. Your rewording omits the discussion of quality that has been present from the beginning of my discussion.

My Answer: BeOS died in the marketplace because of bad BeOS business strategy. Any launch of a new platform requires loads of money pumped into marketing and sales. BeOS tried a mostly "word of mouth" among hackers campaign.

Well, that answers the question you say I asked, but it neatly skirts around several issues by rewording it with assumptions and omissions integrated into it.
Any monopoly can be overcome with enough money, and it does not require "bad business strategy" (except inasmuch as trying to compete with a monopoly is bad business) to fail at it. Trying to forget about our quality discussion is a good way to circumvent that issue, however, good job.
Monopoly = commodity control. BeOS did directly try to address this commodity control, but the degree of Microsoft's control resulted in their failure. This statement does not imply that BeOS had a bad business strategy, but rewording "my question" and ignoring how our quality discussion relates did allow you to blithely make the claim, didn't it?

Discussion #2: Large apps on old hardware. You assert you can have identical functionality on less capable hardware

Actually, I addressed your list point by point. But by trying to argue instead that the absence of absolutely all of the functionality of, for example, Microsoft Word on the Amiga, ignoring the the disparity in minimum hardware between the two, that Word, Windows, et al are efficient despite your stated support for such being incorrect.

My Answer: Yes, you could go find 5 separate word processors that each does a little bit of what Word does today. But Word includes everything. Localization? Word includes over 50 world languages with grammar/spell checkiing in each. Word can do document scanning. Word has an HTML Editor! Word can export as XML. Word has Asian Layout. Word has a Font Selector Combobox that allows you to see EVERY TRUETYPE FONT AT THE SAME TIME. That means 50+ truetype fonts are loaded and rendered into memory and makes it way easier to preview/select a font. If you can't understand the difference between 10 products that do the same as 1 product leading to the 1 product needing more memory, then I can't help you.

You are trying to introduce the fallacy that Word is doing all of these things at once, and therefore its performance is understandable, while at the same time ignoring that all of the functionality I did address, while requiring more than one application, could be achieved with a slower CPU and less RAM. In fact, it specifically occurred within the same framework you said it could not (286 class hardware). I think the "one application" issue is a red herring to detract from that.

(oh, Word also supports handwriting recognition and Speech recognition as input methods)

So the hand writing and speech recognition code is running when you're not using them? And the performance issues with Word on Windows only occurred after such was added? If this not the case, what relevance is this to what I was addressing?

Discussion #3: BeoS quality vs Windows Quality
My Answer: Unix/Windows NT make different tradeoffs vs BeOS in their design points. Unix/NT carry years and years of legacy along with them. It's not a question of "quality" but "design" architecture.

OK, so design architecture has nothing to do with quality?

There are lots of little embedded realtime OSes like VxWorks, QNX, etc that run many tasks way more efficiently than Unix. Does that make RealTime OSes "better" and "higher quality"?

If they can offer the same functionality as Unix, yes.

Quality refers to the stability, robustness, and performance of a design. You can have two separate architectures that differ remarkably in their design, but both are "quality". The Unix and NT kernels are both quality implementations. Yes, they may not run certain tasks as efficiently as BeOS

So then the comparative quality would be less, wouldn't it? You could have saved a lot of typing by admitting this earlier. Did I miss something?

(but what do I want for SMP on my Server, BeOS or Solaris?)

I thought we were talking about Windows and the consumer marketplace? You aren't doing the bait and switch again are you? Its interesting how Microsoft is trying to leverage their consumer OS marketshare into server OS marketshare. Competing on quality, and not control, I'm sure. Ah, monopolies in action.

but that doesn't disqualify them from being "quality". Unix wasn't designed, for example, to be a high quality MIDI workstation. But BeOS probably falls down in areas where Unices shine. Does BeOS have fine grained security?

Hmm...I was talking about the Windows monopoly, and you bring Unix into it. Since I wasn't trying to say that Unix wasn't the result of competition on quality for its marketplace, this seems a bit silly. Being an open source standard with open APIs, it competes with itself.

So it's not a question of which kernel is higher "quality", they are different, not "better" It depends on what task and what features you care about.

Ah, so there are only differences that are not differences in quality. No wonder a monopoly has no impact on the quality of products, difference in quality does not exist. That seems to be what you are saying.

Windows NT and Unix have much larger code bases than BeOS and support way more hardware devices, so the probability of any one system utility, applet, or driver having a bug is higher. But if you look at the kernels only, Unix and NT are rock solid.

Didn't I say in an earlier post that 2k and XP were stable? Didn't I have some other text around it that you ignored to take this opportunity to repeat back what I'd already said?

What you don't seem to understand is that there are oodles of OSes out there that have nice design features, but those nice design features don't neccessarily sell to consumers.

Actually, I'm saying that a monopoly works to preclude the opportunity for these design features to sell to consumers, and also that features can be emulated by the monopoly at their leisure and their introduction is not driven by the ability of the consumer to experience their benefit (competition) but by the ends to which they serve the profitability of the monopoly.

Developers and enterprise customers may appreciate them, but consumers don't care.

Of course, because for the purposes of your selective argumentation features that appeal to former are mutually exclusive to appealing to the latter.

People have limited attention spans and you need a way to communicate your advantages to them in 5 seconds.

BeOS needs more than to just be technically nice, it needs compelling software and compelling marketing to get people to adopt it over what hey already have.

You're right. I'm saying the problem is that it failed to demonstrate this successfully because of a monopoly, not because the advantage did not exist. This seems too radical a belief for you.

Steve Jobs showed that you can still excite the consumer. Apple made a comeback against Microsoft, and successfully got people to switch.

"made a comeback"? To where?

In any case, there is a reason I was pleased by the rumored possibility that Apple may compete in future on similar hardware (x86 based architecture). I think it will make both Apple and Microsoft provide a better products sooner.

Look at Apple's marketing campaign to see what you need to do and what's possible.

Maybe if Microsoft made software for BeOS or gave them, what was it, 150 million dollars?, Be would have been able to afford it. Apple had marketshare and has been in the process of losing it, because they haven't been able to establish a successful monopoly. As a result, they have had to compete on quality. And it shows.

You think Windows wouldn't be better if Microsoft had as much reason. I don't get why.

You seem to think that this all revolved around monopoly power.

Yes, I've noticed that a common theme is you either don't think Microsoft is a monopoly or that monopolies are disadvantageous to the consumer. I simply think you are wrong, and that you have not given any coherent basis for your belief.

I tried to show you how even FREE products not owned by any company can become dominant and prevent smaller players from gaining traction in the market place.

Hmm...past tense? Didn't do a good job then. Let's discuss this example though.

Case in point: The Apache Web Server. Before Apache 2.0, for many years, the web server had an old "inelegant" architecture (no asynchronous I/O, no threads, no high availability, etc), yet it complete dominates above all other web servers in terms of usage. Many small companies tried to compete with better ones (Roxen, Zeus, etc) but no one would buy their "efficient" design. People still wanted "bloated, old," Apache. (because Apache DID ITS JOB regardles of the "BEAUTY" of the DESIGN)

I guess I'll have to take your word for it that Apache 2.0 was not a timely result of such competition, that Roxen and Zeus really did offer quality advantages over Apache at the time. Let's see what info I can find to support your claims:

This seems to indicate that performance wise Zeus was indeed a winner in 1998. Roxen doesn't look so hot, but this comparison is hardly a complete picture. It also states this:

Don't get too excited over the performance figures. Most of these servers have enough oomph to keep a T1 saturated, running on a lowly 100MHz Pentium or maybe even a 486. Very very few web applications need more power than that. So, the fact that Apache is not that fast shouldn't be of concern to most people.

Which is certainly not representative of Windows as a consumer OS. Microsoft and Windows software needs 500 MHz+ to overcome its basic speed issues, and that is going up with each version (though you maintain it is silly to complain about that). Your analogy begins to break down.

To continue:

Zeus claimed increasing market share in 1999 as a result of their higher quality.

Did they overtake Netscape as they expected to?

As of October 3 they are in 3rd place, ahead of Netscape. Did BeOS perform similarly? No? This disproves the impact of a monopoly how?

Many "BeOS-like" companies tried and failed to dislodge Apache, which has a near monopoly (especially on Unix), even though it isn't even a company.

Hmm...it isn't a company, is freely available, open source...and from this you conclude it is a monopoly in the economic sense Microsoft is. To me that doesn't sound coherent.

That's because it's hard to displace an incumbent, be it a politician, an OS, an application without LOTS of marketing/advertising. Incumbents get free advertising (they are already everywhere), strong brand name recognition (ask person X to think of product Y, what comes to mind first?), and people want to use what everyone else is using or knows. It's the network effort, or Metcalf's Law, or whatever you want to call it. You might be the better candidate for president of the US, but who knows about you?

This rationale of "this is the way things are so there is nothing wrong with it" is supremely circular. I don't find myself surprised that your argument boils down to this.

You can blame Microsoft all you want. I blame Be for ruining BeOS's chances. I blame Commodore for sinking the Amiga. I blame Atari for sinking the ST. And I blame Apple for sinking the original Mac, by putting all their effort into trying to win a Copyright on Graphic User Interface in Court, and less effort finishing Copland/Mac OS9/ OSX/etc.

You blame everyone but Microsoft. That's why your comments make no sense and are ultimately nonsensical.

Apple's OS stagnated while Windows when from 3.1 to 95 ( a TREMENDOUS improvement).

If Apple succeeded in achieving Microsoft's position, computing would be in worse shape. Microsoft's monopoly being better than Apple's would have been does not make Microsoft's monopoly a good thing. Your comments makes it sound as if you'd disagree with this.
As Apple did not succeed in achieving Microsoft's position, it is Microsoft I criticize. I do not want anyone to replace Microsoft in its current place, I want more than one competing. This is the heart of capitalism that you grossly distorted in your post that prompted me to reply

It's really ironic, but before 1995, most of the open-source fanatics (like Richard Stallman of GNU/FSF) were railing against Apple trying to "OWN" the GUI via Intellectual Property protection.

Yes, it is possible for more than one company to be capable of wrong doing.

Meanwhile, no real improvements happened to the core of MacOS. Oh, but they were killed by the Win95 monopoly right?

Actually, you brought up Mac OS, made up my stance on it, and decided to attack it. Works wonders as an alternative to addressing the text I actually typed, doesn't it?

I see BeOS making many of the same mistakes as Amiga. Amiga also had a focus on fast multithreading, and low latency video/audio for the Desktop Video Market. They wanted to be to DTV what Mac was for DTP. But DTV wasn't a compelling consumer market back then. Atari ST, ditto for the MIDI world. Too much focus on their niche, not enough on business and enduser.

Pre-emptive multi-tasking in 1985 on Amiga, 1995 on Windows. To me this supports the conclusion that not having a competitor in their market allows Microsoft to add features at their leisure instead of in response to successful competition, because of other factors besides quality. By the way, it is the PC itself (and the nature of its open competition) that defeated the Amiga and Atari ST, not Windows. That's why I'm not criticizing PC makers, and why BeOS existing on the PC illustrates they had learned from the mistakes of Atari and Commodore.

There is a large block of text I've re-typed a few times now that addresses BeOS more completely. Given your record for answering my questions and comments, I hope you'll understand why I don't bother again.

This discussion is over. There is no way I can convince you that Microsoft isn't evil, has the worst programmers in the world, or that Be had any responsibility in the failure of BeOS.

You're doing it again, putting words in my mouth. This is pretty similar with your obsession with labelling people fanbois. You've done a good job in establishing how comprehensively useless it is elsewhere, another example was not necessary.

 

[DemoCoder]

Actually, the Amiga did not have preemptive multitasking in 1985. Due to an obscure bug in exec.library in pre-2.0 kickstarts, the scheduler would never preempt a task that hadn't yielded even if it had used up its timeslice and a task of the same priority was ready to run. The only way a task could be preemptive was if a task with higher priority was ready to run. This was fixed years later in V39 exec (Kickstart/AmigaDOS 2.0), but most people never knew about it because most apps at some point yield the CPU by waiting for an event/message or doing I/O.

By the way: Resource tracking, memory protection, microkernel, device independent graphics and sound: WinNT 1994, AmigaDOS = never. Yep, I loved running more than 1 task on the Amiga, so that when one of the apps crashed and I got a Guru Meditation, I LOST EVERYTHING.


Demalion, you need to go back to college and study business 101. I'm sorry, but there is simply no common ground here to have a discussion. When you create a product, you simply must spend money to market it whether there's a monopoly in your sector, or not. I think you would agree that there is no monopoly in the automobile market right now. Do you honestly think you could start a company and develop a new car and sell a reasonable percentage in the market without huge advertising campaigns? I've just made a better Honda Civic, but I'm not willing to make a commercial to sell it?

Nobody outside internet geeks has even heard of BeOS and that is simply not Microsoft's fault. If BeOS had a case, they should have taken their case to the people. They went public and raised money in an IPO and should have used that money to crack into the market. Whether it was enterprise, or hollywood, or the music industry, or schools, they atleast had to find some message and some market of people that they could take their message to. There are several OS vendors who have succeeded and they didn't even have to spend a lot of money, they just had to pick the right strategy.


And sorry, but you are simply wrong if you think being "real time" makes something better quality. If I need to run online transaction processing for credit cards, I sure and hell aren't going to run it on a real-time OS. You also simply haven't demonstrated anything bad about NT's implementation as not being quality. You don't understand that SMP/Threading isn't the end-all and be-all. You keep bring up the Amiga, and sure, the Amiga had preemptive multitasking before Windows, it never got a modern scheduler, it wasn't multiuser, it didn't have memory protection, it couldn't free resources lost to a buggy program, in essence, it was worse than other OSes in some ways.

And since most people running apps on MacOS and Windows saw no difference between cooperative multitasking and preemtive multitasking, what did preemptive multitasking buy them? It was a differentiating feature that was hardly useful at the time, since the only reason you need to preeempt something is if you have a long-running compute-intensive background task, like a raytracer running.


Anyway, End of discussion. You win. Ok, Microsoft destroyed BeOS with monopoly power. And Microsoft will continue to destroy any company with shitty management that spends money on technology creation, but nothing on communicating why people should buy it to people. And if Microsoft goes out of business or is destroyed by the government because of whiners, then Apple or Linux will destroy BeOS-like companies who simply don't have a marketing message.

In 5 years, you will still be using MS products, BeOS will be as dead as AmigaDOS, and this discussion will have been a huge waste of time. Hopefully by then, you took some business classes at your local community college.

 

[dksuiko]

Demalion,

Imagine a new company developing a processor called NoobProcessor and it competed with Intel and AMD. It was faster than both the Pentium 4 and the Athlon - in every scenario. It was technically more advanced in every way, a revolutionary piece of technology. On top of that, it was offered at a comparable price. However, it did not run any x86 applications. It only offered a fraction of the software when compared to the number of software offered by an x86 processor. A large number of the major developers chose not to develop for it since its userbase was so small.

Now.. If this processor failed, nobody bought it, and its company shut down - would you blame Intel/AMD for its failure? It was technically more advanced than both Intel and AMD's products, but it offered nothing to the consumer since it had no interesting software to run on it. I see this scenario the same way I see the BeOS vs. Windows scenario. It might have a more advanced technical background, but nobody gives a damn about that if it doesn't run the software they want to run and do the things they want it to do.

What good is a car with a fuel efficient engine if it has no wheels? What good is a GeForceFX or a Radeon 9700 if the card has no video out?

-dksuiko

 

[DemoCoder]

dk,
Nice analogy. Such a company probably could take the NoobProcessor into specialized markets like mobile devices, or renderfarms, IF they pursued the right partners. Just look at StrongARM, or MIPS. In fact, they could even get into the x86 market if they used a strategy like Transmeta of translating x86 instructions into their own instruction set.

The problem is, the NoobProcessor will never succeed if they simply put it "out there" and have the attitude "if we build it, they will come". NoobProcessor must be marketed to the right market, and must spend money on ads so that you know it exists.

Windows NT kernel was architected originally so that the OS could have many "personalities". That is, almost all of the Win32 APIs are just user-mode "skins" on top of the kernel. There was also going to be a OS/2 skin, and a POSIX skin, etc. Exposing the core NT kernel thru multiple look-and-feels. (things changed in NT4 to make this less likely)

Lindows is a similar strategy, which is to replace the kernel with Linux, but keep the Windows APIs to maintain application level compatability.

Apple took a similar approach with OS X, replacing the core kernel with BSD Unix, but maintaining the old Mac APIs, as well as some updated new APIs. That way, OS X can run any BSD app, plus any Mac OS9 app, plus any Carbon app.

Done right, Be would have built an emulation layer for Windows apps on top of the BeOS, so that it could run BeOS apps, but also run any Windows app as well. Only then could they convince PC suppliers like Dell to ship BeOS machines. Otherwise, it's a waste of Dell's money, and complete management suicide.

 

[demalion]

Demalion,

Imagine a new company developing a processor called NoobProcessor and it competed with Intel and AMD. It was faster than both the Pentium 4 and the Athlon - in every scenario. It was technically more advanced in every way, a revolutionary piece of technology. On top of that, it was offered at a comparable price. However, it did not run any x86 applications. It only offered a fraction of the software when compared to the number of software offered by an x86 processor. A large number of the major developers chose not to develop for it since its userbase was so small.

Now.. If this processor failed, nobody bought it, and its company shut down - would you blame Intel/AMD for its failure? It was technically more advanced than both Intel and AMD's products, but it offered nothing to the consumer since it had no interesting software to run on it. I see this scenario the same way I see the BeOS vs. Windows scenario. It might have a more advanced technical background, but nobody gives a damn about that if it doesn't run the software they want to run and do the things they want it to do.

What good is a car with a fuel efficient engine if it has no wheels? What good is a GeForceFX or a Radeon 9700 if the card has no video out?

-dksuiko

The problem with the analogy is that companies do this because the x86 instruction set and its behavior is not a secret. Windows functionality doe not parallel this.

The reason I find it extremely frustrating that you and DemoCoder consider this example apt is because I've typed this out in detail over multiple paragraphs prior to this. Like where I stated that as a result of the anti-trust settlement, Microsoft is called upon to share its APIs and related technical information (as in they did not do this before), and that if that this had happened 5 years earlier you'd be right. It didn't, so you're not. I went on to express my concern that the Palladium initiative may preclude this having the desired impact in the future, but we'll have to see how it works out.

I used words almost identical to these. It seems to me it addresses your comment directly. If you find fault with it, it would have saved me considerable aggravation if you'd quoted it and addressed it directly.

Please don't continue to make me re-type my text over and over.

 

[DemoCoder]

Too bad you're wrong. Intel withheld the full specs on the 386, the complete behavior of the x86 execution on the 386 was secret and AMD didn't have the technical details neccessarily to make a complete clone of the 386. AMD was forced to reverse-engineer the 386 in a clean room, the same way the original PC BIOS was cloned because IBM refused to release the full specs.

In every OS there are documented public APIs (interface) and there are undocumented private functions (implementation). Microsoft documents the public API and there is enough reverse engineering of the internals to build a reasonable emulator. There is already Wine, WineX (directX on Linux), cross-platform COM, and even Windows Media player on Linux. Many Office apps run under Wine. And certainly third party Windows apps would run under the public APIS, since by definition, they don't have access to the secret stuff either.

A large number of linux device drivers were reverse-engineered because IHVs typically refuse to give out the details on the underlying registers on the card.

The fact that Microsoft doesn't fully document all the functions in Windows is irrelevent. Every single OS company I have known has done this because of APIE/OO. It's good software design 101, since you don't want programmers calling private functions that may change or go away in the future. Abstraction.

 

[MfA]

Which doesnt exactly jive with their own application software being dependent on such functions.

 

[DemoCoder]

I have written windows software that depends on APIs for which I have no documentation. There are boatloads of COM objects registered in Windows for which there is no documentation, because they are for internal use only. I had to reverse engineer them because the Windows developers didn't think to expose them the first time around. A good example: On windows CE, the Pocket IE COM doesn't expose the DOM of the current web page in the IWebBrowser interface. I had to hack it. Why not? Because the DOM is always changing, and either the developers of Pocket IE didn't think you needed this feature, or thought not to include it until the DOM implementation stabilized. In any case, I needed access to the private DOM implementation IMMEDIATELY for a customer demo, so I had to hack it. It's bad to rely on this, because it will probably break in the next version of pocket ie.


Typically you will notice in large software that over time, the interface has to expand to in each version to keep up with the requests of developers for all the features that the original design didn't think of.

In the case of Office calling a hidden API (do you have any info on which one), is this a case of deliberate INTENT to lock others out on the part of MS, or is it merely one of the Office programmers calling up one of the Windows programmers and saying "hey, could you change this method from public to private in foo.h, or can you add this method getFoo() to return Foo?" I question whether it is deliberate strategy, and all MS programmers are trained to look for ways to lock out competitors, or whether it is just incompetence, poor programmer documentation, last-minute API hacks added, etc.

Sure, MS Applications Group has an advantage in that they can get MS Windows Group to add features to support things they need to do, whereas a regular developer would have to go thru slower channges to get changes made to the Windows API. We hear the same speculation about NVidia, or ATI, or some other IHV having special and unfair control over what goes into, say, DX8/9/10.

But I simply do not buy the argument that this is why Office "won". Perhaps you can point out a compelling feature of MS office that relies on these secret APIs (couldn't be implemented otherwise) and was the thing that made consumers switch from LotusSuite or WordPerfect Suite, etc. It might be "unfair" to outside developers, I'd agree, but I think Office beat the competition not because of some features that were reliant on secret APIs, but because of the way Microsoft sold it.


In hindsight, people go back and try to find out "dirty tricks" Microsoft used to kill competiting applications like secret apis, or deliberate bugs (the Dr DOS allegation), but I think the reality of these failed software packages is more mundane: aggressive licensing, OEM deals, luck, right place at right time, etc. Microsoft licenses basically made it hard for OEMs to not use them once they bought into MS, and MS cast such a net over all kinds of players, like ISPs, etc that further put up barriers.

I think the reality of the situation is that Microsoft salesmen, and marketers crafted very creative marketing/licensing schemes that outflanked many of their competitors. They were punished and restrained by the justice department for this. I think all the other who-hah about undocumented APIs, anti-competitor BUGs, etc are people attributed malice and Machivellian motives to what in reality is probably just laziness and incompetence.

 

[demalion]

Actually, the Amiga did not have preemptive multitasking in 1985. Due to an obscure bug in exec.library in pre-2.0 kickstarts, the scheduler would never preempt a task that hadn't yielded even if it had used up its timeslice and a task of the same priority was ready to run.The only way a task could be preemptive was if a task with higher priority was ready to run. This was fixed years later in V39 exec (Kickstart/AmigaDOS 2.0), but most people never knew about it because most apps at some point yield the CPU by waiting for an event/message or doing I/O.

Hmm, I believe that was actually the v36 exec, wasn't it? And they waited until 2.0 and for some mysterious reason didn't include a SetPatch fix for this in the meantime?

I was happily multi-tasking, including a clear memory of broken programs I couldn't close running in the background and no input code, and running a ray-tracer in the background. I never saw any sign of this bug, atleast that I could recognize.

Because I see no reason a SetPatch wouldn't have fixed this prior to v36, I really would appreciate some substantion.

By the way: Resource tracking, memory protection, microkernel, device independent graphics and sound: WinNT 1994, AmigaDOS = never. Yep, I loved running more than 1 task on the Amiga, so that when one of the apps crashed and I got a Guru Meditation, I LOST EVERYTHING.

Hmm...well, yes there was device independent graphics and sound on the Amiga, which is why I asked when you left the Amiga. Yes, I'm quite sure. That's why there are Amiga sound and graphics cards. I have a graphics
card sitting in my old Amiga right now.

Yes, there was a memory protection API, and virtual memory. The problem was that it required you write for it, and the virtual memory system was horribly slow.

But it sounded more dramatic the way you said it.

Demalion, you need to go back to college and study business 101.

Oh, I see, and your attack on the Amiga's pre-emptive multi-tasking and inaccurate statements support this how? I suppose they are a suitable alternative to actually addressing any significant and pertinent stance or question I provide.

I'm sorry, but there is simply no common ground here to have a discussion. When you create a product, you simply must spend money to market it whether there's a monopoly in your sector, or not.

You know, ignoring the multiple paragraphs of where I addressed this assertion does not mean I did not. When you did this before, I typed it again. You replied, omitting it, and made the assertion again. It gets tiresome.

I could quote your car analogy, which is as broken as your burger analogy, and for similar reason. But then when I gave my reasons, you'd respond ignoring them, and then make another analogy just as broken. I know, I've just seen you do it again.

A hint: a car can use any road and go to any destination regardless of vendor.

Nobody outside internet geeks has even heard of BeOS and that is simply not Microsoft's fault.

Hmm...what a lovely blanket statement. What is an "internet geek"? Someone who uses the Internet? So Internet advertising doesn't count?

If BeOS had a case, they should have taken their case to the people. They went public and raised money in an IPO and should have used that money to crack into the market. Whether it was enterprise, or hollywood, or the music industry, or schools, they atleast had to find some message and some market of people that they could take their message to. There are several OS vendors who have succeeded and they didn't even have to spend a lot of money, they just had to pick the right strategy.

So you're saying they didn't try to crack into "the market"? *sigh* You don't recognize that Microsoft's monopoly impacts any companies attempt to do that, so I suppose it is understandable.

What's your list of successful OSes? Would that be ones without shrinking or stagnant marketshare?

And sorry, but you are simply wrong if you think being "real time" makes something better quality.

What the hell kind of crap are you making up now? Quote my text where I said I think "being real time makes something better quality"? Do you have any other tactic than making up my stance for me?

If I need to run online transaction processing for credit cards, I sure and hell aren't going to run it on a real-time OS.

!?!?!?

You also simply haven't demonstrated anything bad about NT's implementation as not being quality.

That is a nice comparative fallacy. You couldn't have quoted when I clarified why before and gotten away with doing it again, so it is naturally convenient for you not to. It went something like "I did not say NT's implementation was poor, but for these reasons BeOS's implementation is better", and then proceeded to give reasons.

You don't understand that SMP/Threading isn't the end-all and be-all.

And where did I say it was "the end-all and be-all"? Do you get tired of making up my arguments for me?

You keep bring up the Amiga,

Well, actually, I mentioned it as an example of an OS that demonstrated functionality desirable to have on Windows to address your assertion that Windows has not been lacking in comparative capabilities during its evolution. Then I used my experience on it to attack a false assertion concerning what was possible on "286 class hardware" you made to try to dismiss the idea that Windows and Windows software was slow.

What I "keep" doing is addressing your assertions on the Amiga when you bring up something new. Most of which does not have to do with my reasons for bringing it up, like most of your text in this past discussion.

and sure, the Amiga had preemptive multitasking before Windows,
it never got a modern scheduler, it wasn't multiuser, it didn't have memory protection, it couldn't free resources lost to a buggy program, in essence, it was worse than other OSes in some ways.

Well, functionality for all of these existed, with the shortcomings I addressed earlier. Did you ever use a 68030 or beyond on an Amiga? Did you ever use Executive (I think that was it) which offered a new scheduling model?

And since most people running apps on MacOS and Windows saw no difference between cooperative multitasking and preemtive multitasking, what did preemptive multitasking buy them? It was a differentiating feature that was hardly useful at the time, since the only reason you need to preeempt something is if you have a long-running compute-intensive background task, like a raytracer running.

If you can't make it go away, marginalize it. By the way, I notice you didn't touch my web browsing example. *shrug*

Anyway, End of discussion.

You said that before.

You win. Ok, Microsoft destroyed BeOS with monopoly power. And Microsoft will continue to destroy any company with shitty management that spends money on technology creation, but nothing on communicating why people should buy it to people. And if Microsoft goes out of business or is destroyed by the government because of whiners, then Apple or Linux will destroy BeOS-like companies who simply don't have a marketing message.

More coherency. I've addressed every misguided sentiment included here before, by directly quoting and answering. You simply excised it from your universe, and wasted my time.

In 5 years, you will still be using MS products, BeOS will be as dead as AmigaDOS, and this discussion will have been a huge waste of time. Hopefully by then, you took some business classes at your local community college.

Yay, the parting Really Mature Zinger! *blink*
For a minute I thought this was the Usenet. *shudder*

 

[demalion]

Too bad you're wrong. Intel withheld the full specs on the 386, the complete behavior of the x86 execution on the 386 was secret and AMD didn't have the technical details neccessarily to make a complete clone of the 386.

Yes, they did, when they tried to establish a monopoly. And didn't AMD fail to succeed at the endeavor, and ended up being plagued by incompatability? Didn't AMD succeed only after they gained a license and x86 specs from Intel? Didn't Intel lose in a case to try and prevent x86 processor clones? Aren't Transmeata processors a result of this?

AMD was forced to reverse-engineer the 386 in a clean room, the same way the original PC BIOS was cloned because IBM refused to release the full specs.

Yes, and this is a good thing?

In every OS there are documented public APIs (interface) and there are undocumented private functions (implementation). Microsoft documents the public API and there is enough reverse engineering of the internals to build a reasonable emulator. There is already Wine, WineX (directX on Linux), cross-platform COM, and even Windows Media player on Linux. Many Office apps run under Wine. And certainly third party Windows apps would run under the public APIS, since by definition, they don't have access to the secret stuff either.

My goodness. A discussion about WINE! Wait a second, I addressed this exact parallel already! I could repeat it, but that hasn't worked before.

A large number of linux device drivers were reverse-engineered because IHVs typically refuse to give out the details on the underlying registers on the card.

And this works for simple things, and hasn't worked for complex things.

The fact that Microsoft doesn't fully document all the functions in Windows is irrelevent.

Then why was this addressed in a ruling of the anti-trust settlement?

Every single OS company I have known has done this because of APIE/OO. It's good software design 101, since you don't want programmers calling private functions that may change or go away in the future. Abstraction.

Yep, I'm sure that's why they did it.

EDIT: I am struck by how consistently you are eager to lay blame everywhere except on Microsoft. I just saw another post where you spend several paragraphs working to do exactly that again. Doesn't it strike you as a bit ridiculous?

 

[DemoCoder]

Because I see no reason a SetPatch wouldn't have fixed this prior to v36, I really would appreciate some substantion.

Go search in the old Rom Kernel Manuals or 2.0 release notes yourself. It's a fact. I use to converse with Carl Sassenrath on BIX, and I distinctly remember his surprise when they caught this bug in 2.0. This kind of thing is common in realtime OSes. The Mars Lander's little robot crashed permanently because of a priority inversion bug in the RealTime OS they used, and priority inversion is something every CS grad student learns.

Linux's scheduler had a poor implementation bug in it that wasn't exposed for years until people starting running large servers on it. Race conditions are notoriously difficult to find.

Hmm...well, yes there was device independent graphics and sound on the Amiga, which is why I asked when you left the Amiga. Yes, I'm quite sure. That's why there are Amiga sound and graphics cards. I have a graphics
card sitting in my old Amiga right now.

I left the market after Commodore went out of business. As of that time, AmigaOS did not include a device independent graphics or sound driver standard. Such efforts were third party hacks, not supported by the OS. There was no AmigaDOS equivalent of Windows GDI or DirectDraw for 24-bit cards.

Yes, there was a memory protection API, and virtual memory. The problem was that it required you write for it, and the virtual memory system was horribly slow.

There was no AmigaOS (built in) memory protection api for isolating applications/tasks from one another. AmigaOS's entire architecture prevented the concept, since the very way that Amiga multithreading worked was by having threads pass messages by directly modifying other threads. Everything in the Amiga worked this way, the device drivers, the DOS/handler system, the intuition.library, etc. Moreover, there was lots of global datastructures that every amiga process modified that would be broken by memory protection and many of those were inline MACROS that couldn't be trapped by patches. This is the same reason why the Mac couldn't support it, and why Windows NT had to run Win16/3.1 apps in a separate virtual machine. If you whole architecture relies on apps modifying global datastructures directly, you're screwed.

Yes, there were third party apis, but none of them could solve the inherent problems, and certainly none of them were there in 1985, or 1995 for that matter. Up until I left the Amiga (around when Commodore went bankrupt), the AmigaOS, as shipped by Commodore on the very newest machines, was a broken architecture in comparison to other OSes of the day.

So you're saying they didn't try to crack into "the market"? *sigh* You don't recognize that Microsoft's monopoly impacts any companies attempt to do that, so I suppose it is understandable.

It's still no excuse for not even trying. I mean, if I follow your logic, Be, Inc should have realized they never should have written BeOS in the first place because MS has a monopoly.

But if they spent resources on writing the OS, the least they could have done is spent resources trying to sell it. Yes, MS's presence has an impact. So what? So BeOS should just give up and not spend $1 on promoting their product?

I mean, do you just LOVE Be,Inc or something? You can't find anything to criticize about their handling of the company?

What's your list of successful OSes? Would that be ones without shrinking or stagnant marketshare?

Apple (expanding), Solaris, Linux, PalmOS, oodles in the embedded mobile space (WindRiver, QNX, Chorus, etc), Geos/Psion/Symbian, Cisco IOS. Start a company, concentrate on a niche first, don't go for the desktop all at once. Establish a beachhead.

Apple has so infested hollywood, graphic artists, and campus life that I'm frankly sick of getting into arguments with these people as to why I'd rather use Adobe Photoshop on my PC instead of buying a Mac.

It went something like "I did not say NT's implementation was poor, but for these reasons BeOS's implementation is better", and then proceeded to give reasons.

Do you think that there are any things that BeOS does poorly that Windows NT does better?

And where did I say it was "the end-all and be-all"? Do you get tired of making up my arguments for me?

Because Windows NT does many things that BeOS does not, yet all you seem to care about is SMP, multithreading, and IPC when talking about "BeOS's implementation is better". For example, BeOS offers no concept of users, security credentials, and seems to lack any consideration of security all together.

You keep bring up the Amiga,

Well, functionality for all of these existed, with the shortcomings I addressed earlier. Did you ever use a 68030 or beyond on an Amiga? Did you ever use Executive (I think that was it) which offered a new scheduling model?

I had an Amiga's from the 1000 up through the 3000 and 1200. None of the operating systems from Commodore EVER had these features. I don't consider a 3rd party OS patch to be a feature of the OS. And even the 3rd party patches you bring up (like third party virtual memory, MP, retargable graphics, etc) well all major hacks that were not implemented well and were't 100% compatable.

Again, you view history through rose colored glasses. You remember only the good, forget the bad. You correctly assign credit to the Amiga for having modern multitasking features in 1985, but forget it lacked modern multitasking requirements like process isolation and VM, or a device/hardware abstraction layer.

You seem to love multitasking/processing and that's the major feature you care about. Windows NT has features that other people care about. BeOS isn't a better implementation just because it is differnet or concentrates in one area.

The security flaws in BeOS architecture alone would come under harsh criticism today from most architects.

 

[DemoCoder]

I don't blame Microsoft for killing BeOS. I think OS X and Linux are better, 'nuff said. For you, you want to blame every competitor's failure on Windows. It's possible for someone to fail in the software business and not be simply because of Microsoft's position.


90-99% of all new businesses fail. Be failed first and foremost because as a business, they sucked. Look, I am a technologist, I've been doing it for almost 20 years. I'd wager that I know more about operating system architecture than you. And the fact of the matter is, I wouldn't use BeOS. No consider the average consumer: they know absolutely nothing about technology. Now how the hell would they know to even buy BeOS? You think this is Microsoft's fault? How come people BUY LINUX which is FREE if you know how to download it.


You are a BeOS zealot, plain and simple. You're not looking objectively at it, the same as those fools in comp.sys.amiga.advocacy or comp.sys.mac.advocacy.

People purchase Linux, sometimes for $100 a pop. Ask yourself why they buy it, but they don't buy BeOS. Hint, Microsoft has nothing to do with it. I bought BSDI when it came out. I bought FreeBSD. All of these had better advertising in trade journals than BeOS. BSDI cost me $250.


Shouldn't Redhat have realized the inherent pointlessness of trying to advertise and sell an OS against Microsoft? So why did Redhat do so much marketing?


Take a hard objective look at the way Be tried to push BeOS to the consumer.

 

[demalion]

Because I see no reason a SetPatch wouldn't have fixed this prior to v36, I really would appreciate some substantion.

Go search in the old Rom Kernel Manuals or 2.0 release notes yourself. It's a fact. I use to converse with Carl Sassenrath on BIX, and I distinctly remember his surprise when they caught this bug in 2.0. This kind of thing is common in realtime OSes. The Mars Lander's little robot crashed permanently because of a priority inversion bug in the RealTime OS they used, and priority inversion is something every CS grad student learns.

What you are proporting is a fact is that the Amiga kernel did not pre-emptively mult-task until this bug was fixed. This followed from how you justified pre-emptive multi-tasking being "*BROKEN*" (asteriks and capitalization yours) as the result of an "obscure" bug. To me, it sounds like you are distorting the representation of this issue to suite marginalizing the advantages Amiga OS offered, and your statements here further support this conclusion.

I then asked for substantion. Your response is for me to go find it myself. Well, I could unpack my Amiga, set it up, and perhaps I could find the info, but I tend to expect I'll find the issue does not fit the picture you paint of it as disqualifying the Amiga as a pre-emptive multi-tasking operating system. It does not seem a good time investment.

It is a pity...if I were you, I suppose instead of explaining what I'd have to do and why I'm not as part of attempting reasonable discourse, I could propose that since you are making the claim the burden of proof is on you, or some other useless and pointless construction suited to my convenience. Instead I waste copious amounts of time repeating myself.
*shrug* Par for the course, I guess.

Linux's scheduler had a poor implementation bug in it that wasn't exposed for years until people starting running large servers on it. Race conditions are notoriously difficult to find.

Did it disqualify it as pre-emptively multi-tasking as well?

Hmm...well, yes there was device independent graphics and sound on the Amiga, which is why I asked when you left the Amiga. Yes, I'm quite sure. That's why there are Amiga sound and graphics cards. I have a graphics
card sitting in my old Amiga right now.

I left the market after Commodore went out of business. As of that time, AmigaOS did not include a device independent graphics or sound driver standard. Such efforts were third party hacks, not supported by the OS. There was no AmigaDOS equivalent of Windows GDI or DirectDraw for 24-bit cards..

Well, your specific statement was something like "Windows NT = 1994, Amiga = never". Which you've just clearly established was wrong. Glad we could clear that up. And, actually, this was integrated into later versions of the OS. There was even a CPU abstraction layer integrated into the OS later. Does not quite fit "never" does it? But let's pretend that's not what you said.

Yes, there was a memory protection API, and virtual memory. The problem was that it required you write for it, and the virtual memory system was horribly slow.

There was no AmigaOS (built in) memory protection api for isolating applications/tasks from one another.

Is this what you said? No? So when I addressed what you said, was I disagreeing with this?

AmigaOS's entire architecture prevented the concept, since the very way that Amiga multithreading worked was by having threads pass messages by directly modifying other threads. Everything in the Amiga worked this way, the device drivers, the DOS/handler system, the intuition.library, etc.

This is true of many versions of the Amiga OS. If you'd restricted your comment to those versions and stating built in, I wouldn't have argued that you were wrong. What you said was "supported", and 3rd party expansion was the most important aspect of Amiga OS (that benefit from flexible design you conveniently insist does not exist).

To satisfy your new "built in" criteria, atleast some of this 3rd party functionality was integrated as the OS continued to be developed. Hence, again, why your "never" was simply incorrect.

Moreover, there was lots of global datastructures that every amiga process modified that would be broken by memory protection and many of those were inline MACROS that couldn't be trapped by patches.

That's what the replacement libraries for these functions was for. That is why I said you had to program specifically for this...rather clearly before. Oh, but I couldn't have, I think Amiga OS is perfect.

Go ahead and have your spin on it.

This is the same reason why the Mac couldn't support it, and why Windows NT had to run Win16/3.1 apps in a separate virtual machine. If you whole architecture relies on apps modifying global datastructures directly, you're screwed.

I'm familiar with the many shortcomings of Amiga OS, and what amazed me was how its design principles allowed them to be overcome with the same fundamental architecture.

Yes, there were third party apis, but none of them could solve the inherent problems,

They couldn't solve the inherent problems? No, actually they did specifically solve the problems. That was what the purpose of these functionalities were.

In regards to the statement I specifically addressed (Win NT 1994, Amiga never), which you seem desperate to try to bury, I'll mention that there is an Amiga OS 3.9 (and maybe 4.0, not sure) beyond the 3.0 you mention, which is another situation that you seem to practice the "I didn't see it, so it never existed" philosophy with.

and certainly none of them were there in 1985, or 1995 for that matter. Up until I left the Amiga (around when Commodore went bankrupt), the AmigaOS, as shipped by Commodore on the very newest machines, was a broken architecture in comparison to other OSes of the day.

"broken architecture" *shrug* OK, I'll let your amended statement stand I guess, it isn't productive to talk to someone who thinks their opinions stand as fact without justification.

So you're saying they didn't try to crack into "the market"? *sigh* You don't recognize that Microsoft's monopoly impacts any companies attempt to do that, so I suppose it is understandable.

It's still no excuse for not even trying.

They did try. I explained exactly how. You've ignored it atleast 4 times running.

I mean, if I follow your logic, Be, Inc should have realized they never should have written BeOS in the first place because MS has a monopoly.

Yeah, ok, that's what I said...sure.

But if they spent resources on writing the OS, the least they could have done is spent resources trying to sell it

They did. Or we can follow your reasoning and they didn't, since it is obviously more logical. What was I thinking?

Yes, MS's presence has an impact. So what?

I said MS's monopoly had an impact. A competitor has an impact, so does a monopoly. A brick has an impact, so does a feather. By dropping the word monopoly, and using only the word impact, you seem to seek to circumvent every facet of the discussion relating to that to carry forward your statements. A common theme.

So BeOS should just give up and not spend $1 on promoting their product?

From this, it would seem you are saying BeOS did not spend $1 on promoting their product. I addressed this already. The text is still there. You routinely ignored it when I repeated it several times, and here we are, you are ignoring it again. You are a highly efficeint waster of my time.

I mean, do you just LOVE Be,Inc or something? You can't find anything to criticize about their handling of the company?

I realize it is more convenient for you to ignore every time I mention the word monopoly, and refuse to discuss every point I raise concerning Microsoft's monopoly, and instead propose this. But it is not convenient or productive for me to hold a conversation with a person who repeatedly does such.

What's your list of successful OSes? Would that be ones without shrinking or stagnant marketshare?

Apple (expanding), Solaris, Linux, PalmOS, oodles in the embedded mobile space (WindRiver, QNX, Chorus, etc), Geos/Psion/Symbian, Cisco IOS. Start a company, concentrate on a niche first, don't go for the desktop all at once. Establish a beachhead.

Hmm...sort of like BeOS. Oh, wait, they didn't run full page advertisements. But wouldn't that mean they simply didn't "go for the desktop all at once"? Well, since obviously by targetting developers and multimedia functionality, they were not trying to establish a niche, I guess this follows naturally.

Oh, wait, only successful companies' attempts count. And Microsoft's monopoly has nothing to do with the ability of a company to succeed. This construction of thought does work neatly to maintain that a monopoly does not hurt competition, so atleast you are consistent in that regard.

Apple has so infested hollywood, graphic artists, and campus life that I'm frankly sick of getting into arguments with these people as to why I'd rather use Adobe Photoshop on my PC instead of buying a Mac.

Apple succeeded this much, so Microsoft's monopoly has not had an impact on them adn their ability to compete. I see, your logical model is consistent atleast.

It went something like "I did not say NT's implementation was poor, but for these reasons BeOS's implementation is better", and then proceeded to give reasons.

Do you think that there are any things that BeOS does poorly that Windows NT does better?

At the time BeOS was offered? I'd say no. Or are you going to discuss Server OSes again?

And where did I say it was "the end-all and be-all"? Do you get tired of making up my arguments for me?

Because Windows NT does many things that BeOS does not, yet all you seem to care about is SMP, multithreading, and IPC when talking about "BeOS's implementation is better". For example, BeOS offers no concept of users, security credentials, and seems to lack any consideration of security all together.

I'd say this feature is an advantage of NT, and it would have interesting to see how it fared in comparison to BeOS's advantages. Hmm...looking back, how important did consumers find this feature in the marketplace when BeOS was competing? Would it have competed well with the benefits I've already mentioned? Do you presume to know? I don't. I do presume that due to this monopoly we didn't have a chance to find out.

Hence Microsoft brought Windows XP as a consumer OS out at their leisure.

Or are we discussing server OSes again? I've addressed several aspects of your server OS comment before...or maybe I'm just imagining it.

I do note that you glibly ask me for illustration about why the advantages of BeOS affect the consumer, and when I answered you ignored my answers, and discounted them. Yet here you are blithely using the advantage(s?) of Windows NT (not 2k or XP) as a server OS without any of the justification you demanded. I suspect because it would be in inconvenient to acknowledge Windows 95 or Windows 98.

And this is before even the issue of how difficult it would have been to add this functionality to BeOS comes up.

You keep bring up the Amiga,

Why did you quote this and not the text all around it? Trying to misrepresent something, or a simple editing mistake?

Well, functionality for all of these existed, with the shortcomings I addressed earlier. Did you ever use a 68030 or beyond on an Amiga? Did you ever use Executive (I think that was it) which offered a new scheduling model?

I had an Amiga's from the 1000 up through the 3000 and 1200. None of the operating systems from Commodore EVER had these features.

You didn't see it, it doesn't exist. That is a very convenient world view. Did I say Amiga OS, or "operating systems from Commodore", by the way? Not trying to slip a fast one by me, are you?

I don't consider a 3rd party OS patch to be a feature of the OS.

OK. Didn't stop me from using the features though. I guess I was imagining that.

How about it being a feature of the OS that this could be added on at all?

*shrug* I guess for your purposes, you've succeeded in throwing enough text to obscure the statement I've addressed.

And even the 3rd party patches you bring up (like third party virtual memory, MP, retargable graphics, etc) well all major hacks that were not implemented well and were't 100% compatable.

"major hacks that were not implemented well and weren't 100% compatible". Well, except for virtual memory (and I don't have personal experience with Amiga OS on a Power PC to know how well it was implemented there) I'd say all of those are unsupported opinons on things you've even stated you've never seen for yourself. Witness your ending the Amiga OS experience at 3.0 and concluding what things were "EVER" done.
Does that "'shouting' makes it true" concept work for you in everyday life?

Again, you view history through rose colored glasses. You remember only the good, forget the bad.

Umm. OK. Sure. Amiga OS was perfect. I know in your mind that's all I've been saying, so I guess this saves typing.

You correctly assign credit to the Amiga for having modern multitasking features in 1985,

You finally admit this? Go back a few pages and plug this into our early conversation. Clear anything up for you?

but forget it lacked modern multitasking requirements like process isolation and VM, or a device/hardware abstraction layer.

"Windows NT=1994, Amiga OS = never". Yeah, it is my vision that is selective. I guess we might get as far as acknowledging your statement was incorrect if we continued another few pages, eh?

You seem to love multitasking/processing and that's the major feature you care about. Windows NT has features that other people care about. BeOS isn't a better implementation just because it is differnet or concentrates in one area.

A nice summary...omitting most of what I actually said. Again, redefining the context to ignore the entire discussion about monopoly and its role in defining BeOS's failure, yet insisting blindly that ipso facto Windows is better. Any arguments to the contrary get ignored, and your own arguments, naturally being right, do not require even the rudiment of support you insist on from others (not that insisting on it matters, you simply ignore it when provided).

The security flaws in BeOS architecture alone would come under harsh criticism today from most architects.

Yep. And I'm sure since, unlike other people, you don't have to guess, but can simply know, you know it would not have had it today.

I'll do you the favor of practicing what you tried to do, what, 3(?) times now, and actually ending our discussion. Every post you've made wastes my time, and I'm frankly sick of it. Go ahead and raise a fuss, I'm not "waving my hands" and posturing for dramatic effect, but because I am simply tired of continuing this, and have absolutely zero reason to believe anything will change on your end. I'll leave your other post, and let you reply to this one and suitably mangle the point again to have a last word that is to your liking.

Meanwhile, there are others who have raised issues and entered the discussion. Perhaps they can address the text you've ignored previously, or the posts of theirs I've addressed. I'm taking a break for a while, but I'll check back.

 

[Saem]

I believe the way that MS hurts other companies or competition is in the following manner.

First, they're distribution contracts with OEMs. This makes it very hard for other companies to get deals with them. I believe they were poorly punished for such practices. They should have renedered all such clauses unenforcable and had MS pay fees to the competitors they locked out or something.

Second, they exceptionally closed system. Other vendors are continually being locked out because MS can continually change their APIs, file formats and so on to keep vendors playing catch up and make sure they can't offer directly competitive functionality.

Third, the lisencing practices they use with their customers makes it very hard to move away from MS and can be a big money pit in may ways.

Now, this isn't to say, it's all MS's fault, this is merely how they hurt competition, this isn't the sole way the hurt competition. If we use competition drives quality, innovation and so on well then they've suffered in part thanks to (rather no thanks to) MS.

 

[DemoCoder]

I then asked for substantion. Your response is for me to go find it myself. Well, I could unpack my Amiga, set it up, and perhaps I could find the info, but I tend to expect I'll find the issue does not fit the picture you paint of it as disqualifying the Amiga as a pre-emptive multi-tasking operating system. It does not seem a good time investment.

Well, here's your proof, quoted from the RKM Third Addition, Appendix E

Kickstarts before 2.0 contained two task switching bugs. After an interrupt, a task could lose the CPU to another
equal priority task, even if the first task’s time was not up. The second bug allowed a task whose time was up to
hold on to the CPU either forever, or until a higher priority task was scheduled. Two busy-waiting tasks at high
priority would never share the CPU. Because the input.device runs at priority 20, usually the effect of these bugs
was masked out for low priority tasks. The ExecBase->Quantum field had little effect because of the bugs.

Yes, preemptive multitasking worked in some scenarios, but it didn't work in others. The fact is, no one ever noticed the bug because the vast majority of context switches happen because of cooperative yields (Wait()) and not preeemption. But it's remarkable that such a fundamental feature of the Amiga's operating system that so endeared it to you had a major bug for several years and you knew nothing about it.

Linux's scheduler had a poor implementation bug in it that wasn't exposed for years until people starting running large servers on it. Race conditions are notoriously difficult to find.

Did it disqualify it as pre-emptively multi-tasking as well?

The bug didn't prevent pre-emption, it was just merely an inefficiency as high numbers of tasks on the ready-to-run queue would be suboptimal. A serious bug would be if you could run a task under Linux which was supposed to share the CPU but could accidently steal 100% of the CPU forever. This is just what could occur under AmigaDOS in situations where it should

Well, your specific statement was something like "Windows NT = 1994, Amiga = never". Which you've just clearly established was wrong. Glad we could clear that up. And, actually, this was integrated into later versions of the OS. There was even a CPU abstraction layer integrated into the OS later. Does not quite fit "never" does it? But let's pretend that's not what you said.

Burden of proof: On you. I have the latest Autodocs and Rom Kernel Manuals up to 3.1 and I can't fine any mention of task isolation or device independent graphics from any of the OS versions officially released from Commodore. Commodore NEVER released a version of AmigaDOS that supported full memory protection isolation for apps, resource tracking, and retargetable device independent graphics.

But you've been wrong before, and you've proven that you don't know much about the technical details you are talking about.

Long after Commodore went out of business, FIVE YEARS LATER in 2000, OS 3.5 was released by a new non-Commodore company, it sold a total of 13,000 copies. Perhaps this is what you mean, ok, so let me rephrase that:

WinNT = 1994, Amiga = 2000 (after Amiga is dead and 99% of Amiga developers moved on) Ok, so we'll play semantics and let you wiggle your way out of the bind your created for yourself by allowing for people who "left the Amiga scene" in the year 2000. har har!.

Is this what you said? No? So when I addressed what you said, was I disagreeing with this?

Come on dude, stop playing semantics. We were talking about operating systems. For all you know, there might have been a 3rd party extension for Windows 1.0 that added preemptive multitasking. Some consolation.

Basically, whenever you are proven wrong, or your points are shown to be weak, you cry that either I've ignored what you said, or twisted it, or you change the domain of the discussion.

Either you are comparing operating systems, or not. If you want to bring 3rd party hacks into it, then we can delve into all the 3rd party TSRs for DOS and Windows that added extra's as well.

This is true of many versions of the Amiga OS. If you'd restricted your comment to those versions and stating built in, I wouldn't have argued that you were wrong. What you said was "supported", and 3rd party expansion was the most important aspect of Amiga OS (that benefit from flexible design you conveniently insist does not exist).

It wasn't a flexible, modular design. It was people PATCHING LIBRARY VECTORS. It wasn't like Exec included a routine called AddCustomMemoryManager() or SetCustomTaskScheduler(). There was no forethough or modularity to it.

Moreover, there was lots of global datastructures that every amiga process modified that would be broken by memory protection and many of those were inline MACROS that couldn't be trapped by patches.

That's what the replacement libraries for these functions was for. That is why I said you had to program specifically for this...rather clearly before. Oh, but I couldn't have, I think Amiga OS is perfect.

There were no standard libraries for AmigaOS as shipped by commodore. And furthermore, as I stated, due to the way the entire OS worked (*.devices, handlers, message passing) it was literally impossible to have full protection without running entirely separate Amiga operating systems on virtualized hardware.

An errant app could trash exec's task structures, and it would still bring down the whole OS, even if a "new memory protection aware app" were created. It was impossible to protect all the OS structures from being trashed, since by design, they were supposed to be written to directly in C/Asm thru pointers.

I'm familiar with the many shortcomings of Amiga OS, and what amazed me was how its design principles allowed them to be overcome with the same fundamental architecture.

What design principles would those be and what did it overcome using those principals?


[quiote]I'll mention that there is an Amiga OS 3.9 (and maybe 4.0, not sure) beyond the 3.0 you mention, which is another situation that you seem to practice the "I didn't see it, so it never existed" philosophy with.
[/quote]

OS 3.5/9 came out 5 years after Commodore's bankruptcy. Not a single Amiga machine was being produced for that entire span. The new OS 3.5/3.9, from available information I can find, STILL DID NOT IMPLEMENT MEMORY PROTECTION. Anyway, Amiga was long dead. If you bought one post 1995, I feel sorry for you.

Apple succeeded this much, so Microsoft's monopoly has not had an impact on them adn their ability to compete. I see, your logical model is consistent atleast.

You asked for an example, I gave you one. Apple sales are up significantly post iMac era.

I do note that you glibly ask me for illustration about why the advantages of BeOS affect the consumer, and when I answered you ignored my answers, and discounted them.

I asked how could BeOS make me more productive, do my work faster, etc. You never answered it.

You didn't see it, it doesn't exist. That is a very convenient world view. Did I say Amiga OS, or "operating systems from Commodore", by the way? Not trying to slip a fast one by me, are you?

Demalion, I've already proven you wrong N times already. On Arexx, you didn't even know about the existence of WSH on Windows. On Exec.library scheduling bugs. On the existence of memory protection, device independent graphics, virtual memory, et al in the Commodore Amiga OS. I've demonstrated my knowledge of the Amiga core several times already. I mean, did you ever write a single line of Amiga code in your life?


The only person who is trying to weasel around in this discussion is you, by trying to use the existence of third party, non-Commodore hacks to argue for the existence of OS features that simply don't exist. In fact, even your weaseling doesn't help, because not even AmigaOS 3.9 supports real memory protection.

You don't want to face up to flaws in the BeOS business strategy, and you don't even want to face up to the very bad design problems in AmigaDOS's architecture. I mean, get real, OS3.9, 4.0, MorphOS, or whatever other lamebrain resurrect-the-Amiga-in-year-2000+ schemes certainly don't factor into this discussion. All you're doing is trying to score semantic points. Oh, I said "Never" instead of "2000". Amiga was dead. Vast majority of people know what I meant.


I was an Amiga user from day one. I was an Amiga commercial software developer. I had to face bad Amiga architectural decisions every day. People in the industry had to bend over backwards to fit 3rd party graphic cards and sound cards into the Amiga. The lack device independent graphics significantly raised the barrier for anyone to make a cheap, easy to use 3rd party card. You couldn't just create a card and expect DigiPaint to work with it.


Amiga applications routinely bypassed the OS, hit the hardware directly, directly manipulated OS structures via Commodore supplied C macros that couldn't even be patched by later OS revisions. There were oodles fundamentally bad design decisions in the AmigaOS that held back progress on Amiga hardware, and made it significantly harder to upgrade the OS later.


You wanna single amazing praises about the AmigaOS? Sing them somewhere else. Frankly, the only thing amazing about the Comodore Amiga of its day was its graphics and sound compared to CGA/EGA graphics of the day. The OS was already primitive by the time they shipped it compared to what was written 10 years before (full preemptive, protected, secure, resource tracked, oses)

Frankly, the Amiga was more of a GAME MACHINE than a productivy box.

 

[Humus]

On the topic of gcc code performance, a friend sent me this link which is quite interesting:

http://www.coyotegulch.com/reviews/intel_comp/intel_gcc_bench2.html

 

[Rodéric]

BeOS was made for the consumers, and was considered THE media OS.

BeOS is very simple to use, very fast and reponsive, it boots quickly don't require reboot and has no big bugs (like the 32 bit timer in windows making it fail after sometimes)

BeOS from a consumer stand point was really attractive, only the number of app was a problem, as well as drivers, Be inc couldn't make drivers for every piece of hardware out there, which is understandable.

BeOS was planned to be integrated in big brand as secondary installed OS, however Microsoft licence agreement forbid that.
(You cannot use anything but MS bootloader to boot a MS OS, so how do you boot into BeOS ?)

It's clearly MS outlaw behavior that put an end to BeOS.

On the Quality side, Windows (2000/XP) have not even reached BeOS.
Neither on File System, multi threading, reponse time, media capabilities (how fast to handle them), and boot time + reboot.

I still have BeOS on its own hard drive, and I boot it once a week or so, I feel just safe using it, I won't have crash and bugs like windows often did (even if know 2k/XP only seems to crash in the video drivers).

BeOS just does what you ask it to, unlike windows