'Be' is nice, end of story
abortretry.fail
abortretry.fail
I miss BeOS, TBH. I ran a heavily customized BeOS R5.05 install that had upgraded bits from OpenBeOS (later, HaikuOS), ZetaOS etc.
I ran it as a daily driver on a Dell GX1 SFF that I swear was designed for BeOS: Max RAM the BeOS kernel supported without hackery was 768MB - The GX maxed out at 768MB; The GX1 had a Rage GPU that BeOS natively supported (Including 3D Acceleration!) which was a very short list; the GX1 had a supported NIC, IDE controllers, etc etc.
I had upgraded the CPU to a 1.4GHz Tualatin-S that, sadly, had to run at 1.04GHz as the GX1 wouldn't run a FSB higher than 100mhz so my 133mhz FSB Tualatin had to bop down a little. Still screamed.
sigh I miss that system. I would still be running it if lightning hadn't fried it.
It sat next to my NeXTstation Turbo Color that I used for way longer than was reasonable as my primary system, lol. The NeXT was a daily driver till 2008.
33mhz 040 w/ 128MB 60NS EDO. OPENSTEP4.2 + lots of ported goodies friends and I did. Even managed to get SSLv3 working and load newer websites! :D
... wish I hadn't sold it.
Oh yeah! I also had a Pre-FCC certification dual 66mhz BeBox. Cool machine. Slow as shit though. The 603e in it wasn't natively SMP capable IIRC. The GLU chip could either run one CPU with L2 cache or dual CPUs with L2 cache disabled. Coherency was done in software. Unless I am horribly misremembering. It was a nasty hack and showed.
My BeBox! https://www.flickr.com/photos/helfer/albums/7215762193739249...
My GX-1 and my NeXT :D https://www.flickr.com/photos/helfer/6676203105/in/photolist...
Thanks for the info :)
Is there any memoir, book, whatever detailing out decisions on the hardware? Like why did they go with the 603? etc?
I really wish there was some deep-dive book on the early history. I can't (well, couldnt. Havent looked in a while) find much info on the original hobbit based system.
Cale Lewis/Dlazlo - if you read this or if anyone knows him please chime in below ;-) He was like the best guy on those forums. Miss that guy.
Dude.. just found the "for ebay" album on your Flickr profile... omg :(
(Oooh also SGI Onyx2.. nice. I search the term "sgi" on Craigslist every day. All I have is a pair of O2's)
Are you personally acquainted? Or do people routinely run 3rd party intelligence tools to connect online profiles?
I used a NEC Versa P/75 (pentium 75 (no mmx), 40mb ram, 350mb HDD, 4GB CF card, modded ORiNOCO Gold cardbus WiFi card, heavily modded Windows 95b (not 95c, mind you. No no. Has to be 95b for reasons) for years. Dual battery packs I upgraded. Ram for 10+ hours and worked fine for MP3s, IRC, SSH, etc. :). I still have it. I didn’t stop using it on the regular till probably 2015-2016.
I have a Portege 3110ct that I upgraded to 192mb ram, 64gb ide ssd, usb2.0 pcmcia card, usb WiFi, etc. that I still use off and on. It was my primary mobile system for about 8 years. Runs a stripped down heavily customized Debian based Linux.
If you are mostly doing textual things, you can happily exist on anything made in the last 30 years assuming it has a basic tcp/ip stack and preferably SSHv2 support.
I recently dug my 386sx40 I grew up on out of storage. It desperately needs a ram upgrade and an FPU. Right now it has a 512k ISA trident video card, NE2000 ISA NIC, 250mb boot drive, 160mb storage drive, 12mb ram.
Runs windows 3.11 with CalmiraXP, Win32s, TCP/IP stack, lots of mods to the base OS (kernel tweaks for the cooperative multitasking functions etc).
… I feel like an old man ranting about the good old days. I’m only 35. I swear. Lol
I had a Tecra 750CDM that was incredible - I won it at work… I was like a $5000 laptop at the time and included a detachable webcam.
I'm still plugging away with BeOS, have a few Macs running the PowerPC version still.
Edit: also, you're photos made me quite happy. Beautiful machines.
The IRIS3130 went to a museum, at least!
As for the SSL… I’ll dig through my stuff and see what I have. On it. I have a screenshot of gmail loaded up… lemme find that.
https://www.flickr.com/photos/helfer/3947037084/in/photolist...
At AT&T (the mothership) we inherited a bunch of NeXT machines from the cellular company they invested in.
Me, I had next to my engineering books multiple BeOS boxed OS copies.
Carried them with me to <large U.S. airplane company> where I installed them on every machine possible. Great times. Running a NeXT pizza box one one side and BeOS on a Dell Latitude on the other, and showing off the multitasking made some of my co-workers weep. (Turbo Color on the NeXT side with the frickin' sound box and monitor).
One of the <research scientists/future tech> co-workers was absolutely into Objective-C. Those were fantastic time showing what /could/ be done.
Edit: until 2016.
https://www.flickr.com/photos/helfer/albums/7215762370071298...
It's little things like errors automatically prompting you to open a graphical debugger or processes being grouped by application. There's sensible UX that doesn't expect me to be a wizard to understand what's going on or how to dig deeper.
Obviously it has rough edges (as, well, honestly all operating systems do), but the things that do work work really well.
I think after Windows 7 the processes are grouped per application in Task Manager. We’re all grumpy about the redesigns of things we are familiar with, but little UX improvements happen all the time.
[1]: https://processhacker.sourceforge.io/
You're right; I was mostly thinking about problems with Linux since thats my daily driver. It would have been better to say that Haiku gives developers the same UX affordances for interacting with their system as non-technical users.
When using Linux I'm frustrated by an overall lack of UX, but when using Windows/Mac it's developer specific UX that's ignored.
Activity Monitor in macOS can do that too.
I think that shows what the target audience is. I think most people wouldn’t know what to do with that prompt, other than immediately discard it.
That's a fair complaint, and I'm definitely biased since I'm technically oriented. That said, discarding it is totally fine as long as people who want it at least get the option; and if even showing it is too confusing to users, it's a lot better to have a "developer" toggle in settings that enables all this stuff, rather than not having it at all.
Haiku's certainly not crazy accessible right now, but it has some ideas that I think other OSs should take note of.
SerenityOS is also doing a lot to perfect their UX. It might become at least as popular as Haiku down the line.
> It's little things like errors automatically prompting you to open a graphical debugger
If this is the concept of "mainstream" and sensible "user experience" it sounds like Haiku is a complete bunch of horseshit. (And this feature is easily available in Windows anyway)
my business card
https://platinumball.net/be/business_card.jpg
third floor break room. i saw jlg sitting on that couch, holding forth on the company, once or twice.
https://platinumball.net/be/break_room.jpg
a clipboard made out of a bebox motherboard. yep, i still have it
https://platinumball.net/be/clipboard1.jpg
https://platinumball.net/be/clipboard2.jpg
edit: ugh, some people are reporting my links are forbidden. i removed my .htaccess file, which had some rules forbidding hot-linking. maybe that will help?
> Additionally, a 403 Forbidden error was encountered while trying to use an ErrorDocument to handle the request.
one thing, though: i am pretty sure that the haiku error messages were only in netpositive, the web browser, and not in any other beos apps.
Many modern systems do have file attributes, although they're not always as capable as the ones in BeFS, and none of them to my knowledge chose to add all these BTrees because keeping them up-to-date is expensive.
What is the big difference is that BeFS exposed the extended attributes easily in the file manager and APIs - compare it with other systems where the APIs were little known or not used - POSIX extended attributes which are very limited (4k), Solaris extended attributes (which let you open a file as a directory) but are Solaris only, WinNT which had the issue of WinNT being a niche among windows systems, etc. etc.
BeOS in comparison required at least considerable porting work, at worst writing from scratch, and that meant the new features were much more visible.
As it soon turned out, almost everybody. All that BeFS does, in exchange for a bunch of your disk space and CPU, is BTree indexes, for each index you want your filesystem gets a little slower and bigger. It starts out "Quite a bit worse than some of the alternatives available at the time" and goes down from there. If you switch the indexing off so as to get back reasonable performance, now your email folders are just unsorted and sorting them takes forever as it would on a Windows PC (except your Windows PC has an email client so you wouldn't do that).
This makes for a cool demo of course, which is important if your actual goal is to get Apple to hire you instead of Jobs, but there's a reason why your filesystem did not add this fairly simple feature and you still have an email client.
Is it that BeFS was indexing everything and not just the folders that needed to be indexed?
macos spotlight is also routinely indexing the file system, doesn't seem so different to me, naively
Also, an actual email client can specialise. BeFS is very general, the attributes can be typed but there's only one string type (other types include 32-bit and 64-bit integers, floats) so a Subject and a Sender email address both get indexed the same way. An email client might reasonably optimise for @corp.example type address searches for example, but then also do some simple keyword optimisations for subject line too.
You could not turn off indexing for some/ all folders in BeOS only for an entire filesystem (and if you do a bunch of stuff doesn't work)
Before BFS, was OFS which was an actual database and was horribly buggy, requiring the BTree to be regenerated often enough that it was a "BIOS" level boot menu option on the BeBox running that version's Boot ROM.
A lot of the wonder at BFS stems from the marketing of OFS. BFS only retained the bare minimum features that OFS had, to make the database stuff still exist.
Attributes are exposed on the UI level and can be a useful way to organize custom structured data even for an end user: https://www.haiku-os.org/docs/userguide/en/workshop-filetype...
I'm sure it is way more limited than some original failed vision but limiting scope doesn't have to be a bad thing and I think what they ended up with does live up to the hype in some ways. That said, it certainly also has some flaws like not even being able to efficiently use the index for the built-in file find dialog because of case insensitivity.
BFS was written in a way that complied to a file system API that could then support other non-BFS file systems being plugged in. But because it was written in more of a traditional file system way, the database features got massively cut down to just attributes and indexes with live queries. It was a trade off. The database hype built by the DR releases (mainly for BeBox, I think only one made it publicly to Mac) was maintained because the attributes sort of mimicked the bare minimum functionality needed to tick a marketing box. But BFS really doesn't do much more than some other file systems have since done.
The OFS was more like the SQLServer based WinFS that Longhorn was going to have.
Yes it is a trade off. Not sure what other file systems you are referring to but I'm not aware of any others that enable the kind of applications BeOS had using these features. Those are also a key part of it.
That's true but you still did have some control over the indexing as you could choose what attributes to index or not in the first place.
It’s true that indexes can impact performance but “quite a bit worse” is far from what I observed at the time. Even with hundreds of thousands of files it was quite responsive - Apple needed SSDs and 15 more years of development to get close, and Windows still hasn’t.
Everyone who uses more than one device to access his/her e-mail? Sure BeOS had nice features, but it seems to me that it solved the wrong problems.
IMHO, Apple made the right decision with Next because of Jobs. I’d sure like to see the results of the alternate timeline!
I still feel like Be could be great for phones/mobile, maybe VR. (Just as it was so foundationally low latency)
(Note I am aware of Haiku, etc. I’m using “Be” to mean something from its linage)
Side note - Jean-Louis Gassée, founder(?) of Be is very insightful. He writes some excellent short form tech world analysis at http://mondaynote.com/
I'm much happier w/ Quartz (née Display PDF) than the typical image model division which BeOS had.
That said, I really wish someone would get it (and Tron) running on a Raspberry Pi.
They did eventually fix this, but it was too little too late.
For example, in the settings you could choose to turn each CPU in the machine on and off. But it didn't stop you from turning all of them off. And if you turned the last one off, the system halted, just as you'd expect. No warnings, nothing that stopped you. If you thought turning off all the CPUs was the right thing to do, then by golly you go ahead and do it!
I'm not sure I follow why this is a good thing. Having the option is good, but some sort of a sanity check and a warning dialog just seems like looking out for your users.
For example, I recently used Linux on the desktop as my daily driver and wrote about my experiences [1]. At one point, I considered that the system was perhaps a bit of a mess in regards to the installed packages and, not wanting to do a clean reinstall, I figured that I'd just reinstall Python in particular.
There's nothing mystifying about running an apt remove command, nor is there anything outrageous about the idea of removing some software and then later reinstalling it. However, as it turns out, this would probably break certain parts of the system, so I got a helpful warning instead:
After this operation, 1 244 MB disk space will be freed.
You are about to do something potentially harmful.
To continue type in the phrase 'Yes, do as I say!'
By that time I had already entered "y" and pressed enter, not bothering to read the message, since the rest of the prompt looked like it normally would, so instead I got this output: Abort.
And nothing broke. I think systems should look out more for the users like that, which would give the OS less of a bad reputation.Of course, it won't always save you, like Linus from LTT still ignored such a warning and ruined his install somewhat [2].
[1] https://blog.kronis.dev/articles/a-week-of-linux-instead-of-...
Might be conjecture based on small sample size I have for Debian, but what Linus was doing shouldn't really fuck anything up... Steam package in Debian "just installed"
In my experience, DEB distros have somehow been more stable than RPM ones, but Debian and Ubuntu wouldn't be an order of magnitude off from one another (aside from Ubuntu giving you more leeway in regards to proprietary software out of the box).
Well, maybe apart from the whole "snap" debacle, which is more of an architectural issue in my eyes, going around the package manager (and even then doing it in a way that's a bit more counter-culture than what AppImage or Flatpak does; even if I can understand their desire for automatic updates).
Then again, I have no issue using Docker containers and actually liked older Ubuntu LTS Unity desktop environment (which was stable in my experience, contrary to what others experienced), so maybe I'm a bit of an oddball here.
Linux these days is quite a bit easier, but there are still edge cases that can cause problems.
Same dipshits who ran ZFS on top of unRAID. And unRAID is already an abomination.
grumbles
Obligatory demo from the 90s https://www.youtube.com/watch?v=BsVydyC8ZGQ
Source: dual BeBox/133 on my desk
What are your sound problems? I only briefly used BeOS on an actual mac and it, amazingly enough, worked without a hitch.
I never used BeOS much though, as there wasn't much software available for it. This was in ~2000.
Someone should make a programming language where those are the keywords for true and false.
They had their niche
People used Cinelerra and Cinepaint. Now you have Krita and Blender.
I've tried Ubuntu Studio. It's not better except that it had early support for preempt rt. I still think Linux audio is a disaster.
Pro Tools isn't the only DAW game in town, in fact its not even the most commonly used any more and its also not the 'best' DAW for many jobs - merely the most entrenched in peoples minds as a "pro tool".
So many producers are giving it up for cheaper and more effective tools such as REAPER, which not only runs circles around Pro Tools but also runs really, really great on Linux. As does Ardour, which is easily in the same class of any of the other DAW's in terms of capabilities. See also, Waveform, Zrythmn, Qtractor and MusE.
As does BitWig Studio - an Ableton Live analog, arguably more powerful and feature-rich than Ableton Live, which makes sense as it is being built by ex-Ableton folks.
FL Studio? LMMS! Renoise!
Linux Audio really rocks.
Not to mention: ZynthianOS! Oh my! Monome NORNS! Holy heck!
>one major DAW
Sorry, no, this is wrong. REAPER is a major DAW, runs natively. Same with BitWig, Ardour and Harrison Mixbus. So thats at least 4. [1]
>very few plugins
I challenge you to actually install Ubuntu Studio and repeat this claim. Plus, VST plugins run great under Ubuntu Studio. You should probably try it again and update your understanding of the scene - you might've had a point 10 years ago, but as of today: just no. Linux audio is amazing.
>I still think Linux audio is a disaster.
I think you don't have enough modern experience to make this claim. My Presonus-based Ubuntu Studio DAW is the lowest latency, easiest to use and set up DAW in the studio. While the other producers are cussing through the hoops of eLicense and dongle management and unsigned/outdated kext's for super-expensive boutique/cocktail hardware, the Linux DAW users in my studio are happily whistling through the edits with their USB and Firewire-based audio i/o devices, many of which work quite happily 'out of the box', "apt update && apt upgrade" dilemma notwithstanding ..
[1] - If these DAW's can do it, there is technically no reason the other DAW's can't - its just the fixed idea that "linux audio sucks" that prevents this from happening in the industry.
Linux audio does not, in reality, suck - it actually ROCKS, hard. Linux audio only sucks if you say it does.
There are certainly no valid technical reasons for a lack of ports from the market-mindset vendors .. just the mindset you proffer, which is, imho, plain wrong.
Sadly, now I use it to drive a Pioneer XDJ console. And they made me pay extra for DVS support without the Serato box :( But it's a pretty fly 4-deck system!
I very much hope one day to see another integrated Be-inspired system emerge, in the vein of 00s era Apple (and the original BeBox). Narrow product line, tight hardware-software integration; striking, thoughtful and fun UI; notably different paradigm and features from the competition such that it scratches an itch the others are ignoring.
This bit of history gives me so much hope (tinged with sadness that it's just a daydream). If I knew more about OS or filesystem construction, or how to sell to VCs, I'd push the idea hard myself.
That wasn't leaked until after Be stopped operating.
5.0, which was almost usable was only out for about 8 months before Be stopped operating.
4.5 and prior?
What does the dog say? "Ruff"
Yeah. 4.5 and prior could spin some teapots but not much else. By the time "good" BeOS came out all computers could spin teapots and Windows PCs had 3d hardware acceleration. BeOS barely had 2d hardware acceleration.
In the year 2000, after five years of development, BeOS didn't have multi-user support, a stable network stack, or USB mass storage support.
I think that people are remembering with rose-colored glasses about 5.0 (which if I remember right was "oops leaked lol" for free on a magazine disc) or 5.1 with all of the patches which started circulating after all of the laid-off Be employees started dumping files online.
Those of us who struggled for months to get BeOS stable and useful on our Performa 6400/200s have more bittersweet memories.
On PowerPC is was a lot more hobbled. You sound like you were on PowerPC unfortunately.
4.5 and 5.0 were so similar that it really isn't worth talking about the differences - a few minor API differences. The main thing with 5.0 was that Tracker was opened, so they moved the About box in to the Kernel to prevent people using a fake one and preserve their copyright notices etc.. The API was really similar and mostly anything that compiles under R4.5 will run under R5. They also added in a bunch of 3rd party codec and changed some of the icons.
In fact, Stinger (BeIA) was based on the 4.5 codebase, not the R5 codebase. This is apparent when you try to use the API and notice anything newer than 4.5 is not present.
E: Posted this before having time to read the article, glad to see it was mentioned!
> Why did they choose the name Haiku? Error messages from many applications in BeOS are written in haikus. Additionally, they felt that the art of haiku was representative of the elegance and simplicity of BeOS.
It seems Haiku is BeOS, just renamed.
> an open source project was started whose aim was to recreate BeOS from scratch with full binary and source compatibility. This was OpenBeOS. The first release in 2002 was a community update to BeOS 5.0.3 with some open source replacements for Be code. The project name changed in 2004. With everything surrounding the demise of Be being highly litigious, it is no surprise that the project wished to avoid legal complications over their name. They chose Haiku.
In the next decade I see even Microsoft moving way more towards Linux.
Still unclear on whether they’d ditch the windows kernel for Windows, and adapt the UI for it.
If you want a near-instant-boot appliance with Linux, try Librelec. I think it's made for Kodi but you can use it for other things. Lakka (Retroarch) uses it and gets me to a GUI, ready to interact with, rendering under DRM (so, no X or Wayland) in like 3 seconds flat on a Raspberry Pi. Or, FreeDOS isn't a crazy option for appliance-type machines, depending on what you want to run.
1. This is dated 2009 so I may have used my NeXT more like through 2010. Honestly a bit fuzzy on the time frame
2. Here’s a screenshot of when I got Omniweb to load gmail after a lot of fiddling: https://www.flickr.com/photos/helfer/3947037084/in/photolist...
He’s been at some of the most interesting companies in the industry at pivotal moments and crucial roles. There must be a ton of stories.
His Monday Note blog is great.
Dominic Giampaolo did APFS for Apple as well.
My understanding was that the Macintosh Portable was an awful product, because it was just too big and the screen wasn't backlit. I never saw one myself, though.
But you’d have been better off lugging around an SE/30.
it's amusing to see the name of the blog, a popular column in pc magazine in the '90s bore the same name. i forget who wrote it, but it was one of the really well known computer journalists of the time, if i recall correctly.
a notable absence is mention of the cpu utilization lights. to show off all that work they did to build a truly multiprocessor system, they had two strips of leds that ran up the front bezel of the system, indicating realtime utilization of the cpus.
i remember playing with beos many years ago. i think they may have shipped an intel version after shutting down the hardware.
The initial CPU target for BeOS was the AT&T H0BB1T DSP. The alpha copies were called Shark. I actually have system images for the hobbit builds still floating around somewhere. I never did get my hands on a prototype hobbit bebox, sadly. I almost bought one off a guy in France but the deal fell apart.
Be produced a regular PowerPC version that could be installed on Macs (I think it may have required PCI macs?) and then an x86 version that was 32bit only and had a lot of ... quirks. But I used it for over a decade.
I think the truth is that Apple, and Apple + IBM worked on operating systems in this period and had trouble making a good one. Why it is hard for a mature company with legacy systems to create a new operating system is a difficult thing to explain. But it is not unusual. And worth understanding for people who want to innovate in large organizations.
My theory is that you can't really _plan_ how to create and spread an OS: You need to put a dedicated enough person with enough time to write an interesting prototype that would brings you a subsequent advantage. And of course, you need the peoples that will judge wisely about the long run future of this prototype (remember Linux being considered bad?). After this, the prototype have to be mature enough for developers to and hardware makers to support it.
That's a lot of _if_, and I suspect most OS (and in some way, softwares and language) are spontaneous creations that answer a need and become _good_ as a consequence of being used.
It is the reason IBM had to set up a secret ninja tiger team to make the pc. They tried a few times before, and ended up with really nice systems, far nicer than the pc. but they cost 10,000 dollars, in 1980, you are not gong to conquer the small business market for computers with a 10,000 dollar machine. so they set up a secret team, far away from the rest of the ibm mothership apparatus with the mission to make the cheapest computer possible, and they did, it cost 3,000 dollars had nothing compatible with any other ibm and sold like hotcakes.
In general, the hardest problem for such companies is backwards compatibility.
In the case of Apple, it was doubly so: they tried to build a pre-emptive multitasking operating system with memory protection that was backwards compatible (both for applications and system extensions) with an existing OS that had cooperative multitasking, did not have memory protection, and had OS support for patching system calls (a feature that allowed third parties to develop of such things as Switcher (https://www.folklore.org/StoryView.py?project=Macintosh&stor...), After Dark, Adobe Font Manager, and RAM Doubler).
I wonder what did Apple do that BeOS didn't to succeed. And didn't Microsoft basically "save" Apple at some point?
Microsoft invested some money in Apple to help alleviate anti-trust issues, but from my reading the more substantive assistance was a commitment to ship Office for Mac.
https://medium.com/the-techlife/why-did-microsoft-invest-150...
“Later testimony in the United States v. Microsoft Corp. case revealed that, at the time, Apple was threatening Microsoft with a multibillion-dollar lawsuit over the allegedly stolen code, and in return Bill Gates was threatening the cancellation of Microsoft Office for Mac. In August 1997, Apple and Microsoft announced a settlement deal. Apple would drop all current lawsuits, including all lingering issues from the "Look & Feel" lawsuit and the "QuickTime source code" lawsuit, and agree to make Internet Explorer for Mac the default browser on the Macintosh unless the user explicitly chose the bundled Netscape browser. In return, Microsoft agreed to continue developing Office, Internet Explorer, and various developer tools and software for the Mac for the next 5 years, and purchase $150 million of non-voting Apple stock. The companies also agreed to mutual collaboration on Java technologies, and to cross-license all existing patents, and patents obtained during the five-year deal, with one another”
I worked on an internet appliance using BeOS around 2000. The device worked, but was unfortunately slow due to the compromises Sony had to make to hit their price point (most notably running the display in portrait mode without hardware rotation, meaning every paint had to rotate in software!) and the lack of software really contributed - the NetPositive browser was fast and had some nice ideas but you’d run into more sites which would have worked in Netscape or Internet Explorer and didn’t have Flash.
I really liked BeOS: BFS’ database-like capabilities are still unmatched and it was years ahead of the curve on stability (I remember crashing the graphics driver & watching it restart, the open apps repainted, and I kept working) and consistent performance (Windows and macOS still have more UI lag under load despite much faster hardware). The fatal flaws, however, were the lack of permissions and immature networking. It was also harder to port Unix software, which mattered a lot at the time since that meant a ton of technical workers could use a Mac with OS X but not a BeOS system. I’m not sure what the assessment was for running Classic Mac apps either but suspect that already hard task would have been harder with a more limited OS. Even at the time I thought Apple made the right call, much as it meant I couldn’t unload some Be shares I’d picked up at a profit.
We only had one Apple official reseller store in Lisbon for the whole country, and outside a couple of university labs and companies with headquarters on the capital, they were only seen on computer magazines.
Apple already had an ecosystem from the Classic Mac OS. It had lots of apps, lots of loyal users, etc. It had issues, but Apple wasn't starting from scratch.
BeOS was simply new. That means trying to get everyone to support a new platform. If you weren't a Mac user back in the day, you might not know how hard it was - getting printers to work, having to buy Mac-specific accessories, etc. The Mac would have millions of users while BeOS would have none. Getting third-parties to support the platform would be incredibly difficult - it was so hard on the Mac already. This was way before a lot of work on platform independence. Also, if you'd run Linux 15-20 years ago on a laptop, you probably remember the issues around stuff like getting WiFi to work. BeOS was trying to launch a new ecosystem in an era before the web really took over and when native apps and drivers were important and not cross-platform.
Maybe think about a different question: why did Windows Phone fail when the iPhone and Android didn't? Well, it was a third platform and an also-ran with almost no marketshare. Developers didn't want to support a third platform when users saw no need for it. Users didn't want to buy into a third platform without apps.
Still, one of the major things for Be was that Microsoft forbid OEMs from offering BeOS on their machines - if they wanted to be able to get Windows OEM licenses. So Be was locked out of every PC company. If you were Dell, you might tell Be, "we'd love to offer customers the option of getting a Dell with BeOS, but we can't because we need Microsoft's Windows licenses." So if you bought a PC, it'd come with Windows and then you'd need to buy BeOS and install that. Now you had a great OS with nothing to run on it and you'd go back to Windows. Be did make BeOS free at one point, but it was too late (and even at free it lacked a reason to be run).
The Mac had built up marketshare, a userbase, application support, drivers, etc. for a decade when BeOS launched. BeOS launched at a time when Microsoft had sewn up the PC market and was in its full-monopoly, anti-competition stance. They used their power to crush all rivals on the PC.
Apple, even though they had some limited clones, was never really a rival. Apple was a hardware/software combo play and small enough to be ignored as a declining niche player. Did Microsoft "save" Apple? Probably not. They did invest some money and committed to make Microsoft Office available on Mac which certainly helped (remember, this is before the age when things could read .doc and .xls files and Microsoft was still in anti-competition mode where they'd try and break competitors products).
Still, it seems more like Microsoft wanted to keep one player around to stave off anti-trust concerns - and Apple was the safe one to keep around. If they kept Be around, BeOS would need to get some meaningful marketshare (let's say 5%). At that point, what's to stop more people from running BeOS? Nothing - their Pc could just run BeOS. If they kept Apple around with 5% marketshare, what's to stop more people from running MacOS? The fact that they'd need to buy new hardware from Apple at a premium price point with inferior capabilities. That's safe. People can't switch. And Intel was killing PowerPC processors at the time and PCs were available for $500 while Apple finally introduced a budget machine at $1,300. Here's a $500 eMachines with 300MHz processor or an iMac for $1,300 and 233MHz processor - and things got worse from there as the years went on. The point being, Apple was never a threat to Windows because MacOS was tied to separate hardware. BeOS was a huge threat to Windows if Be got any traction. They needed to be killed.
Apple's real resurgence wasn't via Mac OS or even Mac OS X. It was via the iPod. They just started making so much money. Money means you can invest in things. Then the iPhone meant even more money. Then their computers were still expensive, but had a much lower premium than they used to - especially if you account for things like display quality.
Maybe Be could have launched a BePod, but they didn't. Maybe Be could have parlayed that success into a BePhone. Maybe Be could have used all that to become a player on the desktop. But they built the BeIA internet appliance which was just a computer, but worse. Apple removed the floppy drive and Be removed both the floppy and CD drive - at a time when computers were becoming music hubs.
BeOS launched in October 1995. Apple bought NeXT in 1996 and launched the iMac in 1998 and iPod in 2001. BeOS was such a flash in the pan. Blink and you'd miss its 5 year existence. BeOS R5 (the final released version) came out in March 2000, less than 5 years after the initial launch. In a lot of ways, they never really got any traction - but they did get a lot of love from the people who used it. BeOS was magical in many ways (even if somewhat unfinished in others). At a time when the Mac OS had cooperative multitasking and would just freeze up, at a time when Mac OS X was unusably slow, at a time when Windows was a jumble of crap that never quite worked right, BeOS was this clean, modern, happy system. In many ways, it had Mac-like polish and coordination to it combined with such a modern developer experience and so many great ideas.
But it's hard to get over the hump necessary to create the momentum that brings developers and users. Apple already had those users and developers (even if it was a small community compared to Windows). Then Apple had some amazing hits with the iPod and iPhone.
If SGI or BeBox had been the first to market with such a sexy laptop, we'd still have them around.
It was the lack of focus on laptops which killed Be and SGI (and others), but it was the focus on laptops as personal computing devices which saved Apple.
I'd kill for an alternative-universe SGI or Be laptop. Imagine the tiBook with Irix or BeOS on it .. delicious ..
It is pretty amazing to see the thing boot .. the blinkenlights .. the hum and whir of the disks .. straight to the desktop just like the day I got it. And then open the dozen videos that stream seamlessly, which is all I do really.
What a world it would have been if JLG had focused on the laptop as the driver of his revolution, as SJ did ..
I was introduced to it when the company came out to do a demo of it for the computer science department of the uni I was at.
The multimedia abilities they demoed were amazing. Coming from Atari ST and TT this seemed like the next generation. I loved the BeFS.
I could not afford a BeBox at the time and the uni. decided against buying any.
We did have quite a few NeXTstations in the department which were nice. but I would have preferred Be.
Its easily one of the sexiest machines in my retro-computing collection. Sure, the SGI boxes actually have some useful software - but the BeBox has sexier blinkenlights by far. ;)
That’s awesome. Does anyone have some examples of these haikus?
Today, imagine that you are a young Windows programmer and that I'm a venture capitalist and you come and see me and say, “Mr. Gasse, do I have a deal for you.” “Yes?” “I have the word processor for Windows that will kill Microsoft Word.” What am I to do if I'm a caring venture capitalist? I have to open the drawer and instead of pulling out the checkbook I should pull out the Magnum .357 and give you the coup de grace because this will stop what otherwise would be a long, ugly, expensive agony for your family.
I’m sorry but it’s Copland, after the composer Aaron Copland.
Having read altogether too many editions of Macworld in the post Steve, pre Steve era, that error (repeated later in the text) leapt off the page.
(Admittedly, this would be kind of an hilarious name for a failed OS.)