To this very day, this BeOS demonstration video from Be, Inc. blows my mind. I’m not entirely sure about the date of the video, but since we’re looking at Pentium IIs and the Intel version of the BeOS, I’m guessing we’re in 1998. This means that while Windows users were barely getting by with Windows 98, and Mac users did not look at their Macs funny because otherwise Mac OS House Of Cards Edition would come crumbling down, the BeOS was doing the awesome stuff shown in the video.
Taking chronology into account, the BeOS was and is the best operating system ever made. I’m far from impartial on this one, of course, but there has never been a piece of software that generated that same sense of wonder, excitement, and exhilaration that the BeOS did. It sure wasn’t perfect, but it had so much personality, such a strong identity, and even a sense of humour – the stuff we have today just can’t compare. iOS and OS X are clearly designed to lock you into buying as much Apple hardware as possible. Android and Chrome OS are designed to keep you staring at Google ads for as long as possible. Windows is Windows. Linux is the same mess it’s always been (I’m sorry).
None of them are about putting the consumer and technology at the centre.
When the first Haiku alpha was released, I explained how with the demise of Be, something in this industry died with it. I once had the faint, faint hope that the mobile revolution would reignite that spark of insanity, but with Apple and Google dividing and conquering this industry almost overnight, and with ARM devices being an ungodly mess of restrictions and proprietary crap, we’re farther away from those glory days than we’ve ever been, with fewer and fewer signs of them ever returning.
The BeOS is the best example of why our industry is so utterly broken. BeOS exemplifies that the best does not win. And if the best does not win, we are being held back.
I just fired up my trusty Toshiba 3110ct from, I guess, 2000. It has 300Mhz and 64MB Ram. I threw in a 16GB SSD and it’s perfect for Haiku. I really love this tiny machine for it’s build quality, battery-runtime, small footprint and overal cuteness. Everything runs smoothly. Of course browsing the web is not a good experience but I gave it some swapspace and now it’s at least working.
If only they could produce the gluecode for some USB/PCMCIA Wifi cards so I could free it from the portreplicator and the cable.
https://www.haiku-os.org/guides/daily-tasks/wireless
Edited 2015-04-07 12:24 UTC
Grats! You just beat my low end Haiku system!
667 Mhz Coppermine P3, 256 MB RAM
Though mine has a BIG handicap, 20GB ATA HD with about 8 GB for Haiku-nightly, so less space and slower speed!
Aren’t you worried using the SSD as swap might shorten it’s useful life? Why not max the RAM (even 256 will help, trust me) 🙂 and run with no swap? Not even sure if Haiku has some commands like Linux “swapoff”.
I dug out my old (original) black Be Inc, polo shirt a few days ago and decided to wear it all day, yes it still fits me 😉 I also have the white T-shirt, somewhere.
A wonder of an OS, especially BFS was a marvel and way ahead of it’s time for a desktop OS filesystem. I loved the idea of each email being a separate file and you used the file system to search etc… The file system was a database and not just a place to store files.
I built a Dual Pentium II system around a Tyan motherboard just to run it. A the same time started contributing to Setiathome.
Sound like MH mail from ~1980. I used to use exmh as a gui around it. http://www.nongnu.org/nmh.
It was way faster than mbox where finding a message in a “folder” was a linear search from the start of the file.
Most live in a world where Windows and Android (or OSX and iOS) wins.
But I would really like to live in a world where webOS and beOS win.
As a stopgap fix I live in a world where I can use Linux and Sailfish OS and get away with it.
Well, I guess everything is better than a world where WindowsXP and WindowsCE win….
…a world where one of
– Interlisp-D
– Smalltalk
– Ceres
– Lilith
– Oberon
– Lisp Machines
would have won.
What I like in Mac OS X, Windows, iOS, Android is they are all going their own ways in OS design across all layers, instead of plain copying UNIX.
You must have missed the memo, but OS-X has UNIX certification from the Open Group as of Sept 18, 2013. That means it *is* UNIX for all intents and legal purposes.
IIRC, this certification has to be given for each release and either Leopard or Snow Leopard was the last version to receive it.
edit: just saw you mentioned 2013, that’s news to me.
Edited 2015-04-07 14:08 UTC
Leopard was the first version to receive it (When run on Intel processors). All Intel versions except Lion have received certification.
Mac OS X is compatible with UNIX sure, but it offers much more.
Where in UNIX is hybrid kernel, launchd, CUPS, Gatekeeper, Core* Frameworks, IO Kit, *Kit frameworks, C++ driver model, …?
UNIX certification means an existing set of kernel APIs and command line utilities.
It doesn’t say anything about those systems that while offering POSIX compatibility, have decided to take the architecture beyond UNIX base model.
Nor is it meant to. The UNIX spec means that, if you target the spec, your software will be portable across others that meet the spec. It doesn’t mandate that your system doesn’t offer anything beyond the spec.
In your list, there is a bit of a mix-up of machines, operating systems,, and programming languages.
One cannot really mention Lisp Machines without reference to its cousin Symbolics (and their hardware running the Genera operating system).
The Lilith was a hardware platform with its operating system written in Modula. I can`t remember with certainty the name of the operating system though.
The Ceres was the initial hardware platform for the Oberon operating system and language. Interestingly, the Lilith was used for its cross-development until Oberon was self-hosting.
I concur that it would be great that any one of these operating system (and programming languages) had a greater commercial success and/or end-user prevalence today. However, a hardware platform and its operating system has essentially been carried by a “killer” application be it Artificial Intelligence research (Interlisp-D, Lisp Machines, Symbolics), the spreadsheet (VisiCalc for the 6502 and AppleDos and Lotus-123 for the 8088 and DOS), and desktop-publishing (the 68k series and MacOS).
I was aware of it, just writing out of my mind a list of alternative computing models not constrained by green terminals and PDP-11 hardware.
Actually I used some of those systems.
Ah, Oberon. I really liked that language. Object oriented by a simple extension to the Pascal var records. I would have loved to have seen it gain more traction.
I’m not a programming and know nothing about the Oberon language. But the Oberon operating system with it’s “text user interface”… I’ve never used an OS that was less intuitive or user-friendly.
There’s an old P3 computer upstairs that still has an Oberon partition.
I had played with Oberon System 3 (text user interface) along with its extension Gadgets (graphical user interface) and did not find it more awkward to use than Windows 3.0/3.1 which was becoming the standard at my work place at that time.
However, after spending 8-9 hours a day at work in Windows, it was such a mental whiplash to use Oberon during evenings and weekends that I eventually gave up trying.
The greatest difference between the two was how the mouse was treated as input device. In Windows, one was essentially mostly left clicking. In Oberon, one could click one, two, or even the three buttons to access a variety of commands.
*** Linux is the same mess it’s always been (I’m sorry).***
True, Linux is a mess. But it is a working and well supported mess. From what is available today, I’d say it holds the most promise, even if it probably will never be as sexy as BeOS.
Back in the day I got excited by BeOS. I was seriously contemplating leaving Windows 98 behind for it. It seemed to have momentum. It reminded me of what computing could be (started out on Amiga Workbench 3.5). Then the first cracks appeared in the form of the focus shift on BeIA. Soon after it disappeared.
Haiku is the perpetual promise just beyond the horizon. Maybe one day it’ll become more.
Untill then I’ll be using the mess called Linux. Reluctantly it is being remolded for the 21st century. It might even become somewhat good at multi-media….
Everybody who calls Linux a mess does not really understand how open source works. Sure, Red Hat by now might have the money to build a totally cool and easy to use desktop OS and applications, but they sell servers .. so they don’t care.
If there is no billion dollar company that does the majority of the coding, you will get what people like to work on. So you will get a lot of programmer stuff and geek stuff. If you have some kind of love relation with boring MS Office shit then you are out of luck.
But there are so many things in Linux that are way more advanced than anything in OS X or Windows. Filesystems, containers, scalability etc. Nobody can compete with Linux when it comes to things like flexibility, customization etc.
I have to disagree with you in that sentence. I use Slackware as my desktop machine, in my netbook and in my MediaCenter machine. And I call Linux a mess. Slackware is not a mess. But a lot of distribution forgot the basic structures, commands and break a lot of compatibility that any BSD Unix still preserve.
The worst thing I found was that Arch Linux remove the old Net-Tools utilities out of the standard distribution in favor of the IPRoute2 tools. When the second can be used at the same time (some tools in IPRoute2 are not in the Net-Tools).
But sure, Linux is incredibly flexible and allowed you to do whatever you want. And I like that.
Edited 2015-04-07 16:02 UTC
I agree – I am also a Slackware user (of at least 10 years) although I see my mid-to-future desktop being Haiku with Vulkan graphics and FreeBSD for server stuff.
Nothing like a blanket statement…
Or they know all about it and just don’t like it.
***Everybody who calls Linux a mess does not really understand how open source works.***
Wrong conclusion. I for one know exactly how it works and to me it explains why some aspects of Linux is a hot mess. Even if I wouldn’t want to use another system.
For all the wonderful things that FOSS has produced atop Linux, it has also developed some serious warts here and there.
We have a myriad distro’s, but only a fraction of those do really offer something more than just a different theme, wallpaper and a minor tweak in OOTB software. I know very well this is because people can. And with things currently standing as they are, people even should, but it isn’t pretty, nor efficient.
We also have a myriad different repositories, with too many packaging formats. The funny thing is that those repositories carry the same software, over and over. This is because a few projects were scratching their own itch with their own packaging systems, but yet again it isn’t pretty, nor efficient.
It would have been a lot prettier and more efficient had libraries been designed in a way that makes their supported functions exportable and queryable. Then it would be possible to just ask a library if it can provide a certain function and hook in to it when provided. The way we do it now is clunky and pretty fixed to 1 specimen of a dynamically linked file.
The same with making a “different” distro, when at most it is a superficial respin of the parent distro. Which _is_ a seperate OS, even if it brings nothing new to the table. If Linux had had the luxury of a ubiqitous and flexible installer framework, which could work with a “recipe for default selections”, we could have had a few base distributions and have people write their own “installer recipes” to customize it to what they want it to be out of the box. (Awesome Steve’s Mageia recipe, Ultra-minimal Arch recipe (although Arch currently killed the installer altogether), etc.)
Same with package formats. There isn’t some unreconcilable difference between package systems. They all basically do the same. Copy the binary files to the respective directories and add an entry to the package database, so it can all be updated and removed when necessary. If there would have been a universal package format (and a sane, non file based library system) we could have had one repository where application projects could upload their newest versions and all distributions would have had access to the same packages, without artificial compatibility hurdles.
No, this is not one of those “If Linux was just like Windows” things either. With the right amount of flexibility designed in, Linux could be as diverse as always, but it wouldn’t be predicated on historical (and IMO artificial) development differences, locked up in an arbitrarily seperated distro system. Or do you really believe that LibreOffice in Debian totally is a different Office suite from the one in Fedora? Or that OpenSUSE libc is so not the libc in Mageia? (Psst. All that software contained in distro’s is 99% the same!)
I remember hearing back in the 90s that Apple was considering buying Be, this was before they bought Next. I thought that it would have been perfect, I was running Be on a PowerPC something (I can’t remember) back then and loved it. But supposedly Be’s asking price was way too high so Apple passed. Just think what could have been if Apple had bought Be. Of course, this may have been a rumor, I can’t really remember the details from 20 years ago.
Honestly? No Apple. Without Jobs, Apple was screwed, and even though BeOS was a great OS (I used it, loved it, coded for it and have a lot of affection towards it in general), it wasn’t anywhere near complete enough to save Apple. Printing was pretty much non existent, the network stack was extremely unreliable and the programming model was very difficult to port legacy code to. It probably would have taken longer than Apple had left to complete the OS, and without Job the wheat wouldn’t have been separated from the chaff, so the business model would still have been bleeding money. Clones would have destroyed sales of the Mac even further, and who knows if the iPod, or iAnything, would have been released.
Can’t really agree there. Amiga came first and was designed largely as a multimedia platform before anyone really understood what that was in terms of computing. There were groups working on full screen full motion demos for Amiga before IBM clones or Apple could manage the feat. The sound hardware was outstanding. It was light years ahead of IBM PCs & clones and even Apple at the time. But likewise as BeOS, it’s promise never really took off as CBM died and Amiga and it’s promise as it once existed is now long dead.
It’s a pity that BeOS never really went anywhere. I tried it myself and was impressed with its capabilities on hardware that it supported. But not a lot was actually supported. Catch 22. That’s really the problem with totally new ideas. When they are too new and different and there’s an 800lb gorilla in the room with it (Microsoft in this case) it’s not so likely to succeed.
Yup AmigaOS/Workbench was doing this long before 1998.
There are three points here:
i) Which of AmigaOS or BeOS was the best operating system ever made. The answer will always depend on which criteria are felt to be the most important. And whatever the answer is, it too frequently generates flaming comments.
ii) AmigaOS achieved what it did, and when it did, with the assistance of its three custom designed chips. As such, it suffered from the same issues of limited hardware support as BeOS. Circumstances also led BeOS to be somewhat portable in its implementation (forced switching of hardware platform – from Hobbit to PPC and finally to X86).
iii) BeOS, being designed later than AmigaOS, benefited from the lessons learned and its approach to multi-tasking was likely by opposition to that of AmigaOS. Similarly, one of the most successful RISC based chip (the ARM) was reportedly designed by opposition to the most complex one (the NS32K).
Ah the good old days of the last exciting operating system. OS expectations were a lot lower back then, and BeOS still blew them out of the water.
I miss those days, and all the folks I’ve lost track of since then. Hope you’re all doing something cool that makes you happy!
If I remember right, the only half way decent Office suite was Gobe. As a college student at the time, I used BeOs for everything I could, but writing papers just wasn’t possible. I wrote some basic software in c++ ( not too many libraries available) , web surfed, watched tv,listened to music and that was about all I could actually do on it. Most things I needed to do for school could only be done in windows. So despite the frequent crashes, it won.
As soon as I heard the opening guitar riff, my mind teleported me back to my doorm room in 2000…
Gosh I miss that system. It was just so fast….
And I’ve never managed to find a suitable solution for live queries or file system attributes.
I hate iTunes, Mail.app, etc. I want these things to use the filesystem! Gah!
BeOS was a pretty impressive tech demo, but it was so incomplete it was bordering on unusable. Even Haiku has been languishing in development for years and still has a long way to go. NeXT was far more mature, and also quite advanced. Was it as tech demo pretty as Be? No. But it was the better choice – and it also brought Steve Jobs back. If they had went with Be, and brought Jean-Louise Gassee back, Apple wouldn’t have made it to the new millennium. Watch the NeXT demo videos from 1991 and remember what year it is and what you were using at the time. NeXT was the best operating system no one used.
While that might be true, the HFS filesystem that was used in the successor to NeXT (OSX) is to this day a horrible mess and can not even do wild querying even after the upgrade to HFS+.
HFS begat HFS+, begat HFS+ Journaled, etc
HFS was a filesystem designed for a single process to access the filesystem, it really has only seen small advances since the 80’s. If you read up on how it works, it’s pretty horrible.
OpenStep used a variant of the classic UNIX filesystem, so I guess it would have worked for OS X too…. but the issue was that every Mac they were upgrading to OS X already had HFS+ file systems, and the UNIX Filesystem in the OpenStep was going to break a lot of stuff when they tried to run a copy of Classic MacOS in the Bluebox on top of it. I’m guessing that’s why Rhapsody jettisoned the old FS rapidly? It was simpler to use HFS+ and write a new driver than it was to make all the resource fork malarkey work on a non native file system.
In 35 years, I have not found another desktop that made me as productive as the Workplace Shell on OS/2. It also could run 16 videos at full-frame rate at the same time, but strangely, I never found a use for that.
My problem with BeOS is that while it looked pretty, I couldn’t *do* anything with it, because there was no software that allowed me to be productive– that, it didn’t run on commonly available (ie, Intel) hardware.
WPS might not have been as pretty as BeOS, or have all the shiny that AmigaOS had, but it was an actual “object oriented” desktop that was far, far more advanced than anything since.
Just “work areas” would be a nice idea to bring back. Throw an editor, a debugger, an API reference, and some bookmarks into a folder. Check the “work area” box. Now when you open that folder, all the apps/documents (“objects”) return to their former state. You’re done with that? Close the folder. All the related programs shut down / go away.
You want to add a program to files of type “.zip”? Drag your newly installed unzipper shortcut onto the “open with” tab of the “.zip” file object (assuming it didn’t register itself– or you could browse, the old fashioned way if you wanted).
You want to change colors on an object, like a dialog window? Drag the new color onto the dialog. Hold down the alt key if you want it to affect all similar elements (another option key would make it global). Ditto with changing fonts– drag the font onto the dialog– yes, hideous things could happen.
Create a file. Create a shortcut to that file. Now move (and rename) the original file. Under Windows, linux/unix, and OSX, that shortcut will STILL break (unless you did a hard link). OS/2 didn’t care, because the shortcut was just an attribute extending the original file.
All of this, a 32-bit core, 16 bit compatibility, and a windows runtime a la Wine, in 1992.
By 1996, with OS/2 Warp, I could use voice commands as well as dictation– the dictation was a bit clunky, and I realized I don’t need to control my computer with voice commands. Still though, I had the option. They even re-wrote their web browser to work well with voice commands (Primarily with the “links” command, which displayed a list of hyperlinks on the page– today, it probably wouldn’t work well).
IBM made some huge mistakes with OS/2– but very few of them were related to the technical side of the OS.
Unfortunately, the computer industry doesn’t cater to the “power users”– They’d rather cater to the lowest common denominator, and no one likes change, so all of the “current” operating systems are based on 40+ year old technology (Unix or VMS).
It’s a bit ironic that the same site that bemoans the death of BeOS also ran the article “Power Users Need to Shut Up”– when “Power Users” were the only people who cared about BeOS or OS/2 in the first place.
Yeah, people like to gush over how wonderful BeOS is/was, but at the end of the day, an OS is only as good as the applications you can run on it.
Voted you up, but it’s a major catch-22. Most OS projects, even those with technical merit, can never catch up because of this. Devs can spend all their time & energy trying to make the OS great. But without apps, they loose. I think one of the main reasons Linux became successful (especially in server space) is because it was highly compatible with commercial Unix offerings. If, instead of making a clone, Linus tried developing something different (and possibly better than Unix), Linux probably would have failed.
I still think there is more innovation to be done in the OS space (ie beyond windows & *nux), but realistically there’s very little hope for an indy OS to succeed with no 3rd party support. At the same time, it’s hard to convince a 3rd party to support an OS with negligible market share.
I really wanted to use BeOS as my main OS. Gobe Productive looked awesome but the trial version wouldn’t run on my system so I had to revert to Windows whenever I needed a word processor. It was a fantastic OS, though.
There are many more “average” users than “power” users and the general trend has been to “hide” complexity of the operating system under a pretty graphical user interface.
Many of the things you mentioned being feasible in the Workplace Shell, I wish I could do in Windows. There are others things which would greatly help my work, appear so intuitive to be possible to do within an object oriented framework, yet can`t be done.
Unfortunately, my work day is in Windows and I am getting old and can`t tolerate as much as I used to the mental whiplash of going from one operating system at work to another one at home and vice versa.
But, you could say the same about any OS. OpenStep did a lot more of this kind of thing, as did RISC OS. BeOS had things that were far more useful to me, a user, than those that you describe. BFS, live queries – this, this is what I miss and what anyone has yet to duplicate.
The problem is the most perfect standalone invention won’t win over a mediocre, or even bad standard one.
If I could make the perfect car, but it was too wide to drive on the existing roads, or otherwise wasn’t “street legal”, no one would buy it.
BeOS was not open, so didn’t get the GNU/BSD mindshare – nor would there be any ports of the then nacent applications or even the ability to use the GPL/BSD software ecosystem.
BeOS didn’t run on all hardware. This was a partial problem with Linux, but people would code the drivers they could, and it was ported everywhere.
It rejected the bazaar. It wasn’t even a Cathedral. It was a tiny monastery with a scriptorium that could produce something like “The Book of Kells”.
I never quite understood why Be Inc. choose to not open source BeOS after it went bankrupt. It was sold to PALM for peanuts (former Be Inc. got more money from their settlement with MS!) and never went somewhere.
JLG have explained several times that he would have liked to opensource BeOS, bit the problem was that some of the tightly integrated software was licensed and it would have cost too much for the allready money bleeding company in man hours to get the system working without the licensed parts.
It was just not economically viable. They did however open source some of the parts that made up the UI like the Tracker and Deskbar. They also made great strides to document the filesystem so a clean house implementation like the one done by the openbeos group could be achieved so fast.
I understand every geek’s internal desire for BeOS to have succeeded but maybe if their management back then where not so blind folded by the Apple deal they might have still had a chance with it.
The x86 port came too late, in a time when Microsoft was powerful enough to step in as a monopolist entity.
Not only the Apple deal. Be Inc. corporate culture was very similar to Apple itself, no surprise here since it was founded by a ex-Apple executive.
They always wanted to sell hardware, and only saw BeOS as a part of their product instead of a product itself for the most part of their history.
It’s called lock-in into proprietary formats (APIs and file formats).
Quality electric devices like Miele wouldn’t exist if GE had proprietary control over electric plugs, and they kept “updating” the design when the patents were about to expire.
And do you really thing we ‘d have more automakers than Chrysler and GM if they had a patent on the chemical composition of their respective hypothetical proprietary fuels?
Too put it in plain English, the lock-in that comes from the undocumented API features of Windows and OS X and proprietary file formats is the biggest anti-competitive move in the history of mankind. Governments didn’t notice, or didn’t want to notice.
PS: And just wait ’till Google locks in your purchased content (apps, movies, music and e-books) in The Cloud and doesn’t provide a public API to all the features.
Edited 2015-04-07 19:02 UTC
I find ironic you fret and lambaste crappy proprietary hardware while at the same time praise and sing the virtues of Crappy proprietary software like BeOS.
We have open source, and that’s great. But we haven’t done anything in the business end to match the nobel intent of open source. Maybe it’s time to look into scalable democratic businesses?
I’ve been reading a lot about Worker Directed Enterprises, in particular looking at companies like Mondragon in Spain. Perhaps alternative OS need that kind of an Enterprise to match with (or supersede) open source software movements?
I think like democratic government, open source doesn’t pair will with authoritarian, profit exclusive, enterprises. We need something better – then maybe alternatives like Haiku OS or Firefox OS would have a shot.
The business model could start out simply enough – supplying the product in existing categories for other democratic enterprises (for example, a Linux distro or Haiku-OS based hw/sw platform), then move on to real invention. Global business generate the demand they meet through advertising and other measures – this is especially true for technology companies (if you are trying to build something people are already have defined and want in a technology business, you are doing it wrong). The “law” of supply and demand is just a myth. A democratic enterprise can do this as well or even better than exploitive authoritarian enterprises.
So go launch one and beat the “exploitative authoritarian enterprises” in the market. It’s not as easy as it sounds, but it does happen when an idea is truly better and is commercialized by a savvy businessperson. Make it happen.
That’s the idea!
Guys, stop reinventing the wheel. 😉
Of course companies directed by workers would be much more democratic, efficient and less exploitative than any other form of business. Corruption is the gear of Capitalism.
We don’t need some half-baked theories like open-source and its laughable bazaar theory. We have a cooperative model which has been here for almost 300 hundreds years : it is called Socialism.
Socialism is the opposite of freedom and democracy (unless you mean democracy as in the dictatorship of the majority).
Socialism is coercion and initiation of force or the threat thereof.
We need less government and socialism, all countries today have one form or another of socialism ranging all the way from warlord socialism in Somalia, fascism in Burma and Zimbabwe, cronyism or corporatism in the EU, Canada and the US, communism in China and Vietnam, Social democracy in Venezuela and so on. It is all socialism in the end.
Socialism=Collectivism
Capitalism=Individualism
“Adam Smith – Wealth of nations”
Your quick but detailed explanation made me realize I was totally wrong in understanding Socialism. The depth of your thinking is particularly accurate when you say that, in the end, all that you mentioned is Socialism.
I regret that I had no chances to learn such things way earlier in my life. Thank you for opening my eyes, at last.
</s>
It’s been all my fault, I agree. 😉
IMHO one of the best laptop hardware is the 2015 Chromebook Pixel LS.
IMHO one of the best operating systems is Haiku.
Would like to see these two merged…
[on topic]
Used to run the 1.44MB floppy disk with BeOS and loved it.
Sadly the Be engineer that crafted that floppy image died, even before Be closed shop.
Imagine JLG would have sold Be.Inc to Apple (instead of beeing to greedy) and NeXT would still be around doing their own thing… the computing world would be a better place.
[/on topic]
I am puzzled as the 1.44MB floppy disk was a boot floppy which was necessary for running R5 Personal Edition installed on a Windows 2000 system.
Maybe you are referring to the QNX Demo floppy disk which was also circulating around at about the same time&
You are right, it was QNX, sorry for the misinformation.
Well, it was over 15 years ago…
BeOS came on a CD.
BeOS PE was great; that was my first experience with that operating system. Designed as a trial version to run only within Windows, it probably took most users about five minutes to figure out they could run the installer and have BeOS on it’s own partition.
BeOS was undoubtedly a masterpiece. Yet one should remember that starting from scratch is actually EASIER, not more difficult. So the premise Gassée did at the beginning of the video (we decided to start from scratch) made the task easier.
The legacies of existing code and the fact that you must support an existing eco-system make it harder to innovate, rather than easier. Those being able to start from a blank sheet will always benefit from lessons learned by others.
To draw a comparison, when Apple released the first iPhone innovation seemed to be running 24h per day in Cupertino. Now that Apple must support an ecosystem made of hundreds of millions of devices, many complain that Apple is not innovating anymore.
Having said that, BeOS was a masterpiece, an impressive work of talented engineers.
I don’t think it’s a sign of a broken industry. Lets face it: BeOS lacked applications, and lacked market share. It needs one to get the other. It is a classic chicken-and-egg problem, and it exists in many industries.
Look at, say, electric cars. Nobody wants an electric car, because there is the whole issue of finding places to charge it. Or course, without there being many electric cars on the road, nobody wants to invest in a charging infrastructure.
Or, APS film – an easy-load film canister that is capable of storing metadata about the pictures you took. It was easier to use than regular 35mm film canisters, but nobody made cameras for it because people weren’t buying it yet, and nobody was buying it yet because nobody was making better cameras that used it. Also, it came out barely before the digital revolution began.
BeOS had the same chicken-and-egg problem, along with the issue of timing: Before the future of BeOS and Be Inc started looking questionable, Win2K had been released, with modern OpenGL and DirectX allowing great 3D and media, great multi-threading, and a maturity that BeOS decidedly lacked.
Again, not a sign of a broken industry, but just not being good enough to justify the associated costs.
For those of you were not old enough to be around then …
There were several companies that wanted to sell computers with BeOS on them, even as a hidden OS that you could only stumble on if you used the secret handshake (you know how to make that partition on the hard drive the one to boot from), it was shot down by one simple thing.
At the time that BeOS came out Microsoft dictated what was in the contracts all the computer makers _had_ to sign in order to get Windows for even one computer. The clause in the contract was that you could not install any OS other than Windows. It was Windows only for every computer or no Windows at all. Fail to live up to that and Microsoft pulled the plug on OEM Windows licenses for that company and that company would die.
Toshiba tried to go against MS and they were quickly threatened with losing all the marketing money that Microsoft ponied up for them to make advertising. And for those of you who weren’t in that business field, you have no idea how much of the PR budget for most computer makers came from MS, and it was usually a large portion. That alone, with millions of dollars in the balance or maybe even tens of millions of dollars, kept OEMs in line.
So if you are a computer maker and you want to have BeOS run on your computer you can easily do this. But _HOW_ do you let anyone know? Marketing? No. Market anything but Windows and you lose the MS money with threats of OEM licenses drying up fast.
So that leaves it to Be to market BeOS and they tried and found out another way that MS killed them. Software reviewers that reviewed BeOS found quite a few people going on their site and posting negative comments about them. These people worked at Microsoft either through them being what people would consider a “normal” employee. Or they were people who’s only “job” for Microsoft was to be a shill by, in large numbers, spreading Fear, Uncertainty, and Dread. Yes, FUD. And I’m not making this up.
Who am I? I was a customer of BeOS software and a friend of a person who write for “a” PC magazine who got “shouted down” by MS shills when he reviewed BeOS for the magazine that he worked for. And that was nothing compared to the publisher and Executive Editor of the magazine got for letting a BeOS article make it into their magazine if it didn’t drone on about how there were no BeOS applications that had file compatibility with MS software.
Want to edit .doc files? Tough luck. The same with Excel and so forth. Oh, you could edit them but most of the features that people used weren’t supported or worse, it couldn’t read anything but the most basic of versions of Word docs. And then saving them? Well you saved them in a file format called call Rich Text Format (rtf) or for Excel it was “comma delimited” and not xls.
BeOS didn’t succeed for the lack of BeOS trying. It was shouted down and companies that either made computer or wrote articles or published them were threatened by the goons of Microsoft until the ceased and desisted from talking about anything other than MS software in a positive light.
Oh, and that adage about “build it and they will come”? Meaning build a better product and people will flock to it. Ask Novel and BeOS and other companies that made a better product only to be out flanked by Microsoft who went to CEOs and Board of Directors and sweet talked or threatened to use Microsoft software or else.
You scoff or you say that this didn’t happen. You are wrong.