“As World+Dog gets its head around Windows Vista, lets look back at an operating system that might have been a contender, very nearly becoming Apple’s next-generation OS and, but for Linux, almost certainly the key alternative to Windows on x86. Ladies and gentlemen, who remembers BeOS?”
But greed killed it off. It would be interesting to see where OS X would be if it were BeOS based. Finder would have decent networking for once and multi-processor support would be superb out of the box (it’s still sub-par on OS X). However there would be some major omissions. BeOS is not a multi user OS, the lack of UNIX underpinnings and permissions would mean that a BeOS-X would be more akin to OS9, and would probably would have much lesser security than current OS-X. On the flip side, Spotlight would have shipped with the .0
You are thinking about BeOS as it now (5.0). When Apple was shopping for an OS, BeOS was in a far less advanced state; shoddy networking, terrible printer support, etc. It would have taken some work to get BeOS up to Apple standards and Apple felt NeXT offered a better base to build on.
No kidding, I almost thought the original poster was kidding when he said “decent networking” when compared with OpenStep. I used BeOS in those days (4.x) release, and I was a beta tester from way back, and the networking stack sucked in a major way, ugh.
It was a nice system, however BeOS bigots seemed to be under the impresion that the major selling point to an OS is how fast it could boot.
Gasse was miopic both regarding technology and market, the product he had was not worth the hundreds of millions of dollars he wanted to ransom Apple for. NextStep was a far more mature system, the API was light years ahead, and at least it had some kind of software base. Also the engineering team that Jobs was bringing from NeXT was worth much more than anythign Gasse could offer.
Gasse was also painted as much more of a cheerleader than actually one to be grounded in reality – he spent too much time trying to court Apple instead of throwing in the towel on PowerPC back when they made the leap to ELF back in R4, worse still over promising and grand standing in regards perceived evils of its competition – everything really came unstuck with this whole ‘appliance’ fad.
To get BeOS up and running in the required time, firstly Be would have had to drop PowerPC at R3 and made R4 completely focused on Intel, secondly, they needed more programmers, far more than they had to make massive improvements in the system to bring it up to speed; thirdly they needed major cash injection so then atleast along with the operating system, they also provided atleast a decent level of middleware for the markets that they were supposidly going to serve.
All this woul have costed $700million plus, and I doubt that there would have been VC’s willing to through that sort of money into Be given its terrible management team at the time – its the equivilance of giving money to a corrupt government and hoping a turn of conscience will force the leaders to invest the money wisely.
You are thinking about BeOS as it now (5.0). When Apple was shopping for an OS, BeOS was in a far less advanced state; shoddy networking, terrible printer support, etc. It would have taken some work to get BeOS up to Apple standards and Apple felt NeXT offered a better base to build on.
The point made in your first sentence is valid, but you don’t mention that – by the same token – OS X today is hardly identical to NeXTSTEP circa 1997.
Comparing OS X today to R5 is equally misleading – since OS X is essentially NeXTSTEP + billions of dollars worth of resources sunk into its improvement by Apple. If Apple had purchased Be Inc. instead and sunk those billions of $ into BeOS development, then does anyone honestly believe it wouldn’t have improved at least as dramatically as OS X has since it was NeXT?
let’s not forget that Be also had a much larger price tag on it. It all comes down to money in the end.
*sob* *sob*
Me too! I even ordered it with a T-shirt. I got a lot of quizzical looks! I just wanted so bad for it to take off and broaden the market and make computing fun. And then Linux came along…
Just tonight, I evaluated PCLinuxOS 0.94/Test 1 and was very excited at how well it ran off the Live CD on my Dell M70 Laptop. Linux is going to do what BeOS tried to do and failed I think.
yeah pclinux will save the day. uh huh just like Corel linux, mandrake and linspire and suse (which is now ‘evil’) and all the other “yet another linux” distros.
I’m sorry to be so cynical but I remember what the linux community said after Be Inc died. oh don’t worry, a real time patch here and there and ReiserFS and another foundation or “base” and linux will take on Windows. Fast forward to 2007…nope. Still only a couple % marketshare. Linux is free, tried, had a fair shot and failed.
It’s going to take another OS designed by different principles to have a chance on PC’s. It will take a lot of hard work but it’s the right work.
I’m sorry to be so cynical but I remember what the linux community said after Be Inc died. oh don’t worry, a real time patch here and there and ReiserFS and another foundation or “base” and linux will take on Windows. Fast forward to 2007…nope. Still only a couple % marketshare. Linux is free, tried, had a fair shot and failed.
Oh, ye of little faith. I’m sorry to be so realistic but the only way Linux was going to take over the world in a flash, before or after BeOS died was if Microsoft lay down and took it. They didn’t and they aren’t going to, so in order to win, Linux (which is still the only OS which has a chance of replacing Windows, at this time) is going to have to hang them out to dry.
well, everybody always has been hoping linux (or Mac OS X or BeOS or whatever) would take the windows world by storm. Think about the ‘this is the year of the Linux Desktop’. But it never happened, and probably never will. It’ll go slowly. Linux is growing faster than Windows, and has been doing so for years, and if it continues to do so, there will be a day it will have a majority marketshare. But it’ll take years and years…
well, everybody always has been hoping linux (or Mac OS X or BeOS or whatever) would take the windows world by storm.
I’m not really sure that would be truly desirable. I live in a small city and I wouldn’t dream of give up the advantages of living in a place with low population density (no real hardcore crime, no smog, half an hour is considered a significant commute, you can rent a 2-bedroom for $6-700 a month, etc).
While I do wish there was more in the way of decent restaurants around here (for example), I realize that all the other problems of a high population/population density would come along for ride if we had enough people to support those sorts of things.
I find BeOS has the same “small town-ey” appeal to me. And I’ll be indulgent for a moment by carrying on the analogy: I find Windows gives me that same feeling of “I’m being overwhelmed by things I don’t care the least bit about” that I get when visiting some huge, crowded, dirty metropolis (my litmus test: if your automotive traffic moves slower than horse-drawn traffic 100 years ago, then it’s too damn crowded).
Hey, at least I didn’t make an “if-OSes-were-cars” analogy
linux has a fundamental problem that simply can’t be fixed by writing more code. It’s not unified. No one to turn to for a sense of direction or accountability over the entire OS. This is what you need to have a chance against Microsoft or even Apple.
but if linux had a roadmap…it just wouldn’t be the same anymore. that’s not what it is supposed to be: a chaotic hotbed of ideas and choices
Erm, first you say that “Linux has a problem”, and then you deny that it is a problem.
It’s not a problem anymore than having more than one cooker manufacturer (to avoid the tired analogy) is a problem.
It’s a problem if you want linux to take on Microsoft and get the latest games and full featured apps like Quicken and Photoshop.
It’s not a problem if you don’t care about marketshare and just want to use linux basically the same way you are right now. It will improve. It will be there for you. But it’s not the same thing…
Edited 2007-01-31 10:49
It’s a problem if you want linux to take on Microsoft and get the latest games and full featured apps like Quicken and Photoshop.
Why?
Because so far, the Linux community haven’t come up with anything that can match Quicken (or Money) or Photoshop in terms of functionality and ease of use. If you want market share, that’s the level of app quality you’re aiming for.
Now folk like to say that OO has 90% of the functionality of MS Office, but having used both, I’ve found that the missing 10% seems to contain a lot of stuff I want to do.
Because so far, the Linux community haven’t come up with anything that can match Quicken (or Money) or Photoshop in terms of functionality and ease of use. If you want market share, that’s the level of app quality you’re aiming for.
Now folk like to say that OO has 90% of the functionality of MS Office, but having used both, I’ve found that the missing 10% seems to contain a lot of stuff I want to do.
Well that’s your opinion. It’s also possible that you’re in the only 10% of people for whom OO.org is NOT a suitable MS Office replacement. It’s also worth pointing out that Quicken is no use to you if you live outside North America, and that Photoshop is no use to you if (a) you don’t have the money and (b) you don’t want to pirate software or (c) future versions of Photoshop come DRM-crippled.
Regardless, that has absolutely nothing whatsoever to do with the question of whether having X number of Linux distributions (where X>1) is a bad thing.
I don’t know what you’re talking about in regards to “no sense of direction or accountability over the entire OS”. That’s exactly what each distro does–provide a sense of direction/accountability to the chaotic hotbed of ideas and choices that is the OSS community.
The only difference between (Mac vs Windows) and for instance (Red Hat vs Suse) is that those Linux distros get to use the work that the other ones have done, and also have a much easier time of attaining compatibility with one another. Which is a huge benefit.
We don’t need “one unified Linux” anymore than we need “one unified OS to rule the world”. The existence of various distros provides for *focused platforms* to experiment in a variety of directions. That generally doesn’t affect *application compatibility*, and when it does, there’s almost always a way to work around the issue. Linux distros by and large ARE compatible with each other (given a little bit of packaging work), and the situation is getting better all the time (with the likes of Autopackage and other universal install systems getting more and more viable all the time).
All that the “one unified Linux” idea would serve to do is kill the distro-specific innovation that comes from this diverse ecosystem. This diversity is Linux’s inherent strength, not its weakness. And I can’t see any way that it somehow damages mainstream adoption (given the fact that application compatibility really IS NOT an issue).
(edit: typo)
Edited 2007-01-31 17:26
Still only a couple % marketshare. Linux is free, tried, had a fair shot and failed.
When I buy a toaster, microwave or washing machine, I do not want to have to read the directions. I want to just plug it in and use it. A great many people have that exact attitude about PCs–which is the market Windows caters to.
I believe BeOS was competing for the same market–the market of the dumb. But the dumb reach for the dumbest so BeOS didn’t gain any market share.
Many GNU/Linux distros also make attempts to cater to the dumb, but there are too many contributors to the GNU/Linux sphere that favor excellence over simplicity for it succeed in dumbing that far down. So we may just have to settle for the market of excellence. I can live with that.
Hmm I feel like you’re treating the readers of your post as if they are the “dumb ones” of whom you speak. You come up with a dumbed-down equation of “simple=dumb=lesser-quality”, and explain away BeOS’s failure as simply “not being dumb enough”. I’m sorry, but that’s the most overly simplistic, nuance-less, intelligence-insulting BS I have ever heard.
Call me idealistic, but I believe it IS possible to make rock-solid apps from a stability & feature standpoint, AND have them be accessible to a majority of people. That was BeOS’s goal, I think. And contrary to your blanket assessment, I believe that there are in fact many OSS contributors out there who share this worthy goal.
Edited 2007-01-31 17:40
I’m sorry, but that’s the most overly simplistic, nuance-less, intelligence-insulting BS I have ever heard.
Simplistic – yes, I intentionally simplified the issue to avoid writing a book.
Nuance-less – Nuance, like beauty, is in the eye of the beholder.
Intelligence-insulting – I contend that the attitude I described is dumb and there are many people who willfully take on that attitude. That only applies to people who willfully choose to adopt a dumb attitude.
I join you in believing it is possible to make an excellent app accessible (i.e. easy to use), but when that becomes the focus of everything, it crowds out the excellent. The bash command-line, for example, is extremely excellent. Will Windows’ new shell rise to that level? (I don’t know what BeOS had.)
Intelligence-insulting – I contend that the attitude I described is dumb and there are many people who willfully take on that attitude. That only applies to people who willfully choose to adopt a dumb attitude.
The attitude of wanting something to work without poring over a manual? I would characterize that as a sensible attitude, not a dumb one.
but when that becomes the focus of everything, it crowds out the excellent.
Of course – an app is useless if it lacks necessary features / can’t produce an acceptable end result, regardless of how easy-to-use it is. If that weren’t the case, then all software would be designed like the AOL client
But the best applications, to my mind, are the ones that achieve a balance between functionality/flexibility and ease of use. To steal a quote from Douglas Adams, “There is no problem so complex that it cannot be made simple if approached in just the right way.”
Usability isn’t about accommodating the dumb, IMO, it’s about accommodating those who want to get from A to B with as little hassle as possible.
The bash command-line, for example, is extremely excellent. Will Windows’ new shell rise to that level? (I don’t know what BeOS had.)
BeOS does include a port of BASH, and I totally agree on its excellence. The BeOS terminal application itself is a good example of what I mean – it’s small touches in the usability/ease-of-use department are one of the main reasons I typically do command line tasks from BeOS rather than Windows (that and the more useful/more numerous apps). E.g., dragging-n-dropping a file onto a terminal window inserts its path; it works with standard copy-paste commands (unlike cmd.exe); and if multiple lines of text are pasted into the terminal, each line will be executed in turn (handy for quickly testing a shell script, or to avoid creating one altogether for one-time tasks).
BeOS was great. It really was, and arguably still is, but at the same time if Apple hadn’t bought NeXT, they wouldn’t have gotten Steve back.
Edited 2007-01-31 00:12
Yes, Apple sure is fortunate that Teh Steve returned just in time to personally write all the code for OS X, as well as designing the iMacs and iPods (and personally assembling each unit by hand, no doubt).
Yes, Apple sure is fortunate that Teh Steve returned just in time to personally write all the code for OS X, as well as designing the iMacs and iPods (and personally assembling each unit by hand, no doubt).
I get the joke but there’s more than an element of truth in it… Jobs is known to be a complete control freak and does have a big hand in the design of the products.
During the development of the iPod it is said he was involved on a daily basis chopping and changing things (especially UI related).
I do remember it, because I bought it some time ago
Was really nice
they should of just bought both and put them together.
Because of course software engineering really is that straightforward…
they should have bought steve and bought be, inc.
Not that the reg is good with facts.. unless it is syndicated or connected with flogging t-shirts it has turned into the IT equivalent of “That’s Life”…
Shame really.
Regarding the comments from Kroc, it did have POSIX underpinnings to an extent and multiuser, although was not present in BeOS had the fundamental building blocks in place with the file system and process control etc…
OS X would never have been BeOS based, remember Jobs and JLG were both ex Apple execs who split to form their own companies. I don’t think you would have ever seen JLG and Jobs on the same platform.
Jobs is a manager and a businessman, JLG is a businessman who knows and loves technology, this is his weekness, he does not seem to compromise his beliefs for a quick buck and does speak his mind.
If JLG was running the show at Apple I would have more white boxes littering the place than the Mac Shop iPod return department.
Actually, BeOS does not have POSIX underpinnings. It has a POSIX layer that is sitting on top of the BeOS native kernel et al.
http://en.wikipedia.org/wiki/GNU_C_Library
BeOS’ libroot.so, it’s C library – used by all BeOS applications (C/C++) – is Be’s implementation* of POSIX. The kernel in turn provides the necessary interface and services that the C library expects. This is very similar to Linux, BSD & friends. A kernel, a C library, more libraries and programs linked to these libraries.
In short: BeOS does not have a separate POSIX -layer-. It’s very much an integrated part of the system. (Note: This, however, does not make BeOS a unix.)
*(GNU’s implementation, if you like.)
Edited 2007-01-31 17:29
Jobs is a manager and a businessman, JLG is a businessman who knows and loves technology, this is his weekness, he does not seem to compromise his beliefs for a quick buck and does speak his mind.
I get it… you’re implying the The Steve DOES compromise his beliefs for a quick buck.
…I guess it’s a matter of opinion. I firmly believe that keeping the Mac closed is an example of The Steve *sticking with* his beliefs. Are those beliefs to an extent profit-oriented? Yes. But do they allow Steve to keep investing in making the cool technology he wants to make? Yes.
Admittedly, I’m not happy with the whole iTunes/iPod side of things, since I refuse to buy media that limits the devices on which I can play it. But do I believe Jobs potentially averted much more disastrous outcomes with his implementation, thereby opening up consumers and the industry to the idea of buying music online? Absolutely.
Also note: The Steve has had enough experience that he now only makes drastic decisions when he is *sure* he can win. We will see things like the Tablet Mac and the licensing of OS X, only when such scenarios reach *the point of assured success*.
I remember – and still sometimes use my Be machine. It’s an AMD 766Mhz with 512Ram with a Hauppage TV Tuner – and it still feels more responsive than my dad’s 2.4Ghz WinXP Dell… That’s still impressive to me considering the machine’s about ten years old!
The only major problem I found with BeOS was in developing applications – the BeApi was (to me) unavailable and obscure. I can write C++ (it’s not exactly my favorite language), but I could never find any substancial references to work from. Apparently, Java on BeOS seems to be coming together – so that would end up being my language choice at this point.
BeOS was (and still is) an amazing feat of software engineering, and I really hope Haiku gets up and fully-running soon. If you’ve still got an old computer laying around (and who doesn’t these days?) give BeOS Max 3.1/4 a try and you might just find yourself (if nothing else) an awesome arcade machine!
BeOS … yeah … back then … it had some features that Microsoft only offers now with Vista, like the set volume per applicaiton.
But well … Be is a part of the past. And now we have ZETA and Haiku. But ZETA doesn’t feel as fast as BeOS used to be, i think. But well, it has multi-user support since version 1.5, inclusive file permissions and everything.
IMHO, the biggest thing that BeOS had going against it was the lack of a decent web browser. (Ironically I’m running NetPositive right now since Firefox crashed on my last Woot!-off and took my bag of crap with it.) If it hadn’t been for that little fact, I’d say that R5 PE would have caught on. It was fast, better at playing media than Windows, more stable, and the learning curve was extremely slight for Windows users. It also didn’t require that you repartition, so people could try it out fairly easily. Even the media coverage was decent. I actually used it quite a bit back then, but that was because my boarding school’s firewall didn’t work with it. So I could use Napster, usenet, FTP, and even visit certain sites that every 15 year old male wants to go to. More recently, I switched to BeOS for 90% of the stuff I do because Windows did something that really pissed me off (I forget what that was now). Linux was being a pain to install (I tried 4 distros and only one would install), so I decided to pick up the BeOS once again. Needless to say, with Firefox it became perfect for most of the stuff I do.
I’ve seen more in-depth articles about Beos in the Past.
This one does not tell anything more. I also think it doesn’t put much attention to Haiku… which indeed is the most important thing in BeOS world right now (you can argue that Zeta is the most important piece of software right now: i still don’t think so.)
Agreed, this article had nothing new. I was expecting a forecast as to where things “might-have-been” had it been chosen as the basis for OSX. Actually that was the OSnews tag for this article …
MacOS today (it would not be called “X”) would probably look a lot like OS 9, very responsive, no objective C or any flashy effects. They would have kept the usability experts like Tog.
I don’t think microsoft was really worried about linux as a desktop OS. But they were really worried about BeOS as a long term threat and weren’t going to let that happen.
BeOS wasn’t perfect but even if it was, would that make much difference if Apple could get Steve? With Steve they get not just an OS but also the imac, ipod, hollywood, etc. So what if nobody else used objective C? (you can debate C++ vs C vs java but do you know anyone who uses objective c???)
It also didn’t help that JLG screwed up every project he touched.
BeOS even in 2007 still remains the best foundation for a new desktop OS. It doesn’t deserve amiga syndrome which at least had lots of commercial apps and games for it.
Actually attempted to try it out as well. The r5PE i think it was. Used 512K at the most alloted, and had to be run from within windows if I remember correctly. I do know I could never get it to boot, always hung with no indication of why.
It certainly did not run in Windows – it just lived on your Windows partition. Booting into PE would reboot the machine into BeOS. What machine did you try it on? Hardware can be a major factor of it hanging on boot.
“It certainly did not run in Windows – it just lived on your Windows partition. Booting into PE would reboot the machine into BeOS. What machine did you try it on? Hardware can be a major factor of it hanging on boot.”
Well, for the PE it required Windows to install, but no, it did not run in Windows. The professional version did not require any OS to be on the machine. I can’t remember the machine exactly, hell it was back in 1999 or so when I tried it, so would have been a Pentium at probably 100 MHZ or so with what I had back then. I tried it again with a fresh download after seeing this article on a Dell GX110 I happen to have for older testing, and the same thing. No non-standard hardware.
Before you blast me for being off topic, please read to the end.
A potential niche for Haiku to fill could be high end audio. Create a music production system that is “all in one”. Basically, play Apple’s game, but take it to the next step. Integrate X hardware with Y OS with Z software.
Get some pro level gear support for an audio interface, pro level midi, and *working* VST plugin support. Take something like Ardour, give the guy some cash for a native BeOS port (or chip in help, however you want to go about it), sort out the VST issues, get better MIDI editing tossed into the mix, and bam – you’ve got a kick-butt music production system, completely integrated.
We know Haiku has limited hardware support, but in this instance – it’s fine. Pro audio guys don’t care, heck they’re buying macs already, and they only come in a few configurations.
All I know, I shelled out $5k+ on a mac pro, $1k on logic pro, and deal with all kinds of “issues” due to OSX/Logic’s closed architecture. Bugs don’t get squashed if they aren’t deemed important enough to prevent revenue generation for Apple. They are a business, I don’t blame them.
However, if you want to get Haiku adoption rates up there, offer an alternative that performs better, and can do it for much less money – and you’re a winner.
One key is getting third party plugins functional, since most musicians rely on them (and they heavily influence buying habits of DAWs and systems.) That said, if alternative OSS plugins are developed which can replace things such as Ivory, Reaktor, etc – that would make it non-necessary for most new users who don’t mind relearning. That would be the target market, regardless.
——————-
Apple made a good choice at the time they did with NeXT, getting Jobs (as much as I cannot stand his personality, he has done wonders for the company.) As others mentioned, the networking in BeOS (including 5) was terrible, it only got better in the unreleased Dano. There were many other niggles “back then” that made NeXT a better choice. BeOS had some benefits, but in terms of overall worth, I think Apple decided correctly. Most shareholders will agree.
That said, nostalgia for “what could have been” and BeOS doesn’t get Haiku any further. We should learn from the success of modern day Apple, and run with it! Embrace the past, learn from it, and move forward. Haiku is playing catch-up now in some areas, the most important being marketshare. A “killer” use/application/whatever like suggested above would really help propel it into the limelight.
Good luck Haiku, the past is the past, let’s move forward!
if you say the networking in beOS was terrible, I could say the responsiveness in OS X 10.0 was really terrible.
That’s not what really mattered. It was the aqua, icons and dock that sold the OS. Even Nextstep had a cooler UI than BeOS/Zeta/Haiku yellow and grey. yuck
Looks will always win.
10.0 was a “alpha” of sorts, that’s why they gave out free upgrades. It was very prematurely released. Just as a FYI – nobody who actually really *used* their machines ran 10.0 by itself, if at all. Most still stuck to OS9 through that HORRIBLE transition.
NeXT was a much more “usable” product than BeOS at the time of Apple’s decision, which was the intent of my post to clarify. Just because Apple fumbled and released an “alpha” quality hacked-up derivative doesn’t mean BeOS would have been a better choice.
and everyone still says MacOS X is too slow compared to Windows on the same hardware. Hence “teh snappy” and “FTFF” comments.
“Yuck” is in the eye of the beholder. It is exactly because of the red/blue/yellow/grey colours that I first started using BeOS. I’m also a big fan of Piet Mondriaan.
Honestly, I think Haiku, if developed correctly and given enough time, could become what Desktop Linux is trying to accomplish with Beryl.
I mean, sure, the conservatives at Haiku forums are mostly dead-set against the idea of a hardware-accelerated Haiku desktop because of the supposed “frivolities” involved (not to mention the necessary hardware, which is still out of reach for the majority of computer users), compared to the acclaimed simplicity of the BeOS architecture.
However, BeOS, when first released (1991), was made to take advantage of modern desktop hardware from that period. From Wikipedia:
Optimized for digital media work, BeOS made full use of multiprocessor systems by utilizing modular I/O bandwidth, pervasive multithreading, preemptive multitasking and a custom 64-bit journaled file system known as BFS.
So my question is: why not make Haiku (post-Glass Elevator) take full advantage of 2007’s hardware?
That is, unless such a system like Haiku’s will find a niche that none of the other three desktops are destined to fill anytime soon.
I mean, Windows and Mac OS X have both become server operating systems (for those who want a user-friendly GUI), and Linux has become a home desktop operating system (for those who want….uh…better Internet?).
What if Haiku finds its niche in education?
hardware accelerated desktop yes. Absolutely. of course this depends on good drivers.
see
http://www.freelists.org/archives/haiku-appserver/10-2004/msg00046….
What you are probably referring to is the idea that Haiku should allow unlimited theming so that it can have thousands of goth skinz.
“I mean, Windows and Mac OS X have both become server operating systems (for those who want a user-friendly GUI), and Linux has become a home desktop operating system (for those who want….uh…better Internet?). What if Haiku finds its niche in education?”
OS X server is one solid system. It only promises to get better.
BeOS could have been a server system and it would have been great as one.
“BeOS could have been a server system and it would have been great as one.”
Are you sure? I always thought there was some sort of trade off involved like moving parts into the kernel that did not belong there to increase desktop performance, optimizing for responsiveness rather than throughput, …
Of course, the massive multithreading would totally rule on todays multi core servers.
You might know more about the matter but it seems to me that trying to be a jack of all trades will almost certainly compromise performance when compared to a solution that is specifically tailored for desktop or server.
“BeOS could have been a server system and it would have been great as one.”
Only if you removed everything that makes it a good desktop system and replaced that with things a server need.
As it was designed and released, BeOS was entirely unsuitable as a serious server system.
‘What if Haiku finds its niche in education?’
– You mean like RISC OS’s niche in British Education? That did a lot for Acorn and RISC OS.
Totally agree with you on the need for a good niche – I think the whole “Media OS” could still be a reality. At the very least, it could make the ideal “Media Hobbyist OS.”
The big advantage I see is that, in BeOS, simple multimedia tasks actually are simple. I convert a lot of vinyl records/audio cassettes to CD for people – and while I cut up the files using Audacity in Windows, I find that the actual recording is much much simpler in BeOS. Choosing the input source and setting the levels is much easier with the BeOS media prefs panel; it’s also nice that a simple-but-functional audio recording app is included with the OS itself. And when the recording is finished, saving the file is just a matter of right-dragging to the desktop and selecting “Make AIFF Clipping” from the pop-up menu.
A potential niche for Haiku to fill could be high end audio. Create a music production system that is “all in one”. Basically, play Apple’s game, but take it to the next step. Integrate X hardware with Y OS with Z software.
Actually, this has really happened: Tascam made a hardware DAW station using BeIA internally. I couldn’t find a spec sheet for one of the actual machines, but the press release is here:
http://www.crmav.com/46/tascam_and_be.shtml
Man, I want that system that you’re talking about, NOW!
Oh wait, I still have a system running it And it still rocks!
Why is it that whenever somebody says ‘Mac OS X’ somebody shout ‘BeOS’ as if they know better what makes the world turn around…
OpenStep and NeXTStep had superior APIs though BeOS might have very superior kernel designs and speed from the fully multi tasking and multithreading environment it was, but as has already been pointed out – e.g. networking sucked like nothing else…
The wish to see BeOS as a viable alternative to Windows NT was years away…just like Mac OS X 10.2 and above is just soooooo much better than 10.0…. Everything has its staring point and OpenStep was the more mature of itself and BeOS at the time…
I’d love for BeOS and Haiku to gain momentum, but the fact that the open source clone is still about 10 yrs behind the closed source ‘mothership’ makes me doubt that it’ll happen, which in turn make me glad Mac OS X happened, otherwise we’d still run Mac OS 9.6 featuring cooperative multitasking and no memory protection on our state of the art Macs… (That is, I have no faith in Apple having been able to conceive what has been conceived within these past 6-7 yrs on its own…)
Mod me down all the way, but this seems more or less like the battle between OS/2 and Windows NT… I looooove OS/2 but today it’s just old and outdated so I get on with my life!)
Edited 2007-01-31 01:53
“otherwise we’d still run Mac OS 9.6”
who is we? Most people want an OS that will run on any PC. As long as macos is only for mac, it’s not going to affect Haiku. BeOS was more relevant to Apple because of its history but Haiku is going after Windows on its own turf.
And you don’t think that’s exciting?
I said that, and I quote:
‘otherwise we’d still run Mac OS 9.6 featuring cooperative multitasking and no memory protection on our state of the art Macs…’
…thus by ‘we’ I mean ‘Mac users’, because face it: No Mac OS X (or BeOS for that matter) there’d be no Intel switch happening the past year…
Everything’d probably be the same old same old for Apple and Mac OS 9 would probably still be around in some updated version…
I’m not so sure about that because there are a lot more C++ developers in the world than objective c. I don’t argue that Steve is better than Be for Apple but it’s his sense of style…let’s put if this way. If Be made BeOS look like nextstep and Steve made nextstep look like BeOS, MacOS X would *probably* be based on BeOS.
Apple is now primarily about imac and mac mini. Computers for people who basically want to surf the web and check their email and that’s it.
No 3rd party apps for iphone? who cares? It’s cool. See what I’m getting at? There’s something really missing from Apple and it bothers me. It used to be more about doing more with your computer instead of treating it like a toy.
No 3rd party apps for iphone? who cares? It’s cool. See what I’m getting at? There’s something really missing from Apple and it bothers me. It used to be more about doing more with your computer instead of treating it like a toy.
First of all, the iPhone isn’t a computer. Second, maybe it’s not about doing more with the computer but doing the stuff you need to do in a simple way. Many Mac users like Macs mostly because it doesn’t ask much from the user, it just works and lets them get their work done. Granted, this won’t fit every user out there but nothing really does. And I had the opportunity to lay my hands in a Mac Mini and, IMHO, it can do way more than surfing the web.
OT: Remember when phones were used to talk to other people?
“Apple is now primarily about imac and mac mini. Computers for people who basically want to surf the web and check their email and that’s it.”
No, I rather think people who surf the web and check their email usually buy a new PC with “Windows” preinstalled (or let someone install a pirated copy for them), because nothing else exists. 🙂
One feature I did like about BeOS was the ability to dynamically load and unload drivers/processes. This is one of the things I think Linux distributions could learn from. Also BeOS paved the way for MacOS X/Windows XP, and all the MP3/WMA players out there. Too bad it wasn´t marketed enough.
Steveboy killed the ppc clone market because he knew BeOS was going to eat Apple’s lunch. That is what saved Apple.
he killed the clones because power computing was eating his lunch as far as hardware goes. I don’t they could’ve switched the mac community to BeOS by themselves.
Steveboy killed the ppc clone market because he knew BeOS was going to eat Apple’s lunch. That is what saved Apple.
That is the single most moronic comment regarding this BeOS should have been chose debate that has hung around since 1997.
The amount of technologies that Apple got that had yet to be released by NeXT are still being felt with each revision of OS X. Having worked for both I got to see it before such tools have been released.
such as Spotlight? oh wait…that was designed by a Be engineer…
Which one, though? It’s hard to keep track of which former Be engineers are fixing OS X for Apple, the fountain of innovation that they are.
Dominic Giampolo, author of BFS.
Is he still with Apple? I thought he was working with Google these days. Could be Pavel Cisler or Kenny Caruthers I’m thinking of, though.
Too bad they didn’t hire Mike Popovic, I’d really enjoy reading marketing materials penned by him (complete with sideways-looking statements).
Be had a shipping product (as well as momentum) and Apple didn’t squat until 2001.
That must have been murder on Apple’s colon. I’d suggest taking a laxative next time that happens.
BeOS was pretty interesting at the time when you compared it to Mac OS 8/9, Windows 98/NT and Linux 2.0 (maybe RH5 or 6?). It had some cool features and pushed processing into SMP territory which was pretty rare back in the late 90’s on a desktop computer.
The BeBox was pretty nifty by losing the L2 cache in favour of twin 603e CPU’s, but nowadays it’s quite slow from personal experience the last time I checked it out. The later Intel BeOS 5 ran circles around it on 400+ mHz Intel boxes with superior graphics hardware that had proper drivers.
Probably the biggest fault BeOS had was it didn’t really have the driver support it needed compared to Linux’s kernel drivers as well as no true blue Unix compatibility layer (it did support quite a bit, but not as much as GNU/Linux with X, libraries, etc). Just getting critical mass to support the BeOS API’s was a bit of a tough ask.
These days again and again it’s hardware support and application support that matters. All “the snap” and “way cool API’s” don’t mean much if you can’t run the hardware you own and you can’t run the applications you own.
what you are saying is true but at the same time, you can have all the hardware support like linux and not really go anywhere unless you have something ‘cool’ enough to get people to switch.
Steve Jobs has something ‘cool.’ It is NOT about dev tools or whether it has this or that feature. Do people care that MacOSX doesn’t have OLE, a basic feature Windows had for years? That’s not cool-that’s “business.”
If it was about the dev tools…checks the want ads for Objective C experience… Checks the number of business apps and games for macs…
Edited 2007-01-31 05:18
Exactly.
As I was saying, people want to run the programs that they own already (or are familiar with already) and utilise their existing hardware. BeOS didn’t really have it on both fronts (people wanted IE rather than NetPositive, MS Office rather than Gobe Productive), neither does GNU/Linux (but it’s getting there slowly with Open Office). My experience on BeOS 5 on Intel was “yeah, it’s OK. Just don’t see average Joe using this, however. No apps. No games. No nothing, really. Oh, and no good support for my GeForce 256 so I run in 16 colours kind of puts me off.”
As you surmise, the only way to compete is to provide something that’s cool, integrated and it “just works”. That’s what Apple is proving time and time again. However, don’t forget Apple also supports quite a few bits of hardware and software so people can get on with work when they don’t need to do their “cool” stuff.
I’m not sure how successful Apple would be if there were no business applications or games like MS Office, Quicken, MYOB or WoW. That was a major complaint against Apple back in the 90’s. Nowadays, people are more willing to compromise and learn new applications, but there’s still a large number of “indocrinated” users out there that refuse to learn anything that’s new (or different) to a large degree. (IT admins are particularly bad, when I think about it).
Ultimately, I think software developers need to provide the existing applications to users, then work on a limited subset application base that “just works” and does what say, the 90% of users want. We’ve been seeing some of this (Firefox/Thunderbird vs Mozilla/Netscape, AppleWorks vs MS Office, iPhoto vs say, Photoshop, iMovie/iDVD vs AVID/Adobe Premiere) but I don’t think it’s a formalised objective yet.
Users want simplicity and ease of use. If they don’t get that 90% done and need the extra 10%, they can go out and find an application that will do it, but don’t try to provide that last 10% at the risk of alienating the 90% of users who just want something that “just works”.
Well, ok… Steve Jobs would never go for it, but…
Does anyone know how to get Haiku onto an AMD Athlon64 system? I can’t get any version of BeOS to boot on it. The HaikuCD script allows me to create a CD, but it won’t boot at all, even on my Athlon XP system!
Does anyone HAVE Haiku running on their Athlon64 computer (assuming they own one)? I’m thinking the thing that’s stopping the BeOS MAX v3 CD from booting beyond the middle icon, is possibly the ATI Radeon 200P chipset on my MSI RS480-Neo2 motherboard. Any idea if that’s possible and/or how to work around it, if so?
In many ways, I find that BeOS is much more consistent with the general UI design philosophies of the MacOS (Pre-X) than OS X itself. Apple pretty much flushed away the carefully-refined document-centric UI of OS 8/9 in favour of the application-centric UI of Windows (and OS X today). I don’t devalue the results – Apple is just about the only OS maker that has succeeded in making a “Better Windows than Windows”. But it is rather disappointing to see them take the safe path of a app-centric UI, rather than something interesting or innovative.
For what it’s worth, I always liked the UI of OS 8/9 – but despised the OS in general for its fragility, technological backwardness, etc. For me, BeOS is all the advantages of the Pre-OS X UI (utilitarian, the flexibility that is inherent to a document-centric UI) but without all the deficiencies.
Even after 10 years of computing development and greater hardware, nothing comes close to the experience BeOS gave me on an x86.
Linux ain’t there.
OS-X is clunky and in your face flashy.
Windows is just a mess.
BeOS was also the only OS I have used that has given me true Plug and Play.
Changed out my motherboard, CPU, Video card, Sound Card for stuff that was better but supported and then had the OS boot into my desktop giving me full access to all the new hardware with all my old desktop settings intact. The real beauty was BeOS didn’t need to ask me 1 question, it just worked.
Windows 2K which was my other OS installed at the time had a heart attack and died on boot requiring a re-install. Couldn’t even safemode it and get it running that way and nothing has changed with Windows since.
BeOS and Multi-media, nothing came close to its capabilities even now. Just makes me sad to see how far computing has been held back by a crapola company.
Haiku is a salvation though.
“Windows 2K which was my other OS installed at the time had a heart attack and died on boot requiring a re-install. Couldn’t even safemode it and get it running that way and nothing has changed with Windows since.”
There are many ways to fix a non-booting Win2K system, a reinstall is the last resort. Windows 2k/XP/2003 are the most stable things to come out of Redmond, and while other OS’s can claim better stability, I have had only one instance (that I remember) where I couldn’t get Windows to boot, and had to reinstall, in all of 7 years.
Windows XP and Windows 2003 are even better than Win2k, I have no idea if Vista is better than them, but if it is, it’ll be rock solid. Sometimes it’s not the tools that is the problem
“Windows 2K which was my other OS installed at the time had a heart attack and died on boot requiring a re-install. Couldn’t even safemode it and get it running that way and nothing has changed with Windows since.”
There are many ways to fix a non-booting Win2K system, a reinstall is the last resort. Windows 2k/XP/2003 are the most stable things to come out of Redmond, and while other OS’s can claim better stability, I have had only one instance (that I remember) where I couldn’t get Windows to boot, and had to reinstall, in all of 7 years.
Windows XP and Windows 2003 are even better than Win2k, I have no idea if Vista is better than them, but if it is, it’ll be rock solid. Sometimes it’s not the tools that is the problem.
———–
The better way of putting it is that XP sucked less than 2k and Vista might suck less than XP. “The most stable…” meaning the others were horrible, which is true. When you have a billion bugs and you kill a million of them. Then yes, it is less buggy. That doesn’t mean there are hundreds of millions of bugs still there.
Actually, I don’t think they sucked, that’s your opinion, they just suck in different ways then Linux and BSD. No matter what you believe
Given that Windows 2K Pro couldn’t handle the hardware components I had replaced given that I had not uninstalled any drivers prior to the hardware upgrade.
The OS failed to get into it’s bootloader without crashing and none of the repair utilities on hand fixed that situation.
BeOS on the other hand which suffered the same treatment had no issue at all and required ZERO input on my behalf to get back up and running as a fully functional desktop.
Let me say that again incase your eyes have difficulty reading what I am saying.
BeOS had ZERO input on my behalf with a major hardware upgrade in
1. Booting
2. Restoring my Desktop to Exactly the same settings prior to the Hardware Upgrade with the exact same settings and functionality.
Hard to comprehend for most people but that was the experience BeOS gave. Nothing I have used since has had the same level capability.
There are many ways to fix a non-booting Win2K system, a reinstall is the last resort.
True, but it is necessary sometimes. I’ve done dozens of transfers of windows installations/drives containing windows installations to new hardware. Some of the time, you can get by just fine by booting into safemode and letting Windows detect the new hardware. But I’ve also seen Windows bluescreen on the new hardware and refuse to boot without a repair install (not to mention, I’ve also had repair installs make a system unbootable – in cases where the CD used to do the install wasn’t as up-to-date as the system being repaired).
I’ve had a lot more reliable success doing the same thing with BeOS – assuming compatible hardware on both systems, it typically *just works*.
BeOS was one of my first love , when i was a kid with linux just around the corner on the Desktop front the BEos provided a real alternative to windows on desktop but somehow it faded and was later brought by palm forpaltry 11Million $ so it was a big let down considering the good features the OS had i still have BeOS R 5 somewhere in Cd’s i would still make a switch if recent version comes out.
With haiku and other such beos project who knows in future we may have BeOS as a viable alternative on desktop.
Ambuj varshney < http://linuxondesktop.blogspot.com>
It was a great OS, and in some ways still is.
it is however ironically that the current Macs won’t run BeOS/Zeta/Haiku .. well, I haven’t succeeded in installing anyway.
I’m excited for Haiku 1.0 to be released someday, but IMHO Glass Elevator will quickly deteriorate without a visionary leading the project. Look at how Linux and open source in general are relegated to copying functionality of commercial applications. Without a cohesive vision (and a healthy R&D budget) they’re destined to follow, not lead. Still, it’ll be cool to see R5 running on modern hardware.
Just curious, but who owns the Be source code now? Palm? Access? If it’s Access, I wonder if they’d consider releasing it.
you may want to check out the glass elevator mailing list
http://www.bug-br.org.br/pipermail/glasselevator-talk/
I’m excited for Haiku 1.0 to be released someday, but IMHO Glass Elevator will quickly deteriorate without a visionary leading the project. Look at how Linux and open source in general are relegated to copying functionality of commercial applications. Without a cohesive vision (and a healthy R&D budget) they’re destined to follow, not lead.
We have nice concepts for R2 and R3, but you can’t expect anything of it to happen without the acceptance from our developers. This is much more important than a visionary.
Many innovative ideas won’t appear “useful” to geeks because they automate what they’d rather do manually (in the name of having “full control” and “more choice”) or because they make things easier that they don’t have a problem with. Often, the current system is “good enough” and the new idea breaks with concepts that everyone got used to (do you want to change your habits?).
Also, many minor usability improvements appear as nitpicking and the discussions are time-consuming (time-wasting) which raises disrespect for usability people. It’s important to establish a communication policy that gives both groups enough freedom and that doesn’t waste anyone’s time too much.
I think that the problem is more about politics than getting ideas.
This article inspired me revisit the Haiku site, and I must say, the new site is b-e-a-utiful!
“All this woul have costed $700million plus, and I doubt that there would have been VC’s willing to through that sort of money into Be given its terrible management team at the time – its the equivilance of giving money to a corrupt government and hoping a turn of conscience will force the leaders to invest the money wisely.”
Be’s management wasn’t any worse than that of NeXT. How much money did Steve burn thru?
…because I just installed R5.0.3 on an old P2.
Also, BeOS was single-user. A positive for some long-time Mac users, but a big priority for Apple to fix.
When Haiku implements it I want to see file permissions using ACLs in a unified and consistent way. I’d be disappointed to see Owner/Group/Other in the Tracker interface (not sure what Zeta does)
OS X’s ACL implementation is half-baked. To me, only the non-home Windows versions really do permissions well.
> Also, BeOS was single-user.
That’s not true.
BeOS had support for users on the fs and kernel level, and the system library had code to actually handle multiple users, but it was disabled by default in R5 (you had to export MULTIUSER=true in the shell and create users).
The lack of GUI login tool didn’t mean it couldn’t work as multiuser. I actually tried once with R5 and it worked. Wasn’t really safe (big security holes) but it was working.
How’d that work? E.g., could you just “su” between the created users, or did it actually do stuff like restrict access to processes/files/folders owned by other users?
Yes you could su (but it was buggy), and yes the Be file system had POSIX perms from the start on, however though the kernel enforced most of them it had some missing checks. But the layout was there. It was just disabled by Be because it was breaking many apps (badly coded apps which despite a made-for-that find_directory function used hardcoded paths from home directory or even put log files in their app dir).
I remember BeOS because it wouldn’t run on one of my PC’s and it didn’t have drivers for the video card on the other.
E’yup! The everlasting drivers issue. Be, Inc. focused on “Wow” and not nearly enough on “Do”. When the excitement wore off, people were left going “Now, what do I do with this and this and this?” All issues of compatibility and functionality on the Intel platform.
After a couple years of a lack of support for everything Windows automatically supports (and/or drivers are available for), people begin to wander back to what’s familiar and works.
Be, Inc. should have stayed with the BeBox. Improved and refined it. I was “this close” to buying one, when they stopped making them. I was quite upset, to say the least.
Now we’ll never see what the BeBox COULD have been, because they killed it off. What I wouldn’t give for the complete schematics for the BeBox. Any idea who owns (if it’s even in anyone’s possession) the rights to the schematics for the Dual 133MHz 603e BeBox design?
I’d probably be willing to pay up to $1,000 for ’em, as soon as I had it.
It’s not the BeBox, but if you truly want schematics, Genesi has released AFAIK complete schematics of their Pegasos II design, through Power.org. (Anyone can join, free.)
If you’re interested in Power-based hardware, visit http://www.powerdeveloper.org/
If you’re rich or a genious, you may be allowed to buy/test their 2×2 PPC970.
Even the professional version required Windows to install. The installation was a clunky joke: making floppies, then booting from those floppies.
It didn’t seem to run any faster than windows, or do anything that windows wouldn’t do.
“Even the professional version required Windows to install…”
I think this had Sooooo much to do with Windows liscencing issues. I don’t think it had ANYTHING to do with being unable to install without a boot floppy :o)
uh, what? My cd was bootable.
Even the professional version required Windows to install. The installation was a clunky joke: making floppies, then booting from those floppies.
You write about PERSONAL EDITION, which was ment as a kind of demo (it is installed inside Windows, just like any other application).
Professional version does not require Windows at all. It also does not require any floppies to boot.
Neither BeOS 3 Pro nor BeOS Pro 4 required Windows; I did installations of both on a Windowless system.
BTW, the box had prewritten floppy diskettes in it; no diskette creation was necessary.
The way I see it, Be.Inc ‘steered’ its way around various major problems during the period 1998-2000/1. I’m not saying it steered its way around them succesfully, as the company went belly-up big style.
For a small company with very little R&D money, I believe they did just about the ONLY thing they could after the door was shut on the clone market…they HAD to concentrate on the x86 port.
All the talk in this thread about ‘lack of this driver’, ‘crap networking’, no ‘tractor’ apps being the cause of the demise of the company and the OS is of little value. Be.Inc HAD to get BeOs pre-loaded on new machines to garner any sort of real mindshare/marketshare. This they did to some extent. Fujitsu/Seimens amongst a couple of other OEMs started selling dual-booting BeOS/Windows machines.
The rest is a no-brainer, MS didn’t like the competition, got heavy handed with the OEMs and Be.Inc became history and got a paltry 22 million for the priviledge IIRC.
I believe I am right in saying that the company could not survive on CD sales alone (of its fledgling OS).
heh! R5 is now playing ‘surrogate mother’ to my Haiku partition whenever I feel like checking-out nightly builds! R5 still manages to make my jaw drop every now and again compared to any or all of the OS’s available today. It is perhaps, still the greatest OS that not many people ever heard of :o)
Edit: Couple of typo’s/grammatical errors.
Edited 2007-01-31 16:43
The way I see it, Be.Inc ‘steered’ its way around various major problems during the period 1998-2000/1. I’m not saying it steered its way around them succesfully, as the company went belly-up big style.
For a small company with very little R&D money, I believe they did just about the ONLY thing they could after the door was shut on the clone market…they HAD to concentrate on the x86 port.
All the talk in this thread about ‘lack of this driver’, ‘crap networking’, no ‘tractor’ apps being the cause of the demise of the company and the OS is of little value. Be.Inc HAD to get BeOs pre-loaded on new machines to garner any sort of real mindshare/marketshare. This they did to some extent. Fujitsu/Seimens amongst a couple of other OEMs started selling dual-booting BeOS/Windows machines.
The rest is a no-brainer, MS didn’t like the competition, got heavy handed with the OEMs and Be.Inc became history and got a paltry 22 million for the priviledge IIRC.
I believe I am right in saying that the company could not survive on CD sales alone (of its fledgling OS).
heh! R5 is now playing ‘surrogate mother’ to my Haiku partition whenever I feel like checking-out nightly builds! R5 still manages to make my jaw drop every now and again compared to any or all of the OS’s available today. It is perhaps, still the greatest OS that not many people ever heard of :o)
Edit: Couple of typo’s/grammatical errors.
——–
Sorry for copying the whole post but OSNews has no way of letting people easily follow replies. I said easily.
Anyway, someone who actually knows the facts unlike 99% of the people that have posted here.
Edited 2007-01-31 18:25
If Apple bought BeOS there would have likely been no NeXT deal and no Steve Jobs. The company would be out of business by now.
Now that we’re through with the nostalgia, can we go on towards the future of BeOS? Zeta 1.5 is supposed to include multiuser functionality, all I need now is someone who writes a driver wrapper to use nVidia’s Linux drivers for OpenGL
ok, maybe for who doesen’t know the story it’s interesting…
but let’s talk about what’s new, Haiku/Zeta instead pls, or nothing… my 0.02€
Heh! Maybe the title of the thread ‘Might-Have-Been’ or the link ‘who remembers BeOS’ had something to do with all the nostalgia….
Anyways…agreed, the development/new stuff is the most important.
I saw it on I think The Screen Savers on Techtv, man I loved that network. Anyway it looked awesome I might “Be” using that right now if it was still alive.
Edited 2007-01-31 17:23
I’d love to own a BeBox myself. I figure at some point, people will stop overpaying for them and I can win one on eBay at a decent price.
Several people have cloned Apple I, IMSAI and Altair 8800 Machines (all quite nicely). Maybe someday, someone will clone the BeBox Logic Board and make clones available.
I’d certainly like one.
For now, I have an AMD 500mhz Pentium Clone Chip running BeOS 5.03 and it’s quite fast.
I also have a PowerMac 8500 I could reformat and make into a BeOS machine as well.
I wish BeOS PPC ran on the B&W Macs. I have several doing nothing right now. Though I booted Ubuntu on one, and that worked quite well.
I’m looking forward to Haiku 1.0, I would buy ZETA but it’s priced 2x what it’s worth. If they dropped the price to $50 US, I’d buy one right away.
Since I have a working BeOS System, and Haiku is on the way, I’m not going to invest that much money into Zeta.
Linux will not gain a significant market share until …
1 – average Jane and Joe can go into 90% of places that sell computers and can buy a computer with Linux (of their choice) already installed and ready to go.
2 – average Jane and Joe can see and play with and have it demonstrated for them how they can do all their normal functions in Linux.
Right now neither of those are in place. They don’t want to order it on the web or install it themselves or take it to someone to install it. They want it pre-installed and a very simple way to find and install other programs they need. Despite what people that post on OSNews says, Jane and Joe want something as least —LIKE— Linspire.
BeOS didn’t get market penetration for the same reason as above. If you can’t get your OS pre-installed on computers you have little chance of getting past geeks or family members of geeks.
It’s really just that simple.
As a last note. I bought BeOS 5 and BeProductive as a package. I really liked what it could do and used it as one of the seven OSs I had on my half dozen computers I had at the time.
It worked pretty well for what it could do. The big thing was the missing apps which kept sending me back to the other OSs.
I LOVED showing people how I could run 16 movies at one time in Windows on the BeOS 5 desktop. All would run without lost frames and the sound came from the active window. Click on another window and the sound instantly changed to that windows. EXCELLENT and fun.
As far as the look of BeOS. It was too plain. Thankfully some of the apps looked a lot better which made for a better experience. There is a reason why AMC Pacer and Gremlin didn’t catch on. They were UGLY. Sorry but BeOS is too plain.
some background:
i’ve been developing apps on BeOS starting from R3.0 I currently run BeOS (dano) as my primary work desktop, where i develop full time cutting edge distributed live-video software, and I run BeOS at home on a daul-pII along side a mac-mini. the mini has become my primary home desktop however (had to at least be able to load modern web pages and sink iTunes with my phone).
the artical:
it does a good job of restating the wikipedia entry and sparking lots of good (sometimes) conversation.
its core point is: “what if apple would have bought Be Inc.”
all in all, i don’t think things would have been much different then they are today, on the OSX front. the infulance and idealism of Jobs is another topic all together.
and heres why i think that:
BeOS in those days had already been moved from hobit to PPC with “minimal” effort and later moved to x86. I think in general this shows that the ability to migrate (as apple has now done) would have still existed. (not to mention in the end Be Inc did this migration to x86 long before apple did)
people have mentioned already many of the things that would have been a quicker to-market time for apple, most notably BFS’s live queries would have put Spotlight in much earlier versions of OSX. i also think be’s flexible add-on system and multi cpu support would have been huge advantages.
networking is another big place that BeOS was “not so hot on” (eeww net_server) .. however with the resources of apple the BONE (be os networking environment) would have been implimented on a much quicker timeline, and would have turned Be/OSX into a solid server class OS.
and in the later days of BeOS a OpenGL app_server was already in the works. for those of you the worked with the newer version of Be’s OpenGL Kit you knew there were headed for great things.
with BeOS’s low latency which facilitated its extream user responsiveness along with its clean developer API and DOCUMENTATION!! (which btw is a joy to program in)
the BeOS already had a significant head start.
thus:
the be based OSX of today would have probably been much the same, with a personal bias to believe that could have been achieved much sooner and provided a more user friendly experience at a much cheaper price.
plus, they prob would have hired Jobs back anyways. and he would have stuffed all those pretty widgets into be/osx anyways
ps. as a full disclaimer, i will be moving to OS at work also, although BeOS has provided a feature rich experiance, and the user comunity has to some extent kept it alive for long long after it died, the time of accelerated displays, SVG rendering, java, and solid network integration (aka play well with other) is here, and the need to OSX is upon us. sad, is im probably the last paid be software developer outside of yellow tab, uh i mean Magnussoft.
woot
Mac OS X would have been radically different if Apple had bought Be Inc and not Next.
First of all, Mac OS X –IS– NextStep/OpenStep that was changed so that it could run (at the beginning) on PowerPC. So why didn’t they just have Apple customers run NextStep on their Macs? Good question but I’m guessing it had something to do with the temporary backwards compatibility of being able to run Classic.
Since Apple wouldn’t have had NextStep/OpenStep it would most likely looked almost exactly like Classic with BeOS underneath.
Having said that. Who has had the biggest influence on Apple hardware and software since they brought Steve Jobs back. Keep in mind that he is very hands on and nothing gets approved without his blessing (after several more rounds of improvement). The answer is Steve Jobs. He controlls the look of Mac OS X and Apple hardware. So everything (pretty much) is his baby.
On the other hand, what style did JLG like? OS 10 or whatever would have been very different than what we see today. And the guts of Mac OS would have been very different too.
Good to see you YNOP. Been a long time.
As for my BeOS gear, I’ll probably be selling it all before long. I have a LOT of BeOS stuff, much of which is signed.
Gotta move on sometime, and I’ve been using my PB G4 for about half a year now with no regrets. Well I regret the community has become what it is. But other than that, the OS was a joy.
I really like BeOS. And I would like to use it, but it’s too old for my hardware.
But very nice to see that it’s not forgotten yet.
Call me a glutton for punishment, or a retro-grouch, or just contrarian, but I use BeOS (or rather, Magnussoft Zeta) as my day-to-day OS for basic browsing, email, etc. It’s quick and stable and, at especially (to my eye) aesthetically pleasing.
… that a supposedly-dead OS can still generate this many comments in response to a simple article.
BeOS would’ve made a terrible server system for a number of reasons:
1) No security framework to speak of. It was somewhere between Windows 95 and NT in terms of security. You could do all sorts of insecure things in BeOS. For example, the kernel did no rights checking, at all, on data transferred through ports. Zeta has added file permissions, but security is not the sort of thing is not something you can add after the fact!
2) I/O performance really wasn’t very good. Doing something I/O intensive like untarring a file was substantially slower in BeOS than on Linux or BSD (back then even).
3) The thread scheduler optimized for interactivity, at the expense of throughput. The default time quantum was really short (something like 250 usec). There was no server-grade I/O scheduling.
4) The VM wasn’t all there. Some of the stuff you take for granted these days (unified disk/page cache) wasn’t featured.
5) There is no reason to believe it was particularly scalable. The heavy use of threads in the app_server and GUI toolkit really doesn’t imply that it’d be great on modern multi-core servers. All the multi-threaded SMP demos on BeOS were compute-bound processes. These processes spend almost no time in the kernel, so aside from second-order effects from scheduling, the underlying OS doesn’t have much of an impact on their scalability. The kernel itself probably wasn’t much more scalable than contemporary versions of Linux. It surely wasn’t in the league of Linux 2.6 or FreeBSD 6.x.
I remember at the time that the scheduler in Windows NT 4.0 could support several *times* as many executing threads than BeOS R4.5 could. Modern Linux with its O(1) scheduler can probably support an order of magnitude or two more threads on the system before choking.
1) wrong. The fs had POSIX permissions and checked for them. Processes also had uid/gid but some syscalls had comments like /* XXX: check for perms */. But the architecture was there. Zeta didn’t add anything there. Right security is not something you can add after, unless you have the placeholders. As BeOS used most POSIX abstractions it’s not that hard to actually switch on security. The most problematic issues are with native IPCs (sem/ports) which will need more checking to avoid a user to send messages to other user’s apps, while still allowing servers to get both… but that’s for my next job
2) First, for something from 1999 it’s not that bad, and 2nd BFS has to maintain attribute indices unlike ext2 and other Linux fs (but it doesn’t maintain access times). So copying files means also copying attributes, and updating indices.
3) The BeOS scheduler was not that good yes, I even tested it with some IBM latency tests I found on an article once, and was quite dissapointed. Most of the “responsiveness feeling” comes from the threaded gui design. But again it was targetting at Desktop.
4) Again at that time it was considered by Be to be sufficient for desktop.
5) Except Linux and BSD only now come to things BeOS had for a *decade*: preemptive kernel (it can interrupt itself to do something else), no “giant lock” à la BSD, tickless timers (yes, BeOS had tickless timers 10 years ago) (simply, instead of interrupting X times /s you just tell the clock to wake you up for the nearest event you want), modular kernel (and wayyy better than Linux: no insmod crap, other modules/drivers just ask and get what they want; and modules are sorted in a tree), hierarchical devfs (again, 10 years ago), file access monitoring (Linux is still fighting to get one of 3 versions working), extended attributes (something Linux only had recently to get ACLs, and they aren’t typed and indexed)…
The thread/teams limit in BeOS are arbitrary limits calculated at boot time from the available memory (the process table couldn’t grow). Again that was a shortcut to simplify the code, but was deemed a good thing by that time. And once again, O(1) is not 10 years old. You can’t simply compare something from 10 years ago to something that evolved in 10 years (from a way lower position on some issues).
Haiku is likely to remove most of those limitations btw. Many are already gone (it has working mmap() for example).
As much for multiuser which was disabled by default, there are other things that did exist in the BeOS code that were either not enabled or not compiled in by default… (like X11 wasn’t the only one to be able to go through the network…)