“Many low-end Mac users lament Apple’s choice of NeXTstep over the BeOS as the foundation of Mac OS X. Many arguments have been made on technical merits, and many blame the demise of Be, Inc., on Apple and Microsoft. The following is an in-depth look at the demise of Be and reflections on Apple’s choice of NeXT. Hopefully this will debunk some common ideas about BeOS and lay this debate to rest.” Read the editorial at LowEndMac.
LowEndMac on Why BeOS Failed
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
A while back, I read somewhere that the BEOS is much harder to write code for than other OS’s. If this is true then it seems like it may bave been another reason why Apple decided to go with the Next OS. Apple may have been thinking that some software developers might balk at writing code for the new Apple OS (BEOS) since it takes much longer to code for and the roi is not much.
I’m not putting down the BEOS, I wish I could have tried it out. In fact, if anyone knows where I can buy a copy to run on an 8600/200, please post in this forum.
this my problem with the article: isee no reason to choose mc os x over win xp, and win xp has more support less expensive hardware. anyw i can sell u my copy its real, beos 4.5 for $20 @ gnosis92@yahoo.com
I don’t know….Everytime I take alook at MacOSX I thank all the gods Apple and Jobs didn’t get Be. When you look at how they pimped up NeXT, the mind shudders at the thought of what they would have done to Be.
A side note. Over the past year, I’ve noticed the most vehement Be-bashers are the ones who were previously the strongest supporters. Anyone else notice this? The same thing happened with OS/2 advocates…..
I read somewhere that the BEOS is much harder to write code for than other OS’s.
everyone i’ve talked to said it’s far easier than other OSes.
if anyone knows where I can buy a copy to run on an 8600/200, please post in this forum.
hmm, gobe might still be selling it.
Gobe is out
http://www.purplus.com you will need to search at that site, I think it is listed for 38$
okay… TyCom Systems has some:
http://www.tycomsystems.com/beos/store.html
but i’m not sure if that’s the PPC version also, you should e-mail them to check…
And about the 8600/200… only some 8600/200 system work in BeOS, IIRC. It depends what motherboard revision you have, I think.
Kevin, check your netlane email please.
This guy adds nothing to the story we all didn’t know already, & his spin is all wrong. If he would just read Scot’s book & the interviews, he would know that Be was much better off without Apple. And damn it, MS did strangle Be to death with the oem agreement.
Doesn’t he know that “geeks” write the apps he wants to use.
This so called x BeOS fan seems obsessed with the 8 Quicktime video stunt he saw at MacWorld. He seems to have little real knowledge of what BeOS is all about, & it ain’t about playing 8 Quicktime movies. In reality, real BeOS users know exactly what it is all about, reminds me of that StarTrek NG episode when Barkley got a severe IQ boost, he took control of the Enterprise from the Holodeck so that his thoughts were those of the computer. Well when I am using BeOS, I feel like Barkley. When I am using a Mac or W2k, I feel more like I am riding a tricycle or perhaps a trabbi. I exaggerate of course. And I say that as a veteran developer of Mac & NT from their 1st versions.
Now I do like my old slow Macs but I won’t buy another pretty box till it is on fast x86 HW. Until then I run Mac 68k apps on x86 with Basilisk2 at 10x the speed of my old ppc. If only I could find a few more downloads to make the Finder half as good as the Tracker or even Explorer.
As for so called poor BeOS web experience, it is my experience that the rather weak & feeble Net+ can keep running a session with far more windows for far longer than the mighty IE, let alone Netscape. Hopefully Opera will get back in if the OBOS team makes it.
I pass the torch…
I don’t think that technology had that much to do with it. It seems like the real prize that they were after was steve jobs and he didn’t come with BeOS. They needed a miracle worker to revive that place. No one could have pulled that off but steve jobs.
they didn’t buy a technology they hired a man..Steve jobs!
Only Steve had the vision, charisma, and marketing skill to save that place.
Amen to that! A good book to read for NeXT information is “The NeXT Book”. It explains NeXT inside out, not only what they did but how and why they did it. Jobs was/is an amazing person, and Apple did good by re-hiring him.
Scot was writing in his capacity as an PR employee of the (then) BeOS software dev house — now that’s called spin. Go to Qubit’s website 2 years ago and they called BeIA the best thing since sliced bread. Go to Qubit’s website 6 months ago and they called WinCE the best thing since sliced bread also.
It’s not just isolated individuals who were obsessed with running 8 Quicktime windows at once. (As an aside, your claim of BeOS running “far more windows for far longer than the mighty IE” is also similar in nature.) Those claims mean nothing in the real world because no body runs their computers with 8 QT movies and 30 internet browsers at the same time. The multi-media claims are highly questionable now because those demonstrations were done with media files with much lower bitrates and much lower CPU intensive codecs.
Many people touted BeIA to be the best OS for the IA devices because of this media-OS savvy ability. But of course, the problem is (1) you can’t download multiple streams with a 56K modem (2) even if you have broadband-based IA — the CPU load is too much for even running 2 movies at the same time on a 233 MHz CPU [the CPU load for decoding mp3 on a 233 BeIA device is about 40%, so most likely you can’t even play 2 mp3’s at the same time] and (3) even if you have a much faster CPU, you don’t have the RAM and the hard drive space to do it.
“It’s not just isolated individuals who were obsessed with running 8 Quicktime windows at once. (As an aside, your claim of BeOS running “far more windows for far longer than the mighty IE” is also similar in nature.) Those claims mean nothing in the real world because no body runs their computers with 8 QT movies and 30 internet browsers at the same time. The multi-media claims are highly questionable now because those demonstrations were done with media files with much lower bitrates and much lower CPU intensive codecs.”
You’re missing the point. BeOS could multitask better than any other OS I’ve ever seen. Sure, people don’t run 8 quicktime movies at once, but they might want to rip a cd, encode mp3’s, surf the internet, and maybe watch a movie they downloaded all at the same time. And BeOS did that really, really well, without crashing and without stuttering.
Morever, BeOS was truly elegant, as the article tries to deny. It constantly amazed me by how easily and quickly I could accomplish whatever I needed to. I would rank its gui right up there with MacOS in terms of ease of use. It had the best and easiest installer I have ever seen, period.
The article makes some really idiotic points. For example, it cites a problem with Be was that it didn’t have a Carbon layer. But neither did Next! So how is this a con of BeOS in 1996 when choosing which to buy? And I hate it when people say things like, “Be couldn’t even print in 1996.” Yeah, well, Next didn’t run on PPC, wasn’t optimized for multimedia (which is what most Mac people want to do), and Quartz (the evolution of Display Postscript) hadn’t been written yet. Please, if you’re going to compare Next to Be, at least be honest. Next had more advanced developer tools and better networking, but BeOS owned it with regards to everything else. They wanted Steve, plain and simple, and I don’t blame them, as he’s clearly turned the company around.
I think that the LowEndMac article hit the nail on the head (so to speak) – Apple is a hardware company, they need their customers to buy new machines as frequently as possible. Running a slim/efficient OS would not generate as much hardware sales as running a bloated/inefficient OS.
NeXT definately was a better buy (from a hardware company’s perspective). For us consumers, BeOS was the better buy.
’nuff said.
Maybe if you deaktivate all the stuff which Nepositive is not supporting the IE will be also much more stable
And for the programming of BeOS, I remember some complains about writing really big applications. Which will be much more difficult, because of the ‘automatic threading’ for example when you open a new window. Which will give you more overhead for synchronicing (the story was about Pe and others).
So maybe it was easyer to write some small apps, but probably it was dificult to write bigger one’s.
Thöms
Coming from someone who has never used a Mac (why would I it’s only got one button ?) I’ve used Windows (prior to that an Amiga 500) most of my life, until BeOS 5. This was the change for me. I was actually excited about an OS. Sure, I used NT/2000/XP at work and it does the job, but it didn’t make me sing about it. BeOS did. I couldn’t stop talking about it for months (and still can’t). Hell, Intel even said they’ve never seen their hardware perform so well until they saw it running the BeOS. These days, the industry is too money driven. Money is the bottomline and that’s sad. And that was BeOS’s problem. They mismanaged their money, gambled on the IA market and lost. As for all of BeOS’s geeky features, well, they’ll all crop up in the new versions of Windows eventually.
I think will all the focus shifts that Be Inc. kept throwing here and there doesn’t give developers much confidence in supporting a system that might be doing something else tomorrow. Though I have a lot of love for BeOS, I can say the story was right on (mostly without all the sarcasm) and it’s a sad but true tale of what can and did go wrong!
I hope that the OpenBeOS and BeUnited Group takes a read from the article and tries to make a difference for OBOS’s sake
It amazes me, the gall of some people. Unmitigated.
With statements like “I’ve never used [insert product here]
but here’s a laundry list of why it sucks…” and “I was a developer for [insert products here] in the olden days, so I know EXACTLY what’s going on today…”
Use the software. Explore new systems. People, I work for an operating systems company and I’m open-minded. It’s not the best thing in the world. Neither was BeOS, neither is atheos. They’re tools – used for programming, surfing, gaming and other utilitarian ends.
Tools.
Just venting, but whenever I see people pimpin’ their favorite OS to the detriment of others, I just wonder if hardware geeks slam different brands of hammers.
Tools, people.
Just tools. Learn new ones, then slam ’em. *OR* appreciate the infinite beauty and utility that diversity brings, and walk away knowing a bit more at no cost.
Nah, that’d just be civilized. ‘Course, discussion is good – it’s the ignorant (and proud of it!) attitudes that makes me wonder when geeks became proud to be stubborn.
I’m too young to be so disillusioned
For me it, the real situation is that although some say BeOS is the best OS in the world, it still FAIL. IT IS DEAD. Maybe we are hoping on OBOS, BlueOS or YellowTab etc. but can it be the same as the original BeOS? Who can guarantee?
Lets hoping Palm Inc. change their mind and release it to OSS community but it doesn’t have to be GPL if Palm Inc. still want to make money from that. At least both parties are benefited.
The article was a crock IMHO. A few quotes and responses:
> 1. […] being Unix-like is no advantage.
There’s actually nothing particularly Unix-like about BeOS. This was a common misperception apparently caused by people noticing its POSIX layer.
> 2. […] BeOS was designed by geeks for geeks …. You don’t attract developers and partners with superior technology unless you show them how they will profit from it.
Could developers and partners be courted immediately? Obviously the geeks appreciated the technology first. Application developers come later. And Be did have selling points ready for them. Unfortunately about the same time BeOS was ready for app vendors, Be shifted focus.
> 4. [….] No matter how good a GUI looks or how responsive it is, it will not attract and retain users unless it is an enabler.
The BeOS gui *is* an enabler. In fact it isn’t especially good looking. Those yellow window tabs you either love or hate, and in general a lot of people thought the gui is too spartan. It is in fact its capabilities that shine, not its looks (and I’d say responsiveness is a factor in keeping users happy since it is a component of enabling).
> Business and Marketing
> A machine with a configurable and user-programmable expansion port? …. Using these attributes of the operating system to sell it (to geeks) is perhaps one of the all-time worst business decisions in history.
I don’t recall Be ever citing the GeekPort, kernel, file system and so on as selling points in and of themselves. If you followed the early white papers and so on, it was all about capabilities like low latencies and multitasking, in other words the result of those technical details, not the details themselves. No one ever said, “here’s a new computer with a programmable port for you,” expecting this to bring mass sales. Of course it pushed the geek factor off the scale (along with the front bezel LEDs which the author missed, incredibly, as another obvious reason for Be’s failure), but Be was trying to sell an exciting, capable computer for media professionals, mainly. If some guys bought the BeBox for the GeekPort, that’d be icing on the cake. I’ve always been curious what the rationale for that device was, because it wasn’t particularly pushed by the company; in any case it for sure wasn’t considered a significant selling point the way this author is implying.
> 2. […] BeOS was designed to be a personal computer operating system. Unfortunately, at the time of its debut and the years thereafter, the real growth was in server operating systems.
What, *only* servers sold at that time? Obviously there need to be boxes on the receiving end, and workstations in the middle. Growing sales of servers only means that there is also a growing IT market, including desktop machines. Not participating in one sales trend doesn’t mean everybody doing something else concurrently is a loser by definition.
> 3. [….] 3. What’s that Internet thingy?
> BeOS also missed the boat on the Internet phenomenon. … Internet connectivity was added seemingly as an afterthought.
Basically the author is wrong here. Internet connectivity in BeOS, from the user’s standpoint, is integral and very nicely done. If you equate “internet connectivity” to “enduser’s browser” then, true, the BeOS experience is barely adequate, but the Be line on this was “third-party opportunity” for better or worse. Actually Be rationalized the rather poor enduser experience with the idea that BeOS was for “media creators” more than media consumers.
The author is correct in saying “BeOS wasn’t killed by the big boys; it was slowly choked through a series of monumentally bad decisions.” Unfortunately he’s confused about which decisions were were the killers.
— gary_c
the author failed to realize the potential of querying by attributes. i think it’s amazing what you can do. No other OS can do that.
Well said, TLFord.
I agree that no single OS can be the solution to everyone’s problems. Every OS has things it excels at and things that it is not so good at. One should be open-minded about trying out new things and supporting them on their merits rather than just being knee-jerk reactive.
Also, it’s not just the OS, but also things like available skillsets and development tools and development communities that can be big determining factors as to whether you would, say, choose Embeddix or Windows CE or eTRON for an embedded systems project, for example.
A monopoly of any one particular OS will not make everyone’s life easier, IMO, because each OS is fairly optimized for certain applications and/or markets and people should understand this.
I used to be pretty religious about my favorite OS but after having checked out so many facts about other OSes on my own and trying some of them out and expanding my little ‘world’ of knowledge, now I can see why competition and innovation is good for everyone. I find that it is the ‘neutral’ standards bodies who make everything inter-operable between different OSes (XML, WBEM, Kerberos, Wi-Fi, etc.) that really should be defended tooth-and-nail, not so much the OSes themselves.
Long live multi-boot environments!
Hi,
some German BeOS fans have build up a BeOS distribution based on the freely available BeOS Personal Edition. The BeOS Developer Edition comes with many new drivers, tools, apps and recompiled software.
http://bezip.de/app/1194/
Ciao,
Sebastian
This author does not make any new points. The things BeOS got right are areas that need much improvement on Mac OS X, and lesser so, on Windows XP. Mac OS X is a memory HOG by any stretch, requiring at least 512mb of RAM to get real work done instead of listening to the hard disk page all the time. I still remember how to write a Be application in my head (and haven’t written BeOS C++ code in over a year), yet Carbon? Please. Be did right by designing an OS from scratch. The UNIX argument doesn’t fly, either. UNIX is old stuff, and too complex. Be had all the tools people needed. What a bummer that people had to work a little harder to port thier code. This is not what killed BeOS.
When we’re on the topic of Mac OS X, by the way, name 10 new applications for it… There aren’t 10. All the apps are Carbon ports — not a leap of engineering by any means. Shameful state of reality this is, indeed.
Usability-wise, Be’s Tracker is a much better file browser than Mac OS 9’s “classic” Finder or the horrible kludge that is the shoddy OS X Finder today. What a waste of time it is to even use it.
BeOS was a far superior OS, technically. Of course, the market does not appreciate this as the author states, but most of us do.
Steve
“> 1. […] being Unix-like is no advantage.
There’s actually nothing particularly Unix-like about BeOS. This was a common misperception apparently caused by people noticing its POSIX layer. ”
Um, *cough*, Bash.
and then reading the comments, I can see that the author is right on.
he says technical merits are not selling points to the masses.
a guy here says, they did not use technical merits as selling points. then he turns to explain how its gui was so much faster and its memory foot print was so small and it was such an efficient os.
heh…makes me laught a bit.
> Um, *cough*, Bash.
I have a UNIX-like command shell on my old Atari ST. Does that make TOS a UNIX-like OS? No, certainly not!
Come on, in BeOS the Bash is just an application you can start of you want to do something on the CLI, but in UNIX everything is started from the shell.
I heard about all the power of BeOS and tried the personel edition a couple of times. Each time I looked at it, I walked away feeling, neat, but not useful. For me it was about a lack of 3rd party apps, not whether I could get it preinstalled on my Dell machine. It didn’t run any of the things I used, none of my games, none of my 3D software, and I didn’t need to boot into another OS to play MP3s and listen to CDs. I don’t choose my OS because of any love of the OS, I choose it because it runs the applications I want to run and if a new OS has just the same old check list of apps (email, browser, mp3 player) without offering a new killer app I can’t get elsewhere, there is no reason to change. I only watched BeOS from a distance, but for me that was the perceived downfall.
>>…. While there was an enthusiastic BeOS community, Linux surely consumed most of the enthusiast developers. Be did not get solid industry partners until it was far too late.
The lesson: You don’t attract developers and partners with superior technology unless you show them how they will profit from it.<<
They didn’t design it “for geeks” to impress ISV’s you mororn. They did this to impress geeks (I know that doesn’t sell software). If you can gain a FAITHFUL_LARGE early adopter audience, it would be a great boon. Be did. It was. Again, this had ABSOLUTELY NOTHING to do with Be’s demise.
>>3. The BeBox
The lesson: You don’t make money off anything unless there is a market for it or you create a market for it. This is even truer of proprietary hardware.<<
No $h!t Dick Tracy. Out of all the drivel your puking out this has to be the dumbest bit. Be was attempting to Create a market. Your “segment” on the BeBox has absolutely no point whatsoever.
>>4. The GUI will save us
…
The lesson: A slick GUI won’t save the operating system. (Even Apple had to re-learn this one after releasing Mac OS X — iTunes and the other bundled apps are all enabling tools to attract users.)<<
Who the fsck said “Oh, gee guys our sleek GUI will make everything better!”. Uh, NOBODY.
Of fscking course they added a sleek GUI to their OS. And somehow you see that as a reason for failure???? I think you just like to talk.
I could go on but you get the point. This guy rambled on for pages and said nothing. Stick to your Mac dude, where only seeing the surface is a good thing.
>> “> 1. […] being Unix-like is no advantage.
>> There’s actually nothing particularly Unix-like about BeOS. This was a common misperception apparently caused by people noticing its POSIX layer. ”
> Um, *cough*, Bash.
OK, and Bash :^)
— gary_c
The author misses the point of the Quick Time stunt. I media creation such as Audio and Video editing, it is required to play and mix many audio or video files while adding effects or making edits. BeOS would have been perfect for this kind of thing. It would compete with SGI and Mac very well in this market. But you can’t stay in business by giving your product away. It is a shame that an OS with such great potential to do high end commercial content creation for audio didn’t have the needed apps or driver support. Professional studios don’t use a soundblaster audio card. Imagine BeOS apps from Steinberg or Digidesign running with ProTools, MOTU, or Ensonig audio hardware and Mac and SGI have a HUGE problem in the art-geek market. It is infuriating to think about such wasted potential. The author would do quite well with Windows or OSX for his normal consumer needs. BeOS was a new beast. I can’t believe how much they screwed up the simpler concepts while doing amazing work in the “impossible” areas. I was convinced that Intel hardware was crap until I loaded Be. I was spoiled. Anyway, I really don’t care if I can’t run Office or do my taxes on BeOS. I would use Windows for that kind of thing. BeOS should have been heaven to Art-Geeks.
My favourite bit is his comparing AmigaOS to Unix then saying AmigaOS was his second favourite OS… How can it be his second favourite OS when he clearly hasn’t a clue about it? AmigaOS had *nothing* in common with Unix other than what all OS’s have in common. There was a public domain 3rd party posix compatability layer available and hence many ported posix tools, but that was it. And as for that being the reason for AmigaOS failure, no, that was Commodore being run by a bunch of money grubbing clueless jerks… No real investment = dead OS.
When Apple decided to axe Copland (was to become MacOS 8) in 1996, they decided to look for other options to make a more modern OS architecture. Options were continue to develop 7.X and merge work from Copland in it (this actually happened too, Coplands user interface for instance is what we now know as platinum, introduced with MacOS 7.6, and premeptive multitasking and process management introduced during system 8.1 to 9.2),
or to shop for a modern OS foundation to build a MacOS on.
Apple decided that BeOS could be an interesting option, it had an interesting advanced foundation and it already worked on the same hardware MacOS runs on.
so serious talks between Apple Inc. and Be Inc. about a takeover were conducted.
But Jean Louis Gasee (former apple employee and CEO from BeInc.) overplayed his hand. He thought he had all trumps and refused to sell BeOS to Apple, instead offering to LICENCE BeOS to Apple. This was ofcourse unacceptable to Apple.
And Jean Louis Gasee totally overlooked another small OS company, Next. From that other former Apple employee.
Next was not as technologically advanced as BeOS, but had some very interesting features, that the geeks this editorial talks about of course overlook: very advanced development tools and Display Postscript… two important ingredients to not only make an OS but also software for Apples main area… graphics and publishing software.
And other than Mr. Gassee, Mr. Jobs had no problem selling his company to Apple.
These are the two real reasons why Apple did not buy BeOS:
BeInc. wanted to licence to Apple which was not acceptable to Apple because they would not have control over their new to develop OS (and also it would not add software engineers to Apples team that had experience with BeOS… the head of the Next software development team is now head of Apples software team), and the other alternative a big advantage that geeks often overlook, and that is a foundation to write insanely great software on (Next’s development tools Project builder and Interface Builder with Objective C, and Display Postscript).
Out of grudge Mr. Gassee did not develop BeOS for Apple hardware anymore, alienating his user base… and began to develop it for the Intel platform. The Intel based BeOS does not come close in performance to BeOS running on PPC hardware, and BeOS fell into the trap all OS-makers including Linux fall, and that is driver hell.
On an Apple mac pheripherals need to behave a certain way towards the hardware and OS, on (W)Intel it is left upto a driver how the OS and the hardware work together, making it very difficult to let an OS run on every (W)Intel computer.
In the years that Apple spent to make MacOS X and add lots of cool features and software, BeInc spent its time writing basic drivers to make BeOS run on Intel hardware, keeping BeOS a nerd OS (fast and cool, but nothing to actually do with it).
Mr. Gassee killed BeOS himself, by not selling to Apple and then switching to Intel hardware.
Apple didn’t buy Next for Steve Jobs. That’s the kind of reverse-engineering of history Apple partisans love to do. They were picked over Be because Next had existing credibility in the enterprise, and everything Gil Amelio did at Apple was canted toward getting the company into the enterprise. Go out and read what the man said (and look at what Apple did). You think Amelio wanted to buy his replacement? In a post-Apple interview he was asked how things would have been different if they’d bought Be, and his wry reply was, “Well, for one thing, I’d still have my job.”
Granted, BeOS partisans also constantly reverse-engineered history. Be themselves claimed after the fact that R4 was the first “ready for consumer release” and people have insisted that R4.5 or even 5 (after they’d dropped the “R” entirely) were the first “consumer” releases. But, well, that’s what R3 was being called when it came out. That’s even why it was just “Release,” not “Developer Release” or “Preview Release.” And there’s the laundry list of promised improvements (and even bug-fixes) that took years to land or never landed at all, but we won’t go there.
As I noted on Low-End Mac’s forum, Be did have a fair amount of developer support; if you measure from the release of R3 as a “1.0” through 1999, they developed commercial interest at a much faster rate than Linux did after its 1.0 release. And, given the resources and time that Apple put into turning OpenStep into Mac OS X, BeOS would have certainly taken no more work to be at that level by mid-2001.
BeOS was certainly easy to develop for in most respects–although it’s worth noting that Next’s development environment has long been considered the gold standard. Be never had anything remotely comparable, even commercially, to the Next-derived tools Apple makes available for free.
well.. i used to play 8 hot quicktime movie a the same time
Ok, so Be has bash. Yes I suppose it is a unix-like command line. Tell me, how hard would it be to add bash to say, windows or MaxOS 9. Not very hard, me thinks (though I’ll admit I’m no programmer). Would that make Windows or MacOS 9 Unix? Me thinks not.
I did enjoy the article even though I don’t know the validity of the author’s statements. Overall, though, a lot of it made sense. I do feel Microsoft at least helped kill Be with its OEM practices, but I wonder if BeOS would have gotten the software they needed to make people want to order their PC from Dell with BeOS rather than Windows.
For me, the thing the article got right is that most people really don’t care that much about their OS (the readers of this site obviously fall into the exception). I imagine the average joe wants the OS to work, get on the internet, maybe do some word processing, stuff like that. Essentially, windows already did this fine, so there really isn’t a need for any of BeOS’s advantages there (sure people could benefit, but it wasn’t essential). Add to that some people stating that BeOS was more aimed at media creators rather than media users (home users, end users) explaining the weak internet experience, and the conclusion I draw is that BeOS didn’t really intend to target the home user en masse. If that’s true, the OEM stuff shouldn’t have affected them had they stuck to their business plan.
But they didn’t, at least that’s how it seems (and the BeIA endevour lends credit to my belief). Maybe that R&D $$ should have went to buying Avid or ProTools or the company that Apple bought Final Cut Pro from. Final Cut Pro and a few other media creator products keep Apple afloat in the professional community. BeOS needed to go that route and didn’t. It almost seemed like they developed this great core technology then missed, ignored, or got shut out of every opportunity.
Bottom line is I really feel like BeOS was doomed to fail from the start. It’s only chance, even from the beginning, was being bought out by Apple. It could never have competed with Windows (fairly or unfairly). Its real competition (for what BeOS was intended to do) was Apple, and Apple desperately needed a new OS. I’m sure programming for NeXT isn’t total hell on earth, Apple got Steve Jobs back, and what did BeOS offer that was so important to have?
This is a rambling post, it probably doesn’t make much sense, but for me, the bottom line is BeOS needed to create a demand and failed. The build it and they will come approach to business will kill a company everytime. I bought the BeOS Professional Edition/Gobe combo just for fun and to throw my fiscal vote to BeOS. It’s a great core with no reason to use it. BeOS, just like Apple and Windows, needed to create a reason to use it. *nix OS’s are having the same problem right now, great core, no real reason to use it.
after reading some of the other posts here, I believe BeOS’s biggest hurdle might have been the owner himself.
Ok, so Be has bash. Yes I suppose it is a unix-like command line. Tell me, how hard would it be to add bash to say, windows or MaxOS 9.
*cough* Cygwin *cough*
Hmm…I guess that makes Windows Unix-like, eh?
Arthur, have you used Be?
Like there wasn’t enough misinformation in that Low End Mac article…
Out of grudge Mr. Gassee did not develop BeOS for Apple hardware anymore, alienating his user base… and began to develop it for the Intel platform.
Wrong.
Where they had previously given active assistance on supporting their hardware, post-Jobs Apple declined to share new hardware specs outside the company (in addition to squashing its fledgling “clone” licensing program flat). Be was faced with the sole option of reverse engineering support for new machines and facing possible (some say probable) lawsuits.
In the meantime, Intel stepped up with funding and active engineering assistance. Plus, x86 support offered an order or two of magnitude increase in the possible userbase.
What would you do?
To suggest that JLG focused on Intel out of spite (or, out of grudge) is crazy; you obviously haven’t read any of his comments on the matter.
The Intel based BeOS does not come close in performance to BeOS running on PPC hardware, and BeOS fell into the trap all OS-makers including Linux fall, and that is driver hell.
I don’t know where to start here. Driver support on the Mac is no problem? Tell that to the tons of people without drivers for scanners, printers, etc.
And PPC performance was better than Intel?
cough
BeOS engineers were on record saying how, after getting the kernel up on x86, they were shocked at the performance improvement.
On an Apple mac pheripherals need to behave a certain way towards the hardware and OS, on (W)Intel it is left upto a driver how the OS and the hardware work together, making it very difficult to let an OS run on every (W)Intel computer.
Also wrong. To oversimplify the situation, it was necessary to support more driver models in Windows than in Mac OS, but to imply that Windows has no driver model (and just as bad, that Apple had only one) is ludicrous.
Ayesquiddy, HTML is not supported anymore on the OSNews forums. Please read the fine print underneath the posting form on how to use bold and italics from now on (I have only converted to UBB the first two tags you used).
As for the HTTP URLs, they will auto-parse into HTML, no need to use any kind of tags, just leave a space between the first and last character of the URL (or just write a plain url in its own line).
http://www.unix-systems.org/what_is_unix.html
I guess the phrase “like UNIX” can mean anything you want, but for me the most useful way to think about it is whether UNIX software works there, and the best index of that is compliance with the X/Open Single UNIX specifications. POSIX 1003.1 is a start, but as noted by others already it’s a pretty easy mark to hit. The interactive shell, etc., is too superficial – important, but easy enough to add later on a platform that’s enough like UNIX.
In my estimation, BeOS does indeed qualify as UNIX-like on that basis – they went way beyond POSIX 1003.1, and you can treat it as a UNIX flavor and port lots of software. I’m not saying BeOS is a kind of UNIX – it isn’t, just similar in a convenient way. I think that was a good idea, there wasn’t any realistic alternative, and I think the little flaws in their implementation did hurt them a little, but it’s hardly a major cause of their demise. Yet to be seen whether Apple’s approach is really going to pay off – can they bring their BSD up to speed with the alternatives, or will it be just a better 2nd rate UNIX than BeOS is? At least BeOS doesn’t have to drag around a complete UNIX kernel.
I did enjoy the article even though I don’t know the validity of the author’s statements. Overall, though, a lot of it made sense. I do feel Microsoft at least helped kill Be with its OEM practices, but I wonder if BeOS would have gotten the software they needed to make people want to order their PC from Dell with BeOS rather than Windows.
The people Be was trying to reach with OEM distribution don’t care what OS comes on their computer. They just want to browse the internet, send email, balance their check book, etc.
This is just my speculation, but the user base that OEM distribution could have given them might have been enough to get larger software and hardware developers to release BeOS versions of their work. Which would have given BeOS users Quickbooks, Photoshop, drivers, etc.
While I did not agree with a lot of the article, I do have to agree with this point. The shift was occuring before Be made it’s debut. The desktop was already won. By not positioning a “Server” version of the BeOS – Be, Inc. IMHO did itself a great deal of harm. Not to mention the *still missing* multiuser functionality.
And if any OBOS developers are listening – the only reason (besides the fact that I really like it) I have BeOS 5 on my work network is WON. I can nearly seamlessly access the rest of the network – which runs Windows 2000. My suggestion is to expand on that functionality. Hell, I think that the OBOS folks should strive for nearly total network compatibility with Windows 2000. It would be an easy task to convince a network admin to replace Windows with OBOS – if it respected the security* of a windows network (NTFS permissions, Active directory, etc.) Obviously, you can’t beat ’em – so join ’em.
*and before you flame me – a windows network IS secure in the hands of someone who knows what they are doing.
He wrote, “Microsoft pressured OEMs to not bundle BeOS on their systems. Big Deal. Microsoft has done this with every competing operating system. Is it fair? No. Did Microsoft willfully and purposefully destroy Be? Probably not.”
Yes, Microsoft has done this with every OS – AND THE RESULT IS THAT MICROSOFT IS THE ONLY OS. DUH! OF COURSE MS WILLFULLY AND PURPOSEFULLY DESTROYED BE – THAT’S WHAT KEEPING OTHER OFF OF OEMs MEANS! DUH!
Sorry for shouting, but it was such a stupid statement that it set me off. What does he mean, “Big deal”? It’s like “I killed your mother. Big deal. Is it fair? No. Did it mean I purposefully and willfully killed your mother? Probably not.” Stupid.
dave_ asked if I ever used Be. I have, I ordered the professional edition with gobe and tried it for a while. Had some issues with a geforce I was running, so I use, but I eventually got full color with VESA (is that how the geforce does it?). When I used it with a radeon, it went a lot better (also, switching from scsi drives to ATA drives simplied things as well).
Gobe worked pretty well, didn’t blow me away but it definitely wasn’t trash. Surfing the web was also okay, but the browsers just haven’t really kept up in BeOS. I got excited for a while when I heard somewhere that Digidesign’s ProTools was going to use BeOS, but I guess that didn’t pan out.
Someone else thought that OEM sales might have given BeOS the base they needed. Maybe that’s right. I still feel like most people don’t care which OS they use, so they would probably go with what most other people used, Windows. I think Windows was too ingrained in people’s minds in 96. But who really knows?
Anyway, I still have my BeOS and Gobe, maybe someday I’ll stick it on something to play around with it again. Until then, I’ve finally dumped windows and got a mac. So far it has been a great experience, but I think the need to dual boot makes things more complicated than Mac Apple would like.
There was a meeting at cupertino’s office in which apple and his next associates came in totally showed up Amelio. It was said at the time that all in the room realized that Amelio had very little vision regarding the future of apple or computing while jobs did. The Jobs team were described as being full of enthusiasm, vision, etc.
While Amelio may have wanted to keep his job. the board wanted jobs and guess who wins out in the end? Yes the board does. That board actually cared about the company.
The NEXT decision was not executed solely at the executive level (ie: Amelio). It was executed above as well and they wanted to get rid of him. jobs made way more of a difference than BeOS, or any os could.
NEXT with its poor revenues and high expenses would have certainly failed. Period. NEXT would have followed Be, os/2, etc fate.
As a 604e apple powermac owner 7500, with a generous 128 MB of memory and 2GB SCSI hard drive. I’m unhappy mac os x will not run on my machine.
I also own a dell p3-733, and i see no reason to choose
mac os x over win xp.
IF apple bought be, and built mac os next generation on be rather than next, i could run Mac os NG on my old 604e
AND it would offer something win xp can not touch
-superior digital performance, superiorer user experience,
on my 604e with only 128 Mb and 2Gb Hard drive.
also, If Apple built its NG Mac OS on top of Be, with Apple’s muscle behind it and installed user base, we could see Nvidia-Be OS games that blow Win XP out of the water.
Games like quake 5 or unreal 6 or halo 3 on a Geforce 4
as opposed to the cheesy geforce 2mx the new imacs come with.
also, Mac OS NG could be ported to x86, and, the cloners need not have been killed, spurring better powerpc development.
as cute as the new imac is, and it is cute, i don’t see any reason to prefer it to a Athlon XP $1000 clone running win XP.
>>And PPC performance was better than Intel?
cough
BeOS engineers were on record saying how, after getting the kernel up on x86, they were shocked at the performance improvement.<<
Where did you dream this up?! I own a BeBox (Rev-6 PPC driven) and compared to a similarly equipped PC… Don’t even get me started, that is all I got to say!
I am so sick of BeOS flamers dissen the Mac because of the demise of Be Inc.! Yeah I am mad to see an awesome OS like BeOS get screwed over by a mismanaged company like Be Inc, but it is not Apple’s fault, and Microsoft is not totally to blame (though they didn’t make it any easier)!!
Get over it!!!
“Where did you dream this up?! I own a BeBox (Rev-6 PPC driven) and compared to a similarly equipped PC… Don’t even get me started, that is all I got to say!”
I remember specifically reading Be engineers talking about how Intel’s PCI implementation allowed them to do things that simply weren’t possible with PPC in “The BeOS Bible” by Scot Hacker. I also remember reading somewhere that Be benchmarked higher on Intel hardware per MHz than on PPC. Considering that the PPC Macs at the time (the 120 MHz all in one Powermacs) were the absolute worst computers I’ve ever seen, I’d believe it. In fact, seeing Be turn one of those POS’s into a usable environment was actually what first turned me onto it.
Looking back on my time at Be both as an evangelist and as an engineer I have to wonder if one fundamental decision made during that time might (in retrospect – ain’t it a wonderful thing
have been the wrong one.
Be decided NOT to go into the application business, but rather stay focused on the OS itself (3DMix was an exception made at the end of the desktop era, and only made after much deliberation). The reasoning was two fold :
1) We’ll p*ss of our partners (developers) if we’re seen to compete with them.
2) We don’t have the resources
Whilst I certainly can’t argue with item 2 (though more people could have been hired if the desire was there) I have to wonder if item 1 might not have been false.
When I look back on the success Apple has in drawing people to the Mac through iTunes and iMovie I have to wonder if a strategy to develop similar products (esp. considering the Media OS angle) in house for BeOS might not have paid off.
Andrew
c’mon guys. he led a company to ruin, yes, but he did it for longer than most other companies would have done. that comment is ripe for flame but remember he found jobs for his staff after their disolution. yes, gassee can be held accountable on all counts of his companie’s failures, but give him his credit. he fought agianst huge odds and did it for ten years. plus, he’s ferociously intelligent as well. which is nice. when was the last time we had a CEO sooo witty?
Out of grudge Mr. Gassee did not develop BeOS for Apple hardware anymore[i]
What? Your talkin about the G3s and G4s? Last time I checked BeOS didn’t have support for the G3 and G4 macs because apple refused to give them the necessary infomation, and they didn’t want to reverse engineer the systems.
[i]*cough* Cygwin *cough*
Hmm…I guess that makes Windows Unix-like, eh?
I thought cgwin was just an emulated command line…
Had some issues with a geforce I was running, so I use, but I eventually got full color with VESA (is that how the geforce does it?).
drivers have been written for alot of the geforce 2 and 3 cards, check bebits…
Where did you dream this up?! I own a BeBox (Rev-6 PPC driven) and compared to a similarly equipped PC… Don’t even get me started, that is all I got to say!
I didn’t know you could do SMP with the old Pentiums 133s…
Just had a thought. AmigaOS. AmigaOS is well known for its efficency and many strenths. But also known for what it is lacking.
Could have Apple made an OS out of AmigaOS???
I don’t need to use something to know if it’s no good or not, when there are articles and colleagues who have used and abused them and know their pitfalls. I could prattle on all day about the PalmOS problems. I’m open to new OSes.
Over the years, I’ve run CPM, AmigaOS, All Windows, BeOS, Redhat, FreeBSD, WinCE, PalmOS, QNX and I am now on the OBOS Coding Team.
Who could afford a Mac anyway ??? They are 3 times the price of an Intel PC. Maybe it’s different in the US, I assume that’s where you’re from. But I never come across Macs. Most businesses don’t use them. I’ve never seen a Bank using them.
I always get to these discussions late…
“[the CPU load for decoding mp3 on a 233 BeIA device is about 40%, so most likely you can’t even play 2 mp3’s at the same time]”
Well, since we’re making assumptions, I’ll assume that the BeIA is quite similar (if not, more improved) to the BeOS. On my 166Mhz machine with 64MB of RAM, I could run 3 .mp3’s at the same time with 1 of them playing backwards in the BeOS. Just for the record.
“So maybe it was easyer to write some small apps, but probably it was dificult to write bigger one’s.”
Would this explain why the more sophisticated applications I downloaded are less than 10MB?
*cough* Cygwin *cough*
Hmm…I guess that makes Windows Unix-like, eh?
I thought cgwin was just an emulated command line…
Uhhh…this comment was in answer to a previous statement that claimed that a BASH shell made the OS UNIX-like. That would be a command line, now, wouldn’t it?
And as a matter of fact, Xfree86 can run under Cygwin as well.
>>What? Your talkin about the G3s and G4s? Last time I checked BeOS didn’t have support for the G3 and G4 macs because apple refused to give them the necessary infomation, and they didn’t want to reverse engineer the systems.<<
It depends on what G3 you’re talking about… as far as I know, aftermarket G3 upgrades for Macs would run BeOS with no problems!
>>Where did you dream this up?! I own a BeBox (Rev-6 PPC driven) and compared to a similarly equipped PC… Don’t even get me started, that is all I got to say!
I didn’t know you could do SMP with the old Pentiums 133s…<<
I am talking from clockspeed level and not the number of CPUs on the motherboard!
Well, I also heard that Apple was not too helpful to Gassee re getting Be up on the G4. And Intel threw some bucks at him. No wonder he talked up the X86 version of BeOS. One hand washes the other.
But, about why OS’s rise and fall– who knows? Why does Ford motor company still exist? Because people are too lazy to do a little RESEARCH, I think. I can’t figure why Win2K still sells , given its tendency to go south if you look at it sideways. This is memory protection? I can’t use it an hour without praying that it doesn’t go down in the middle of an important job. I primarily use OS X, and it is far from perfect, but it’s the best thing out there. If I was afraid to spend more than $5 on a computer, I might be using some distro of LINUX or BSD.
I believe, are related to the motherboard chipset, not the processor. That’s why the upgrade cards work, but not the Apple-built machines.
Why did this guy write this article? Did he ever use BeOS? Did he ever read their marketing? And why do he put words in the mouth of "The BeOS zealots" as he call us? I for one do think that Microsofts OEM dealings did things worse for the Be team, but not that it is the whole explanation. The thing that imho killed BeOS is none of the factors he mentoned, rather lack of drivers, lack of professional software in some areas and lost trust when shifting focus to BeAI.
And when he starts mumbling about Amiga OS, I couldn’t help myself from laughing out loud. Did he really try to draw a connection between their demises? It sure sounded like that.
Actually, I don’t think he got one reason of the Be failure correct, but that might just be me. Actually it was some of the worst piece of **** I have read in a long time on a supposedly serious site.
This is my first time using BeOS and as you can see, I am already writing and sending a response for what I have read here. I have used all versions of Windows and RedHat Linux 6.2. And everybody knows the issues, complexities, and weaknesses of these OS’s. Should we stand still? Definitely not. Linux has benefitted of thousands of colaborators around the world, 24/7. We should do the same thing. BeOS can become a spectacular success, but that success lies only in our hands. It is up to us to develop it. ONLY WE CAN MAKE A DIFFERENCE.
Yes, the BeOS APIs were (mostly) a subset of Unix. In contrast Cygwin is actually built on top of Win32 APIs (because the POSIX APIs are deliberately crippled) and you can see how strained this relationship is all the time when using it.
The most pressing issues in BeOS R5 can be described as “It isn’t Unix enough”, and the response from Be, and now from OBOS is “let’s make it more like Unix”.
Since Eugenia and others will otherwise probably write poorly considered flames in response to this, let me defend my position up front:
(1) No memory mapped files, proposed solution is to add Unix-style mmap() system call
(2) B0rken shared lib implementation, proposed solution is to add Unix-style dynamic ELF object loader on mmap()
(3) Poor net performance, stability and utility, proposed solution is kernel-based BSD-style (Unix) sockets
(4) No hardware 3D acceleration, proposd solution is OpenGL with DRI-style (Unix) drivers based on Mesa.
(5) Threaded Windowing system is frustrating to program, proposed solution (not for OBOS R1) is non-threaded, and resembles X (Unix) much more strongly than OpenStep or Win32.
Apple got this right, for all the ranting on forums like OSNews, most users don’t care what the kernel looks like, so it makes sense to benefit from experience and use Unix.
I tried BeOS on my Windows machine. It is a great OS, and I would take it over Windows.
However, despite all its advantages, I don’t think 95% of the people would appreciate (or need) the improved design of BeOS. As mentioned above, it wouldn’t be noticable in day to day work for most people. The lack of programs would be noticed.
I do think Apple made the right choice. Going with a Unix based OS gives them more of a tie in to the outside world. A tie in that was needed. There was already plenty of Unix programs out there, and people were already looking for a Unix OS with an easy to use interface.
Where BeOS didn’t take off on the PC, I now see people asking Apple of an x86 version of Mac OS X.
And yes, Be management blew it, they asked too much $$ and took thier presentation to Apple too lightly.
And I don’t think Apple would be doing near as good if Jobs hadn’t come back there.
P.S. Apple also looked into using Win NT and Solaris as their new OS.
I just wanted to point out that Carbon was only considered after Apple bought NeXT. The story goes as follows:
1996 Apple buys NeXT.
1997 Apple starts working on Rhapsody. The OS will use OpenStep for native applications and offer a MacOS emulator (also known as Yellow Box and Blue Box, now Cocoa and Classic). Adobe thinks that rewriting their apps in Cocoa is decidly uncool and tell Apple that they will jump ship. This leads to:
1998 Steve Jobs announces MacOS X with Carbon.
A quick word before my quick response…
Since Eugenia and others will otherwise probably write poorly considered flames in response to this
Relax…responses/dissent != flames.
BTW, I can’t remember Eugenia flaming somebody, ever, with the possible exception of her rising to her own defense in response to some sort of unwarranted attack. Why people like picking on her, I’ll never know.
Yes, the BeOS APIs were (mostly) a subset of Unix.
That’ s just not accurate. The BeOS APIs were a unique application framework many coders loved, among their other advantages (and drawbacks).
Perhaps you were thinking of the POSIX compatibility layer (which was mostly, but as you note far from entirely, POSIX-compatible in some crucial ways)?
The most pressing issues in BeOS R5 can be described as “It isn’t Unix enough”, and the response from Be, and now from OBOS is “let’s make it more like Unix”.
That may be a side goal, but it’s not the principal goal. OBOS is attempting to remake BeOS 5.03’s general feature set with binary compatible executables and APIs, with open source. I can’t remember Be ever was ever “make the product more like Unix” (unless providing better interoperability and compatibility is the same thing).
Maybe we’re picking nits here, but OBOS is not out to turn BeOS (may it more or less rest in peace) into Unix. Part of Be’s rallying cry was always that Unix sucks on the desktop due to its moldy cruft.
I think you’ll find little disagreement that any future BeOS-ish OS would benefit from some completely rewritten components, including (perhaps first on the list) rewritten virtual memory that implements mmap in the POSIX layer.
On a couple of your other points…
(2) B0rken shared lib implementation, proposed solution is to add Unix-style dynamic ELF object loader on mmap()
Broken how? Being different from Unix != broken. If it was broken, I would (genuinely, seriously) like to learn more about that.
(3) Poor net performance, stability and utility, proposed solution is kernel-based BSD-style (Unix) sockets
I believe (but I confess I’m not certain) that the OBOS team is trying to imitate BONE, Be’s unreleased rewrite of the entire networking layer. BONE certain implemented BSD, but was certainly not equivalent to BSD, and included no BSD code whatsoever.
(4) No hardware 3D acceleration, proposd solution is OpenGL with DRI-style (Unix) drivers based on Mesa.
Partly correct here…Be’s unreleases (yet somehow ubiquitous due to file sharing) OpenGL rewrite was in wide use. It kicked ass, actually.
(5) Threaded Windowing system is frustrating to program, proposed solution (not for OBOS R1) is non-threaded, and resembles X (Unix) much more strongly than OpenStep or Win32.
I believe this is a complementary solution, not a replacement solution. As for “frustrating,” some programmers really love(d) the threaded Windowing system once they were able to get a grip on the relative complexity. (It was one of BeOS tougher learning curves.)
I really don’t know whether such a complementary approach will succeed, since there are drawback to the Unix approach.
Apple got this right, for all the ranting on forums like
Of course those features are in OS X…OS X is BSD Unix. (Sort of…you know what I mean).
OSNews, most users don’t care what the kernel looks like, so it makes sense to benefit from experience and use Unix.
This is the only statement with which I strongly disagree. The world has enough Unix (and Unix-y) distributions out there. Remaking OBOS into yet another Unix, with few if any unique advantages growing out of its BeOS heritage (which presumably you’d like to see removed), make zero sense to me. The duplication of effort is a huge waste of time for developers and users alike.
If you want to use Unix, use Unix–don’t use BeOS.