Linked by Eugenia Loli on Thu 7th Oct 2004 05:42 UTC
Zeta In this screenshot you can see an updated /lib folder serving BeOS. In the folder, every library present, with the exception of libbe, libroot (and libGL which it comes from the MESA project) are Haiku libs. Hopefully, soon Haiku will be able to replace and extend on the original BeOS. On related news, IsComputerOn is now the distributors of YellowTAB's Zeta in North America.
Order by: Score:
...
by Thom Holwerda on Thu 7th Oct 2004 06:20 UTC

Congratulations to the Haiku project! Let's hope we'll see a release soon ;) .

v BeOS > Linux
by Zachary on Thu 7th Oct 2004 06:30 UTC
v RE: BeOS > Linux
by Eugenia on Thu 7th Oct 2004 06:33 UTC
BeOS != Linux
by Zachary on Thu 7th Oct 2004 06:44 UTC

I would sorta like to see this project outshine Linux on the Open Source desktop side of things. Seems to be a mildly superior architecture and I've always loved the responsiveness of Be. A desktop like this would potentially be competion for Windows.

--This comment watered down to please the masses.

Good for yT
by mario on Thu 7th Oct 2004 06:47 UTC

I am glad yT has picked ICO as a distributor for Zeta, ICO is a household name and emphasizes the continuum between BeOS and Zeta.

Who knows, maybe BeGroovy could join in the fun in some way. In any case, good luck to yT and hope they seed the market with many copies of Zeta.

@Zachary
by Lumbergh on Thu 7th Oct 2004 07:09 UTC

Seems to be a mildly superior architecture and I've always loved the responsiveness of Be. A desktop like this would potentially be competion for Windows.

I'm curious to what you find about BeOS "architecturely superior"? Is it the NewOS kernel, the heavily threaded model (which Eugenia and her hubby(who worked at Be) have both stated is a bitch to program for)?

Personally I'm pulling for BlueEyedOS, but it doesn't seem to have much of a following.

Re: Lumbergh
by Zachary on Thu 7th Oct 2004 07:22 UTC

Though I haven't programmed the API's and honestly haven't investigated them much, I've always heard good things about them (perhaps mistakenly). What I was referring to, however, was the design of the overall system. Things like storing each contact as a file and then being able to search for contact meta data seems unique (at least in full scale implementation). Similarly so with mail items. Having all of that integrated is really quite nice. Also, I like how it has well-defined global services (translation, audio, video, etc). We have some of that in Windows, but it seems better implemented in BeOS (I think i have 5 different video players at the moment in Windows, none of which can play all my videos). So, my architecture comment referred to the user experience and not necessarily the developer experience (though I believe there to be a correlation between the two).

I hope Haiku succeeds
by Sean on Thu 7th Oct 2004 07:26 UTC

There's no question that BeOS was technologically superior pretty much anything Apple or Microsoft could come up with at the time. Even today, BeOS mops the floor with Linux when it comes to usability. Haiku will make BeOS live forever and be competitive again. Staying away from the tempation to use the Linux kernel is also a good choice. The people need a real desktop OS, not some ginormous behemoth thrown together from spare parts.

Yellow tabs future...
by kaiwai on Thu 7th Oct 2004 07:32 UTC

With Haiku-os delivery around the corner, the question is whether there is actually a reason to purchase Zeta if there is a free implementation available? if it were *me* I would would suggest that Zeta would be better off, for example, working with Haiku-os to get Haiku complete and to make money off selling ported applications.

One easy one, for example, would be to get Wine working without the need of using X11, it would be just a matter of implementing a BeOS based display driver and sorting ouut some other issues, once that is done, they could license the technology developed by Codeweavers, make a natice BeOS front end, and resell it.

Another one would be to work with Troll tech to get qt natively available for BeOS along with GTK on BeOS as well; of course, with a nice BeOS native them so that these toolkits integrate nicely.

ehrm.
by robert renling on Thu 7th Oct 2004 07:46 UTC

is that that the openbeos appserver?

Re: ehrm.
by bsdrocks on Thu 7th Oct 2004 08:09 UTC

No, check the date. As far from what I understand that anything that has 2004 is Haiku and 2000 is orignal from BeOS.

RE:Ehrm
by Bryce on Thu 7th Oct 2004 08:14 UTC

Yes the haiku appserver compiles and links and does various diffrent things. draws lines even. however it is not yet in a usable condition and as far as I know it is being tested inside beos and not used as the actual appserver, as of yet.

Re: Lumbergh
by Mat on Thu 7th Oct 2004 08:28 UTC

I am a BeOS user and a hobby programmer. The way I see it, both Linux and BeOS OS architectures have their pros and cons. In my opinion, BeOS is much more oriented towards purely desktop use. There is focus on providing the user with responsive, high-performance GUI applications, so the architecture is quite streamlined and simplified, that may also be the root of some of its cons. Linux is more pragmatic and universal in that respect. The moral of the story: you do not want BeOS in a server room.

The "heavy multi-threading" mentioned above is was called pervasive multi-threading by Be Inc. The trick here is that many BeOS API C++ classes try to promote responsiveness of your apps by implicitly spawning separate threads, most notably the essential BWindow. To make things even more insidious, there are few restrictions on what your thread can do to others running in the same "team" - so there is plenty of room to shoot yourself in the foot, I guess! But the problems are *no* greater as with any other multi-threading capable OS. As usual, you have to deal with issues like race conditions, locking and such. It isn't any harder, really. Fortunately, the BeOS API is one of the most elegant and starightforward application frameworks you can come across. It is easy to pick up even for those just starting with BeOS programming, C++ or even the whole programming business. Plus, there is simply fewer things to know. I mean, the BeOS Book pretty much contains *everything* there is to know on BeOS programming, yet if you print it out, you will notice it is much thinner than most specialised Windows literature on a *particular* topic!

With Haiku-os delivery around the corner..
by Max on Thu 7th Oct 2004 08:36 UTC

The most difficult and central pieces of the project - the app server and the kernel have been barely started. AFAIK they don't even have real kernel hackers - not a single one! Haiku is still years away from it's first useable release. However if one of the Haiku devs reads this and disagrees - please post! You don't believe that Haiku is just "around the corner" right?

BeOS programming
by Mat on Thu 7th Oct 2004 08:37 UTC

@Zachary: Take a quick look at Programming BeOS. This book is freely available in PDF format at http://www.oreilly.com/catalog/beosprog/book/ and is an excellent introduction to BeOS programming.

v RE: With Haiku-os delivery around the corner..
by Jonathan on Thu 7th Oct 2004 09:40 UTC
Given up on Zeta
by AndrewG on Thu 7th Oct 2004 10:08 UTC

I bought RC1 almost immediately, paid for the upgrade to RC3, but I wonder if they will ever get V1 out the door. I mean I remember reading about 2 years ago that the first release was imminent. Also think about, RC1 is over a year old. How can you go a full year from a Release Candidate and still not have it finished. Clearly RC1 was more of an Alpha.

...
by Timerever on Thu 7th Oct 2004 11:37 UTC

Just for curiosity...

Ain't BeOS dead?

I mean... why not put effort on a new OS that surpasses BeOS instead of trying to recreate it?

RE: Yellow tabs future...
by Inglorion on Thu 7th Oct 2004 11:38 UTC

I think you have misjudged the difficulty of porting WINE off X11. AFAIK, WINE is precisely an implementation of the win32 APIs on top of X11. This means that porting it off X11 would be about the same work as writing it anew, and we know how long that takes...

On the other hand, many applications have had BeOS ports at some point, and these would work on Zeta and Haiku just as well.

Re: ...
by Anonymous on Thu 7th Oct 2004 12:13 UTC

@Timerever: You've got to start somewhere. Haiku will be an BeOS compatible system, built from scratch as far the coding goes. This sounds like a flaime bait, but BeOS is already a pretty modern OS, in many ways more so than GNU/Linux or Windows XP (flame on!). The only reason you can call it dead is lack of support - you know apps, drivers, codecs and like. Haiku is an attempt to make a modern implementation of BeOS which already "surpasses" BeOS in many areas and brings it up to date. Besides, Haiku R1 will be an excellent fundation to build future versions on, while still being "old" BeOS compatible, so we'll have something to play with while waiting. I for one am looking forward to it.

RE: Timerever
by Thom Holwerda on Thu 7th Oct 2004 12:37 UTC

No, BeOS is not dead.

I always have the strong conviction that an OS only dies when no one uses it anymore. This is far from the case with BeOS. It's still used by many (including me, it was my main OS untill I bought an iMac a few months ago, now it's my 2nd OS), and even new users are coming every day (I got lots of people to use and try it out too).

BeOS will simply never truly die, because there's no other OS out there that posesses the same strong points as BeOS has: speed, responsiveness, clean design, ease of use and easy to maintain. And, the only OS where I could sanely manage my windows (UI behaviour in BeOS makes a million times more sense than in other OS's).

Re:Inglorion
by zele on Thu 7th Oct 2004 12:40 UTC

I think you have misjudged the difficulty of porting WINE off X11. AFAIK, WINE is precisely an implementation of the win32 APIs on top of X11.

The wine libs run on reactOS which does not have X11. I remember reading an interview with a wine developer who said that WINE is very abstracted and portable. The greatest obstacle for wine on BeOS is the lack of mmap. Not a problem for Haiku.

I'm awaiting the normal eugenia post....
by $crew j00 on Thu 7th Oct 2004 13:53 UTC

I'm just waiting for the "It'll take them another seven years, my husband says it's impossible" bullcrap that normally gets spouted here.

C'mon Eugenia, I know you're just dying to slam Haiku. Oh wait, what's that? You're seeing too much progress from the Haiku people and starting to get a taste of humble pie?

Keep up the great work you Haiku devs out there.

screenshot
by johnMG on Thu 7th Oct 2004 14:09 UTC

And, the only OS where I could sanely manage my windows (UI behaviour in BeOS makes a million times more sense than in other OS's).

Man, every time I see a screenshot of BeOS I always remember how darn cool it is/was. Love that BeOS GUI -- Be just nailed it.

v RE: I'm awaiting the normal eugenia post....
by Anonymous on Thu 7th Oct 2004 14:09 UTC
Good job!
by Ronald on Thu 7th Oct 2004 14:15 UTC

This is great. If they finish Haiku, people won't have to go to all that trouble learning Linux for an open source alternative Desktop.

Looking forward to a release
by TaterSalad on Thu 7th Oct 2004 14:15 UTC

I can't wait til they have a release of Haiku. I know its probably still a few years off, but I still look forward to it. I just hope it has great hardware support. I tried BeOS PE 5 on two different PC's, it wouldn't run on one, and it would run on the other but with poor graphics and no networking support.

Bring it on!
by Galley on Thu 7th Oct 2004 14:33 UTC

I've still got an old 100% BeOS-compatible PC in the closet, waiting for Haiku to be installed on it. :-)

Re: With Haiku-os delivery around the corner.. (Max)
by Anonymous on Thu 7th Oct 2004 14:56 UTC

The most difficult and central pieces of the project - the app server and the kernel have been barely started. AFAIK they don't even have real kernel hackers - not a single one! Haiku is still years away from it's first useable release. However if one of the Haiku devs reads this and disagrees - please post! You don't believe that Haiku is just "around the corner" right?

Barely started is hardly the truth to be honest, but I'd say their probably far from completion (what does that mean anyway?). Linux kernel isn't complete neither, nor is Windows kernel... it's a constant work to make it better. The big question is how long it is until we see something usable (Yes I know that you can use parts of Haiku on current BeOS).

There is a code bounty for whomever makes that first boot off the Harddrive and brings Haiku to a GUI which works (limited as it may be). I too long for that day more than hearing that OBFS works good or Print kit seems ok etc etc. Once it boots that's when I think things will really take off.... and for some reason, I don't think this moment is very far away. Please Axel/Ingo/Marcus and others... prove me right will ya =)

ok...drivers?
by stew on Thu 7th Oct 2004 15:01 UTC

Be it Haiku or Zeta, they both share one big problem that the original BeOS itself had:

Who's going to write drivers and applicaitons for it? It's great and nice to have the latest MediaOS on my system, but what good is it when it doesn't recognize my audio card and is far from an application that could actually take advantage of it?

I'm not badmouthing BeOS or any of it's successors, I like it and I still boot into it every now and then. But I cannot share the "we'll surpass Linux in a year" enthousiasm, because it takes a lot more than a kernel to get an exciting system.

Haiku OS is several years away.
by paul on Thu 7th Oct 2004 15:11 UTC

Let me give you a perfect example. Breadbox Computer nows owns the source code to Geos. They have only managed to release breadbox ensemble which has the newer html brower.

Geos is a good os but its 16 bit. An they say it will take time to create Geos 32.

Haiku does not have any old source code to work with they are building it own their own.

My suggestion would be try an get the person whois doing triangle os an see if he can help them cause if not is released in the next 2 to 4 years Haiku will be a after thought. The kernel is not even in aplha yet.

I dont see them releasing a beta version of Haiku for at least 2 to 3 years.

RE:With Haiku-os delivery around the corner..
by Anonymous on Thu 7th Oct 2004 15:56 UTC

Syllable doesn't have any kernel hackers either and they seem to be making (slow) progress.

Beos Max 3.1
by Anonymous Penguin on Thu 7th Oct 2004 16:05 UTC

Slightly offtopic (I hope I'll be forgiven): what happened to BeOS Max 3.1? I know there was a beta, but I can't find a place where to download it nor their homepage any longer. Pity because 3.0 was IMHO the nicest BeOS I had ever seen.

commentary
by obelix on Thu 7th Oct 2004 16:24 UTC

Haiku is 90 percent completed. It's the last 10 percent that is perhaps the most difficult, and energy consuming. I mean, if you don't want your OS to turn out like Windows or Linux.. hehe..

But no, BeOS isn't dead. Be Inc is and has been dissolved for a time now, but unless you knew the people behind the corperation, and the people behind the community, you'd assume also, that we also too, were dead. We are not.

BeOS lives on, and is quite vibrant.

Haiku Kernel
by mario on Thu 7th Oct 2004 16:36 UTC

I think they forked from the NewOS kernel sometime ago, and NewOS was in quite a good shape back then. It has multiprocessor support already in it... if I recall correclty. What's there left to do with the Haiku kernel? Isn't it so that some sort of Haiku is now bootable?

Way to go :))
by Di pa on Thu 7th Oct 2004 16:47 UTC

Go Haiku, go ...and keep BeOS alive ;) )
It will take time,
but, imho, even replace and *update* 'here and there'
old code BeOS is good.

I think best effort people without deep programming skills
(like kernel hackers?)
can do to support Haiku,
is develop their own BeOS apps.

Spot:

"Innovative ideas don't concern only the OS,
so...if you can, Be creative with your apps"

:))

RE: BeOS != Linux
by Anonymous on Thu 7th Oct 2004 16:51 UTC

I think that your comment is wrong.

Right now, Haiku isn't anything that is usable by anyone for a desktop OS. It is incomplete. There are major pieces of functionality that are just missing. Linux is currently usable. Maybe not for you, but it is usable. I can install Linux and use it. I can install new software with yum and apt. I can browse the web, I can write a paper. . . Now, there are things I can't do with Linux. I can't config my hardware, I can't mount volumes/partitions. . .

Of course, it would be foolish to say that people in the community don't want to make the things I can't do easier. Gnome has started making hardware easier to config and the like (I haven't been able to use Gnome 2.8 yet), there is a project that aims to make driver installation painless for the compiler-challenged, etc. Of course, this all takes time and work.

Haiku aims to have this functionality in their first version. Does that make it a better desktop OS or merely one that is less likely to produce a final product or whose product isn't going to be available for longer? No one in the Linux community (at least I hope no one) thinks that everything in Linux is implemented in the best way that it can be implementd. Many people work on improving this implementation so that it is easier for people like me. Just because Linux exists now and somethings aren't implemented as easy as they could be doesn't really make it less suitable for the desktop than something that isn't finished, but has the goal of having everything be easy. Heck, I'm sure that tons of people in the Linux community have that same goal of ease-of-use.

Just as a tangent, people always seem to think that Linux is harder because you can directly edit the files that control everything. Of course, this is the easiest way to implement something. Linux has built a lot of tools on top to make it so that you don't have to edit those files to change things. Are all of those tools available? Not yet and that represents a lacking, but people seem to think that if the only way to do it today is through ways that are too difficult for many people, that it will always be that way. Linux is improving, yet has a usable product today. Haiku is improving, but doesn't have a product now. For Haiku to achieve its goals, it has to go well beyond Linux to implement GUIs to do everything. Linux is implementing them, but has shipped something prior to it.

*Note: I have used linux to refer to everything that is typically in a GNU/Linux distribution. Also, when I walk about products and released stuff, I'm talking about things that have been put out there for general consumption.

Booting the haiku-kernel
by Jonathan on Thu 7th Oct 2004 16:54 UTC

There is a short tutorial on how to boot only the kernel @ http://waldsjo.se

Haiku is not BeOs
by ryan on Thu 7th Oct 2004 17:33 UTC

though they are copying many elements of the beos, its sounds like haiku is also making some changes.

Its going to end up being its own thing that pays homage to the design principles of beos. The reason for this is that beos stopped development many years ago (was it 2001?). Haiku will, i hope, evolve the general concept to include more recent ideas over time.

Haiku is now haiku.

Beos Max 3.1
by Software on Thu 7th Oct 2004 17:35 UTC

Yeah, been waiting for it... But realy don't know what happen... Have they given up? Are they waiting for Haiku?

Max 3.0 was/is a real nice Beos package.

RE: BeOSMax
by Thom Holwerda on Thu 7th Oct 2004 17:48 UTC

BeOSMax got kicked off of sourceForge because SourceForge believes that BeOSMax is illegal. Therefore, they are currently working on getting their own hosting and such.

Drivers done also
by Kian on Thu 7th Oct 2004 18:12 UTC

Not pointed out in the ICO article either, but its now possible for you to use virtually none of Be's drivers. On my desktop computer, the disk, graphics, sound, gameport and USB storage are from Haiku, and its now possible to get your keyboard, mouse and SCSI drivers from Haiku as well.

My system is in the similar state to NathanW's. Its not totally stable, but I've reported any actual crashs or inconsistancies I've had. I'm using a few R5 components just for my own comfort - BFS, input drivers, SCSI drivers - but if I wanted to I'd be virtually down to the app_server and kernel from R5.

Doesn't seem like 7 years off now, eh Eugenia?

Re: Drivers done also
by Carl on Thu 7th Oct 2004 18:51 UTC

Come on you guys! This flame against Eugenia (or anyone for that matter) is pathetic. She has the right to write an editorial with her thoughts on the Haiku project. If you disagree with her, why don't you write a paper with the opposite conclusion instead of hitting the comment field with easy remarks?

Flaming
by Kian on Thu 7th Oct 2004 18:56 UTC

I'd see what I said as more of an "in-joke" than a "flame", as I was one of the most argumentative people in the comments on that editorial, and kept trying to insist to people that Haiku was, this is not a quote but something like it "closer to 80% done" than whatever "not done" or "ten percent done" comments people were making.

Re: With Haiku-os delivery around the corner..
by Gabe Yoder on Thu 7th Oct 2004 19:38 UTC

As one of the lesser Haiku developers and a member of the app_server team, I would like to make a few comments. It is true that completion is not just around the corner, and the app_server and kernel are probably the biggest pieces left. However, it isn't accurate to say that they are barely started. The app_server is coming a long nicely and if the members of the team can find enough spare time, I expect that there will be some interesting developments in the next few months (not finished, but hitting interesting milestones). I can't say as much about the kernel, but I doubt that things are as bleak as some people think. Axel does a bunch of stuff that doesn't have any obvious results for the spectators. Once enough of these underlying pieces get built, there will probably be some relatively large jumps in perceived functionality. I am in the process of beefing up my low-level skills, and once the app_server is in good shape, Axel should have a bit more assitance than he does now. Overall, I would say things are progressing nicely, but it is true that R1 is not just around the corner.

BeOS "I'm not dead yet!"
by Anonymous on Thu 7th Oct 2004 20:15 UTC

BeOS isn't dead. Seventeenorbust.com has a new BeOS Beta client based on the latest code that screams.

Mindshare is the name of the game. Many drivers have been written lately, it supports the latest (mostly) ATI and nVidia cards, supports some wireless.

BeOS - Get back into it!

@ Inglorion
by dpi on Thu 7th Oct 2004 21:06 UTC

I think you have misjudged the difficulty of porting WINE off X11. AFAIK, WINE is precisely an implementation of the win32 APIs on top of X11. This means that porting it off X11 would be about the same work as writing it anew, and we know how long that takes...

There's a precedent you might find interesting: BeWine ("old port of WINE for BeOS").

Who's going to write drivers and applicaitons for it?

Exactly! The (potential) problem any new technology has: backward compatibility. Hope they got some kind of X11 compatibility and/or get a WINE (+ Qemy for PPC) port.

RE: @ Inglorion
by Mat on Fri 8th Oct 2004 08:07 UTC

"Hope they got some kind of X11 compatibility and/or get a WINE (+ Qemy for PPC) port."

There are X11 servers for BeOS like this one http://www.bebits.com/app/337

and there is even a Qt 2.3.0 port (uses the X11 above)

http://www.bebits.com/app/2420

but this is just to provide some X11 porting capability for those who really need these X11 apps so bad that they can't live without them - the whole thing is kind of crufty and not very robust and integrated.




RE : drivers / Blue Eyed OS
by Ben on Fri 8th Oct 2004 16:00 UTC

We all know how it's going to end : someone will use all pieces of software developped by the Haiku team on top of a linux kernel and make something really great with that.

I mean come on, could someone here tell me one single reason why the linux kernel is so bad compared to the beos original one that it is worth not using it, not using the drivers, and redevelopp a whole kernel instead ?

Remember i'm talking about the kernel here, not the crap that is usually put on top of it.

@Ben
by Kian on Fri 8th Oct 2004 16:16 UTC

Its a macro kernel. Thats enough reason for BeOS to not be able to use it. It also has latency "issues", although I admit 2.6.x has reduced these.

Haiku is going to have its own kernel. Nothing going to change that. If you really want "BeOS on Linux", go help B.E.OS

@Ben
by ModeenF on Fri 8th Oct 2004 19:09 UTC

I'm not a Linux developer but I intend to be a BeOS developer.
What I have heard is that Linux has a lot of drivers built in the kernel (way rebuild if you don't need to)

BeOS don't have that or don't have as many drivers built in the system.

So BeOS only needs to load those drivers that are needed not a kernel with all the drivers.

Haiku kernel
by paul on Sat 9th Oct 2004 22:17 UTC

First the linux kernel is different from the linux kernel.

Is beos totally compatable with the linux kernel.

Syllable is similar to beos but its not beos.

My opinion is we should wait for Haiku to finish

what they have started.

Also you could might be able to use syllable an port beos applications to run on it.

My question? Is syllable closer to beos than linux?