http://www.osnews.com/story/5472/Interview_with_Michael_Phipps_Project_Leader_of_OpenBeOS Exploring the Future of Computing en-us Copyright 2001-2009, David Adams adam+nospam@osnews.com Tue, 10 Nov 2009 05:06:07 GMT http://www.osnews.com/images/osnews.gif http://www.osnews.com http://osnews.com/thread? http://osnews.com/thread? I really hope OpenBeOS succeeds, it's great to see that they are making good progress! Sun, 21 Dec 2003 07:39:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? One of the best reviews on OSNews in recent memory. Sun, 21 Dec 2003 08:36:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? "Keep the faith"? Sounds pretty damn much like a religion of sorts.... Sun, 21 Dec 2003 08:54:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? A nice little interview. My favorite part: he guesses that in a year, it'll only be in early beta, and that's only R1. Maybe this will keep all those BeOS fans here from constantly posting about how OBOS is going to rock everyone's world any day now.

When it comes out, it will be fun to look at. But it's nice to hear some more realistic estimates of when it will be available than what the fanboys always post here. Sun, 21 Dec 2003 09:17:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? I first tried BeOS in 1999. I'd grown sick of Windows and how much it kept messing up. I tried Linux (which I revisited a year later and switched to) and it just wasn't good enough for a desktop for me. I then stumbled upon BeOS and it blew me away, so fast! So easy, exactly what i was looking for: MacOS for AMD. I never switched fully as I could never get my sound working, but I was seriously impressed nonetheless.

But now it is 2003, Linux and GNOME/KDE as well as Mac OS X have become seriously impressive, and are getting better all the time. Longhorn looks to push a few boundaries itself in 2006.

Is there really any place for BeOS in 2003? Or will it just appeal to a select few Be followers who refuse to let the past die? Sun, 21 Dec 2003 09:20:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? A nice little interview. My favorite part: he guesses that in a year, it'll only be in early beta, and that's only R1. Maybe this will keep all those BeOS fans here from constantly posting about how OBOS is going to rock everyone's world any day now.

...We've known that for a while before this interview.

The way I see it is this,
R1 is not a meager acomplishment, because once they have completed it they will gain a familiarity and intimacy with understanding the OS itself that could only have been gained via this route. They're going to know their code inside out and I believe that once R1 is released, this will result in quick progress to R2. Because by then the gruntwork would have already been done, they just need to evolve it. Sun, 21 Dec 2003 09:24:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? But now it is 2003, Linux and GNOME/KDE as well as Mac OS X have become seriously impressive, and are getting better all the time. Longhorn looks to push a few boundaries itself in 2006.

I don't get this. Is Linux a server OS or a desktop OS? It's definitely gonna need to do the server trade offs if it wants to be a desktop OS now don't it? Therefor Linux can and will never be as fast as you can be OBOS =)

OS X is a different story, still I prefer the BeWay (tm)...

A year from now, BeOS will STILL be faster than OS X, Linux, whatever... because it is THE fastest OS around.. Sun, 21 Dec 2003 09:47:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? I don't get this. Is Linux a server OS or a desktop OS? It's definitely gonna need to do the server trade offs if it wants to be a desktop OS now don't it? Therefor Linux can and will never be as fast as you can be OBOS =)

No, you DON'T get it. With Linux 2.6+ the OS is focused on desktop as well as on server tasks. You can now choose at compile time many options that enchances the desktop. Linux as of today scales REALLY well from almost calculator level (there's even a Linux port for the TI series...), to big mainframes and clusters like NASAs new 512 x Itanium 2.

OS X is a different story, still I prefer the BeWay (tm)...

Honestly I believe OS X has got MORE of the "desktop trade off for server" in kernel than Linux.


Myself I still prefer BeOS, but not because it's superior in any way. My Linux/KDE desktop is faster and more featureful in almost any way.
Still BeOS feels more right.... Sun, 21 Dec 2003 10:23:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? I don't get this. Is Linux a server OS or a desktop OS?

Neither. Linux is a kernel, not an operating system. The Linux kernel is distributed in operating systems intended for desktop and server use, however.

No, I'm not nitpicking. I'm making a point. Linux by itself - Linux the kernel - is just fine for a desktop OS. But a desktop OS is much more than its kernel. The surrounding framework is what makes it comfortable - or doesn't. The keyword is "surrounding" - *Linux* doesn't have to trade off anything to be used in either a server OS or a desktop OS. The distributors chose the focus of their Linux-based systems.

But enough thread hi-jacking, let's talk about OBOS. Sun, 21 Dec 2003 10:46:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? Thanks Koki, I hope more of the resourceful Japanese community join the OSBOS movement. Sun, 21 Dec 2003 10:51:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? the big problems with Linux involve Xfree and the bloat that runs on top. BeOS was nice and quick because it didn't have an app communicating to the desktop manager communicating to the X server (or client, whichever terminology is in vogue this week) communicating to the kernel. Less overhead. More speed.
Trading speed for network transparency is great for servers, but hardly does a desktop user any good.
I'm impressed with the progress of OBOS, and waiting in the meantime for a Zeta I can use.
Nice interview Sun, 21 Dec 2003 11:47:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? the big problems with Linux involve Xfree and the bloat that runs on top. BeOS was nice and quick because it didn't have an app communicating to the desktop manager communicating to the X server (or client, whichever terminology is in vogue this week) communicating to the kernel. Less overhead. More speed.
Trading speed for network transparency is great for servers, but hardly does a desktop user any good.


Do you have any idea of what you are talking about, or are you just chanting? By the way, the kernel plays a key factor in determining how responsive the desktop is.

bad kernel != nice desktop Sun, 21 Dec 2003 12:21:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? > Is there really any place for BeOS in 2003? Or will it
> just appeal to a select few Be followers who refuse
> to let the past die?

BeOS R5 I admit is not a truly competitive OS now. There are bugs in the media kit, networking is slow, modern hardware support is poor.

Looking through all those issues though - it is still a very elegant OS, with a more modern internal design that most "current" OSes. Windows XP still has the old Win9x C API at it's heart whereas BeOS has a very nicely thought out C++ API, it is heavily multithreaded, and remains the most responsive OS I have ever used. I still use it 90% of the time as I much prefer the user experience over that in windows.

OBOS R1 will not come out tomorrow. OBOS R1 will not revolutionise the world, or contatin lots of new features. However, it will fix many of the annoyances with R5 (the media kit bugs, poor POSIX support, networking living in userland, 1 Gig memory bug, poor support for modern hardware, etc). On top of that it will be completely open source - so there is no danger that the OS will suddenly disappear (as happened with the original BeOS).

But the thing that I really like about the OBOS project is how the whole thing is integrated. The kernel, networking, the app server, the file system, everything is developed by one project. All the components are designed to fit together as a single, cohesive OS. All of them designed with the same thing in mind (giving the best user experience possible on a desktop OS). There won't be any need for distributors to put all the pieces together, you won't be expected to compile parts on your own - you can simply go to the OBOS website and download a complete, consistent OS.

Hopefully R1 will re-enliven interest in the BeOS platform, bringing more developers and with them more apps, ready for future releases which will be much more capable of competing with current OSes. Sun, 21 Dec 2003 12:33:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? Besides OpenBeOS there are several other projects that have recreating BeOS as their goal (BeFree, BlueEyedOS, Syllable, etc.).

Just to set the record straight, Syllable is not a clone of BeOS. We are not a member of BeUnited, our API's are not compatible, we are not trying to be BeOS. While I respect what Michael & the OpenBeOS team are doing, and I wish them the best of luck, it doesn't help anybody when people look at all these projects and say "Hey! Why don't you all work together; you're all the same, right?" Sun, 21 Dec 2003 12:43:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? Since Syllable when it reaches version 1.0 shall functionally be the same or better than BeOS, why are they duplicating several efforts to clone BeOS rather than working on Syllable? Syllable is more complete, and supporting legacy BeOS applications is probably not a good reason to clone the OS. Sun, 21 Dec 2003 13:22:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? I'm hoping for SkyOS more than Syllable. Sun, 21 Dec 2003 14:04:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? "I'm hoping for SkyOS more than Syllable."
I was,but I can't bring myself to support a closed source OS.Besides, Syllable is remarkably advanced too. Sun, 21 Dec 2003 14:49:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? OBOS get a number of advantages by sticking to the R5 API:
- No time is wasted discussing new APIs.
- By and large, the API is consistent across kits, as Be Inc took 10 years designing it and improving it.
- Existing BeOS programs will still work.
- People who have done programming on BeOS before will instantly be able to code for OBOS.
- While R1 is being written, and with it a lot of insight gained wrt the issues that matter writing an OS, R2 with the API extensions can be planned.
- Keeps developers focused on writing code with the best possible implementation possible, without really having to worry about how the small part they are working in fits in the bigger picture. Sun, 21 Dec 2003 14:55:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? Maybe since ELQ hasn't posted the story, the subject of the story will not be cursed to failure. Sun, 21 Dec 2003 15:34:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? A nice interview. Its good to see that there is substantial progress being made. Hopefully, in 2004 we'll be able to get our mits on a Beta. Like everyone else the world of OS's have moved on (I'm using OSX more and more had to get a new iBook G4) but I still believe that there is a place for a BeOS style OS. I still get amazed how fast R5 feels even when running on PII 500MHz Dell. Linux distributions have yet to get a desktop version right, its not the kinux kernel but due to KDE and GNOME. They are still not as good or consistent as BeOS's GUI. There are things that I would like to see updated in the BeOS's GUI, esp since I've been using OSX more and more, but I can wait for these to be implemented.

I'm looking forward to run, OBOS on my Shuttle XPC Nforce2, 1GB of RAM.

Don't knock EQL. She was one of the most dedicated BeoS advocates around.

Harj

: o )> Sun, 21 Dec 2003 16:25:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? WE are not a BeOS clone. As much as I actually like BeOS and I have used it for quite a while, I strive for something different. We don't want to be a BeOS clone. Sun, 21 Dec 2003 17:37:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? There will be no beta in a year's time. When the project started two years ago, it was predicted there would be a downloadable iso in a year's time! See a pattern there? Sun, 21 Dec 2003 18:44:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? OBOS supports new networking technologies like zero conf.

I also hope they get Samba ported. those two things will make using OBOS much nicer on a mixed network with PCs and Macs.

also, what is their printing system? will they use Cups or is that not how they want to go? Sun, 21 Dec 2003 18:58:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? Please re-read what I wrote, I was questioning the point of cloning BeOS at all and not suggesting Syllable be a BeOS clone. Sun, 21 Dec 2003 19:03:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? > Please re-read what I wrote,

Heh. You do realize that you are posting anonymously?

---

Also, regarding OBOS, I see 2 fundamental mistakes (of which the 2nd will be their undoing):

1. They are going for this "binary compatibility" which is probably turning off many experienced devs (and thus they aren't interested in joining the project). My guess is that an experience kernel/OS developer does not want to spend their time making sure every old app on bebits runs on their new creation.

2. They chose the wrong license. Although I think the MIT license is great for apps, or something you want distributed as widely as possible, it's a poor choice for a desktop OS. Using the GPL would attract many more OS devs and allow use of already-available GPL'd OS code.

Even if OBOS were successful with the MIT, there's nothing stopping some company from coming along, downloading the code, making their own "enhancements", closing the code, patenting said enhancements, and then suing OBOS'ers at some point down the road for patent infringement.

This is particularly important for a project like OBOS which will have mass-market appeal. Corporations are cutthroat and will do anything legal (and sometimes things illegal) to make a buck. A successful OS project needs to protect itself. Sun, 21 Dec 2003 19:38:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? You mean BeOS had a fixed-position-size GUI layout engine?

That was really a bad idea. It really, really sucks from a uability POV. At least for any non-english-speaker, it means you have to redo all the dialog layouts for each translation.

Madness! Primitive! Sun, 21 Dec 2003 19:38:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? As for Binary Compatibility, I suggest you look at this (http://open-beos.sourceforge.net/nsl.php?mode=display&id=11#34)

Also, I disagree with you about the license, I believe that the license is perfect! First off, OpenBeOS will always be free, no company will really be able to compete with that, two, how can they patent something OBOS created? I mean, we don't exactly have that problem with BSD, which is nearly the exact same license (save for that little quibble with ATT a few years back, which turned out to be their fault.)

Plus, the MIT license will attract companies who want to make drivers, and also commercial application development for OBOS. Sun, 21 Dec 2003 20:41:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? They chose the wrong license. Although I think the MIT license is great for apps, or something you want distributed as widely as possible, it's a poor choice for a desktop OS. Using the GPL would attract many more OS devs and allow use of already-available GPL'd OS code.

Seeing the difference between Linux and BSD today, we can definitely say Linux has attracted media attention and a huge amount of zealotry.... I'm pretty convinced OBOS wants to be spared from this no offense. BSD's has been very successful and even today are gaining more and more developers and leading the serverplatform amoung the OSS options.

the MIT license will prove to be a very wise license in the long run. I'm still a bit scared about what will happen with Linux when media won't keep it as their pet anymore and the zealotry becomes "uncool"....

I wish OBOS the best of luck and hope to use it someday... Sun, 21 Dec 2003 20:53:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? Michael mentioned WinBE. What's that?

Is it like using Windows NT as the base but getting rid of Explorer sort of like Litestep, instead having BE like kits.

Actually I would love the OpenBeos kits to be ported to windows. Then I would know all my hardware works. Sun, 21 Dec 2003 22:52:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? If I'm not mistaken, it was a windows emulation layer for BeOS. Kind of like wine.

And what you just described is pretty impossible. Sun, 21 Dec 2003 23:20:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? "how can they patent something OBOS created"
easy, I don't think you need to prove YOU thounght it up. Just get to the patent office first. I heard MS has patents on the linux kernel. Mon, 22 Dec 2003 00:21:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? And look at how much of a rut linux is in now, man they sure have suffered! Mon, 22 Dec 2003 01:16:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? That's not the point, the thing is that they could, at anytime,cause massive problems for linux. Mon, 22 Dec 2003 02:46:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? wing wrote:
As for Binary Compatibility, I suggest you look at this (http://open-beos.sourceforge.net/nsl.php?mode=display&id=11#34)

And I hope it is just "3 easy steps". But, you know what they say: the difference between theory and practice is much bigger in practice than in theory.

First off, OpenBeOS will always be free, no company will really be able to compete with that, two, how can they patent something OBOS created?

They won't. Like I said, they can simply add some patentable technology to their fork and then patent *that*. Then, all of a sudden, everyone wants to use this new fork. Maybe later, the OBOS'ers (with their significantly dimiminished user base) add features to their original code that resembles the patented additions to the fork. This is when the sh*t starts to hit the fan.

I mean, we don't exactly have that problem with BSD, which is nearly the exact same license

That's because BSD is a server OS and doesn't have the mass-market appeal that OBOS (a *desktop* OS) will have. Just wait till the right (wrong?) folks start to see dollar signs in their eyes.

Plus, the MIT license will attract companies who want to make drivers, and also commercial application development for OBOS.

I think that's debatable.

Then Anonymous (cm-upc.chello.se) wrote:
the MIT license will prove to be a very wise license in the long run.

Why? To avoid GNU/Linux zealotry (as you'd referred to)? I'm not sure that the zealots are doing Linux any harm.

Zealotry aside, there's a reason Microsoft hates the GPL so much. If GNU/Linux was all MIT licensed, do you think it would still be around? How many lawyers does MS have working on ways to eliminate GNU/Linux as competition? Mon, 22 Dec 2003 04:24:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? And I hope it is just "3 easy steps". But, you know what they say: the difference between theory and practice is much bigger in practice than in theory.

Okay, even though it's been proven in practice (by the OBOS folks.) Have you done any research? Mon, 22 Dec 2003 06:52:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? "Actually I would love the OpenBeos kits to be ported to windows. Then I would know all my hardware works."

"And what you just described is pretty impossible."

Really ?
http://alpha.luc.ac.be/~ef00/

http://burton666.neoni.net/TextViewShot.jpg Mon, 22 Dec 2003 07:32:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? I thought he meant something like porting the underlying architecture (the servers, etc.) to Windows or something, which I don't think is possible. I could have been wrong.

Hey Jack long time no see Mon, 22 Dec 2003 08:26:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? "..Linux and GNOME/KDE as well as Mac OS X have become seriously impressive, and are getting better all the time. Longhorn looks to push a few boundaries itself in 2006.
Is there really any place for BeOS in 2003?"
Well let me tell it like this:
As long you really nead MM performance and you can't afford SGI Machines ;-), there _will_ be a reason to push BeOS/Zeta or both.
I tried to work around with different apps, different Hardware on different platforms (x86 and PPC and Gx/OS X).
If you really looking for a performant MM OS, there is simply nothing better - unless you pay _way_ more.
- Capturing [DV & TV]
- Ripping [DVD & mp3/OGG]
- Editing Video & Audio
- Internet Radio Stations ( http://www.tunetrackersystems.com/ )
I think if the Audience just take a brief look that it is a shame how new OS'ses just bloat their OS to urge the user to buy steadily new more performant Hardware, wether they can afford it or not.
Even Apple is nearly at the end of the road in HW Ways. The G5 is a terminal attempt to get some %-ages in the big PC market back (hope they will).
But all of this is not _really_ needed: Keep your OS thin and clean and remind that MM Apps=low latencies in the OS.
If you sum it up, there is only one new OS which fit all these rules: BeOS/OBOS/Zeta.
That simple.
I think, the Devs have to review their Platform choice. It would be like a snowball effect: if some key Applications and Hardware Support is initialized (Zeta pushes that in a fantastic way), it's up to the Devs which fight on and on on Wintendo Lunix or Applet, try to patch this or that or wait for even more performant hardware to be distributed.
There is no need for all that.
If you want to do MM Stuff, you have to revise all your opinions once you saw BeOS/Zeta...
We need more experient Coders and non-coward companies which step into a solution in which their software and or driver makes the most sense!
... Mon, 22 Dec 2003 08:40:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? You mean BeOS had a fixed-position-size GUI layout engine?


Yes, just like MacOS X. I can't understand how Apple could do that to the "most advanced OS". Mon, 22 Dec 2003 08:59:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? Yes, just like MacOS X. I can't understand how Apple could do that to the "most advanced OS".

I don't know about OSX but there's third party layout engines for BeOS, and I assume that there are for OSX as well. Though, I agree, it should be needed. Mon, 22 Dec 2003 11:14:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? Enjoyed reading this interview. Thanks for posting it.

I definitely plan to use OBOS as it becomes available.

BTW if you have to "hijack" this OBOS thread to tell me how great some other OS is . . . Mon, 22 Dec 2003 16:04:00 GMT donotreply@osnews.com (Anonymous) Comments http://osnews.com/thread? http://osnews.com/thread? I enjoyed this interview; thanks for posting it.

About Michael's guess as to when a beta of R1 will be available... hey, what's wrong with a little cautionary positive thinking? Sure it could be delayed much beyond that guessed date, but that's why he called it a guess and why he said things change with the wind in open source development.

Positive thoughts! :-) Mon, 22 Dec 2003 22:16:00 GMT donotreply@osnews.com (Anonymous) Comments