Blue Eyed OS, one of the premiere open source BeOS replacements, released their first downloads yesterday, a beta version of their kernel_server. It can be downloaded from the Blue Eyed OS website. They’ve also added a new screenshot as well as an interviews section to their homepage.
Blue Eyed OS Releases Beta Downloads
About The Author
Adam Scheinberg
Technology Executive • Web Developer • Father • Foodie • Music Snob • OS enthusiast
Follow me on Mastodon @[email protected]
what is the point of releasing such a small chunk of the OS? There’s not really much you can do with it. This looks quite like nothing more than a publicity stunt imho.
well you can replace the parts of the original beos with these parts.
you have to start somewhere
Yeah, replacing parts is what most of these projects are for. For example, in the OpenBeOS project, the file system is almost in stable, and a lot of the developers simply erase the vanilla BFS and put in the OpenBeOS File System.
Check out the binary, it only weights 70 kb !
Everyone can download a file with that size …
—
http://homepage.mac.com/softkid
hmm.. from what i know and read in their readme it’s for linux. You can’t replace BeOS parts with it.
This release is only for curious developers, its interest is very limited, because limited to the Kernel API of BeOS.
Of course, you need a linux kernel to use it.
If you want to see more interesting things, just wait the release of a beta version of the app_server and some small test apps.
Regards,
Guillaume
Microsoft leaked a whole full beta of Longhorn. Blue Eyed OS needs to keep up!
It’s called release early release often…it’s open source…nobody forces you to use it if you don’t think it can do anything…how is this a publicity stunt?
stop trolling…you should know better than to troll BeOS on this board 😛
-bytes256
I think I understand what this is. It’s a perverse plot to take BeOS, which is well-known for being responsive, and making it slow and bloated by implementing it using Linux and X11, those well-known race horses.
“and making it slow and bloated by implementing it using Linux”
He guy, you are the greatest troll on the planet, the kernel_kit is now available for download, make a real benchmark instead of inventing such stupidies.
Regards,
Guillaume
Of course I am the greatest troll on the planet. I was behind the greencard spam back in ’93, I go to great lengths to spell like a monkey, use lots of insults and post under pseudonyms.
As you can see.
In the meantime, have you got any benchmarks to support your claim that Linux and X11 are fast and responsive?
I won’t be able to test it myself, since I’m on the other side of the world from my BeOS machine and don’t have any desire to install Linux on anything.
I don’t know how many people can get it runs, because your libroot.so has the wrong name of file. In the tar.gz, it comes as libroot.so.1.0.0, which it’s supposed to be libroot.so.1.
=============================================
$ ldd kernel_server
libroot.so.1 => not found
libc.so.6 => /lib/i686/libc.so.6 (0x42000000)
/lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
=============================================
I had to rename from libroot.so.1.0.0 to libroot.so.1, then put in the /lib/ to make it works. Now, it loads perfect, but what am I supposed to do with it? 😛
==============================================
$ ./kernel_server
kernel_server: Starting up
API_THR: Team Table Init
API_THR: Thread Table Init
kernel_server: out_port: 0
kernel_server: in_port: 32769
kernel_server: sem: 0
kernel_server: Server waiting for msg of size 52 on port 32769
===============================================
Anyway, I do really look forward for B.E.OS to arrive the final version. It sounds good when I read in your README about possible to get it runs in the FreeBSD, since it’s POSIX compliant. I won’t put the high hope, but look forward to get it works in FreeBSD. 🙂
BTW: I am using RedHat 8.0 on the diff computer.
Regards,
bsdrocks
Cool! I’m happy to see progress in the BeOS inspired OS projects.
Anyhow, is there a chance of a souce release? Currently, the only Linux installation I have is on my iBook, it won’t run the x86 binary.
..about, well, not a whole lot.
BTW, what do you folks think, what are the odds that MacOS X will be ported to something x86-like? After Adam Scheinberg’s excellent review (“Month with a Mac”) I got a liking to MacOS X. It seems like an OS I wouldn’t mind paying money for.
If projects like this didn’t show us something as they progress, we’d be complaining that said project is dead.
The libroot is called: libroot.so.1.0.0 and not libroot.so.1 because of the version managing of the lib.
To create symlink or rename the lib solves the “issue”.
“Now, it loads perfect, but what am I supposed to do with it? :-P”
you can use the KernelAPI of BeOS under your linux, it’s not really usefull in comparison with the app_server, but you can start to code things by using OS.h, debug, test, benchmarks…
“Anyhow, is there a chance of a souce release?”
Yes.
“In the meantime, have you got any benchmarks to support your claim that Linux and X11 are fast and responsive?”
Sure! I did benchmarks a year ago, results where positive. What’s more I wrote some lines about kernel performance months ago and released a “proof of concept” (http://blueos.free.fr/Static.tar.gz) some weeks ago.
Regards,
Guillaume
Guillaume Maillard,
Nice wallpaper, so I just need to run the test file in linux right.
I’ve done benchmarks of Win32 GDI vs X11, and they show X11 very positively. I’ve got some free time now (home from college for a month!) so I’ll dig them up and post them. If anyone wants a copy, just email me and I’ll send them when I’m done. As for Linux, it’s performance is a dead horse. Everyone and their mother knows that Linux is blazingly fast for the UNIXy operations a kernel is supposed to do. I suggest taking a look at Andrew Morton’s audio latency site or entering “Linux performance benchmarks” into Google.
So I can understand, that “proof of concept” is designed to demostrate the feel of how quick your project will display windows on the screen with Xfree 4.x.
It is basically as fast as launch an xterm. But how does this represent system responsiveness.
Anyway keep up the good work, I am very excited to have a play with all the “Open” BEOS variants in the future.
It’s called release early release often…it’s open source..
I’m sorry to say that B.E.OS is not Open Source( http://opensource.org/ ), nor Free Software ( http://www.fsf.org/ )
Really sad, I can understand that there is some people that don’t like the BSD license(I love it, but it’s a fact that some people just prefer the GPL, which I respect), but they could at least used and Open Source approved license!
Even if you are a control freak(that is the only explanation I can think of) there are quite good licenses for that, eg., apache license that doesn’t allow derivate works to use the name of your project, or the GPL where you make sure you will have access to the src of all future modifications…
IMHO there is no excuse to reinvent the wheel and create Yet Another (Incompatible) License… please, the licenses mess is already bad enough as it is…
\k — that will never understand why people fears the greatest open source feature: The Fork
I don’t like BlueEyedOS.
It’s not because of the Linux kernel – even if I think that the Linux kernel is a bit bloated, it could be ported down to size and successfully used in a reimplementation of BeOS. And the difference between microkernels and monolithic kernels is becoming less important than it once was.
It’s not because of initially being implemented in XFree86 – as much as I hate XFree86, I think the replacement should be the Fresco project, not necessarily whatever homegrown thing the OpenBeOS guys come up with. And they could always switch.
No, I don’t like BlueEyedOS because it’s proprietary. If I’m not mistaken, the source isn’t even available to anybody except select developers. And even if the source is made available, the license still prohibits forks from being made. It’s eerily reminiscent of Microsoft’s “Shared Source” plan.
There are tons of free licenses to choose from. Guillaume Maillard and company purposely created their own because they wanted to retain control over the source code. If I wanted to use a group-controlled product, I’d use Windows. OpenBeOS may have as many flaws as BlueEyed, but at least it respects the community – and acknowledges that, if someone, anyone, wants to take it and improve it, they can do so at their whim.
Guillaume, if you decide to change your license, I’m sure you’ll soon gain a huge following. Until then, as the great RMS said, “Share the software/You’ll be free, hackers, you’ll be free.”
I know the open Be teams seem to have not-pure open source opions, but there are people (hardware, software vendors) behind some of these projects that had an investment in BeOS technology. Their products like high-end multimedia require access to the hardware that Apple will never provide, and access to source that MS will never provide. These investors are a little to shy to be commonly known, but the teams trying to recreate BeOS do have a market to actually make money from of they pull it off!
Something to remember…
I think I understand what this is. It’s a perverse plot to take BeOS, which is well-known for being responsive, and making it slow and bloated by implementing it using Linux and X11, those well-known race horses.
Well, congrats to myself, I just caught myself the troll of the day. Linux, underneat, is way way way more faster than BeOS, while XFree86 have a potential to be responsive. If you try the latest latency patch for the kernel, the same patch used in BEOS, you might even think that Linux is faster than BeOS.
The only reason why people seem to think that Linux is slow is because of a unresponsive UI. And one of BEOS goals is to fix that. BeOS itself is very slow, for example when you are doing things like transfering large files from one directory to another. But the reason why people think it is fast is because it *feels* fast.
“And even if the source is made available, the license still prohibits forks from being made. It’s eerily reminiscent of Microsoft’s “Shared Source” plan.”
Ask yourself:
“Why do I need the source code of B.E.OS?”
If the response is:
1/ “I’m curious, I want to see how it works”
You will be able IMHO to do it when B.E.OS will be completed
2/ “I want to improve it, port it, etc…”
Join us, it cost nothing, but you will have to really do it, not only “chat”, of course.
3/ “I want to be sure to never have to pay for it”
B.E.OS is free and will stay free
4/ “I want to create a new OS based on B.E.OS”
ask yourself why you want a new/different OS and why not make evolve B.E.OS in the right way
5/ “I want make money with it”
You will never be allowed to do that.
If you have an other answer, post it.
Regards,
Guillaume
“So I can understand, that “proof of concept” is designed to demostrate the feel of how quick your project will display windows on the screen with Xfree 4.x.
It is basically as fast as launch an xterm. But how does this represent system responsiveness.”
I think that you forgot to read the README
—— Press SPACE to create a new window —–
Regards,
Guillaume
I think that there are no chance for any new operating system today. It is a chicken-and-egg problem. The world is Windows and the only one alternative witch real chance is linux, because of opensource concept and of zero cost.
All other O.S. will only be adopted in specific niches. There are no chance for another comercial O.S. This is the reason why BeOS and OS/2 died.
B.E.O.S use a modified linux kernel to partially resolve the chicken-and-egg problem with drivers and basic software, but it isn’t free (as in speech) and it isn’t opensource. It will be only a geek toy like BeOS was.
4/ “I want to create a new OS based on B.E.OS”
ask yourself why you want a new/different OS and why not make evolve B.E.OS in the right way
Well yeah. Unless that person don’t share the same goals as you do. Just say that developer couldn’t care less about source compatiblity with BeOS, would he be successful in making B.E.OS evolve the “right” way? Or what happens he decides to change Linux for FreeBSD? Or completely change the UI? Or something like that?
Besides, one major problem you would get from this is that no business on earth would want to support B.E.OS. Just say I have a company. I personally wouldn’t want to push B.E.OS as a product of mine. Why? Can I differentiate my product from B.E.OS? Nope. Sure, maybe I could just distribute B.E.OS, but anyone can join in and I wouldn’t have a big competitive advantage anymore.
Then I’m quite sure you would say “Then why not they just build apps for B.E.OS”. Why should they? There is no company behind B.E.OS, and therefore no market which means hardly any users. Do you think Linux would be this famous without the distributors and companies behind them? Doubt it.
5/ “I want make money with it”
You will never be allowed to do that.
And guess what? You would never have any commercial interest behind it and therefore B.E.OS would stick as a hobby OS. B.E.OS has a lot of potential, tonnes of it with its concept of taking the best from Be OS and Linux. You would have a good product. But for sure you would not have good marketing, and therefore it would stick as a hobby OS.
You loose practically nothing from relaxing the license. Linus Torvalds don’t make money from companies like Red Hat and SuSE, yet he sure is a millionaire. This project wouldn’t even exist if everyone had the same philosophy as you.
But before you flame me, I think it is your choice. You guys write the code, you guys decide what license it would be under, you guys decide its fate. I do not code for B.E.OS, so I have no say in it. But what I wrote is my opinion. If it isn’t your opinion and you completely don’t agree, disregard it. But first, think about it first.
Personally, I like B.E.OS’s concept. I think it has a much higher chance of succeeding than OpenBeOS. I would probably use it when it is release if it is good. (I won’t if it isn’t, obviously).
“I think that there are no chance for any new operating system today. It is a chicken-and-egg problem. The world is Windows and the only one alternative witch real chance is linux, because of opensource concept and of zero cost.”
Then you also agree that OSnews is a very unnecessary site on this net. After all, if there are only to be Windows and Linux, a site reporting on OSes in general serves no purpose.
Hmm… Perhaps we should rename OSnews to YALDnews instead. OSes are dead, as Marcelo said.
“You loose practically nothing from relaxing the license. Linus Torvalds don’t make money from companies like Red Hat and SuSE, yet he sure is a millionaire. This project wouldn’t even exist if everyone had the same philosophy as you.”
You misunderstood the philosophy behind it.
RedHat and Suse CAN’T make (and never did) money with Linux itself, what they sell is a bundle (kernel, libs, apps, docs…) and support, nothing more. When Redhat pay engineer to improve Linux, they think at their profits.
It’s the same for B.E.OS, if you build bootable CDs with B.E.OS, add apps, support and documentation, you are allowed to sell them 100$ if you want, what’s more noone will request you to release the source code of your installer or apps.
IF for example, you want see a B.E.OS under FreeBSD, you can join us and do the port. What we will not “accept” for example is someone proposing to break the complete API or to write a monothreaded app_server, because that’s not B.E.OS.
Regards,
Guillaume
I should add that IF 20 news coders [ convinced that B.E.OS must be GPL or BSD ] join B.E.OS, and that they represent (in term of produced code/activity) the majority, B.E.OS will become under GPL or BSD licence when completed.
As I said before, B.E.OS is what B.E.OS coders decide.
Regards,
Guillaume
The only reason why people seem to think that Linux is slow is because of a unresponsive UI. And one of BEOS goals is to fix that. BeOS itself is very slow, for example when you are doing things like transfering large files from one directory to another. But the reason why people think it is fast is because it *feels* fast.
You are exactly right. But then again, FEELING FAST is the POINT! If the user gets annoyed from waiting for the UI to respond to him/her, then who gives a crap if he can copy a gigabyte from one folder to the next in under a nanosecond?? Responding to the user is the PURPOSE of the OS. Everything else comes second, including all the geekness. The file system, vfs, vm and all that kind of stuff are important, but the OS is there for the USER, not for the data.
2/ “I want to improve it, port it, etc…”
“I think that there are no chance for any new operating system today. It is a chicken-and-egg problem. The world is
Windows and the only one alternative witch real chance is linux, because of opensource concept and of zero cost.”
Then lets fold up OSNews and go and read WindowsXP instead!
Whoops MS already folded that as well.
Seriously, there is plenty of room for alternative OSes but they must meet ever rising expectations and that becomes harder & harder over time. Its easier for BeOS clones to get back in since we all know what they are supposed to do. But if one of these groups was to create an entirely different OS with no familiarity then that would meet far more resistance.
One way to significantly help an OS chances is to make it easy for other OS developers to recompile for and that means the new OS must bend over backwards to support atleast one & preferably several of
Java, .Net, Qt, GTK, wx, Posix, any other good x platform tool that reaches some of Windows, MacOS, Linux. IIRC BeOS had minimal or zero support for most of these. There were some partial ports of GTK & Qt but I don’t think any gui BeOS app was ever created with these. Luckily if any any of these does get ported fully, then all the BeOS groups can benefit. Ofcourse Blue Eye gets the benefit of whatever Linux has already.
nuff said
“The only reason why people seem to think that Linux is slow is because of a unresponsive UI. And one of BEOS goals is to fix that. BeOS itself is very slow, for example when you are doing things like transfering large files from one directory to another. But the reason why people think it is fast is because it *feels* fast.”
What’s wrong with you people? That’s right, Linux ‘feels’ slow and other OS’s don’t. How long will you loons keep calling linux fast when you really only mean it’s fast without the GUI. Don’t you realize that by now we’ve all tried a few distros and can plainly see that Linux is slow, regardless of what numbers are posted somewhere?
Silly penguins, *nix are for kids.
Linux doesn’t feel slow to me. I abuse it quite heavily, running compiles, learning Houdini (the apprentice program is wicked cool…) doing renders in the background, and it behaves nice and smoothly under big loads. On an onloaded system, sure it’s slightly slower than Windows XP* but while doing real work, it slaps XP silly. Of course, if you haven’t used Linux for more than five minutes, I wouldn’t expect you to know that…
*> Even that depends on the desktop. KDE is slightly slower, GNOME is about the same (the two only differ because of app startup speeds in KDE) and WindowMaker/FluxBox ownz XP.
Many people around here really seem to confuse low latency with performance.
XFree86 doesn’t make Linux slow and unresponsive, XFree86 is slow and unresponsive in itself.
It doesn’t mean Linux is only fast without a GUI either. I think many would agree that FreeBSD is a pretty fast system, yet if anything, XFree86 feels much slower and less responsive on FreeBSD than it ever did in Linux (even if I use linux 2.2.x).
XFree86 just needs to improve a great deal soon or be replaced or linux may never lose it’s reputation for a crappy UI system (even if it ever is no longer the case).
However, gnome2/kde still feel noticably faster than win2k’s UI on the same hardware for me, it doesn’t feel faster in every way, but overall it does.
Dont like BlueEyedOS? Go and help OpenBeOS for gods sake! They need all the help they can get!
> Join us, it cost nothing, but you will have to really do
> it, not only “chat”, of course.
That’s exactly it. People need to be able to contribute as loosely as they want. Read ESR’s “The Cathedral and The Bazaar”; bazaar-style development is what made the Linux kernel you’re relying on so great.
> 3/ “I want to be sure to never have to pay for it”
> B.E.OS is free and will stay free
Software libre is what we want. Gratis is nice, but not as important. Aside from the fact that I’m on limited income, I would buy a boxed copy of Debian in a second, or at least give them a constant stream of donations.
> 4/ “I want to create a new OS based on B.E.OS”
> ask yourself why you want a new/different OS and why not > make evolve B.E.OS in the right way
Because I don’t want the right way! Many people often talk about forking as a problem, when in reality it’s one of the greatest features. The fact that someone can make a different product from a codebase is what’s wonderful. When some Blackbox people got fed up, they made the better Fluxbox. When some Mozilla people got fed up, they extended it to make Galeon (not strictly a fork, but close). If someone develops a fork, and the fork is not popular, it will just close; if it turns out to be better than the original, i.e. more well-liked by the public, then the public is getting the product they want.
> 5/ “I want make money with it”
> You will never be allowed to do that.
What’s it matter if someone could? If it means that your product is better able to reach its ideal audience, why not? If someone who sells your product is able to offer something worth the money, then so be it. I don’t think it’s a coincidence that RedHat is making a profit, while SuSE, who does not offer their current OS du jour for free download, is almost bankrupt.
> RedHat and Suse CAN’T make (and never did) money with
> Linux itself
RedHat is making money. All of their business involves Linux in some way. Because of RedHat, many more people use Linux than would otherwise. Isn’t that what matters?
> I should add that IF 20 news coders [ convinced that
> B.E.OS must be GPL or BSD ] join B.E.OS, and that they
> represent (in term of produced code/activity) the
> majority, B.E.OS will become under GPL or BSD licence
> when completed.
Unfortunately for you, there exists a project called OBOS. Regardless of its technical merits (though I do believe they’ve got you beat, especially considering they’ve got Axel doing BFS), it’s currently BSD-licensed and is attracting a very wide variety of coders. Most importantly, anybody can submit even a ten-line patch to make something work better, and the OBOS team will accept it if it’s good.
I’m sure you could find 20 dozen coders who would make a good contribution to your code and vote to make it BSD or GPL. But every single one of those coders will join OBOS instead. Sorry.
Hello all.
1 – BeOS users: I’m a BeOS user. I love BeOS.
2 – Linux users: I’m a BeOS user. I like of Linux.
3 – BlueEyedOS: I’m a BEOS user. I work in BEOS.
4 – OpenBeOS: I’m a OpenBeOS user (I use OBFS, print_server, opentracker/opendeskbar, mail daemon replacement, stylededit, etc). I like of OpenBeOS.
In an entire year, our developers has worked hard to have a kernel responsive than BeOS. In BEOS site, downloads section, we show one. Do you can make your tests with BEOS and BeOS kernel to see what’s is more responsive? Which have more drivers? Which is more POSIX compliant? The source code will be released later. No panic.
We have a kernel 93% finished. We have an app_server 80% finished. We have a XML kit 99% finished. The bigs kits of the BeOS we already have a replacement to be released (ok, still misses media kit), but what’s our problem? I’m planning a new design to BEOS website in the next year (Be website style+Gonx style). What’s still misses to BeOS cummunity?
Please think about our christmas gift…
Michael Vinícius de Oliveira
BlueEyedOS Webmaster
I read in BUG-BR:
Who needs BFS? On Linux, we have ReiserFS & XFS, which can replace BFS.
This is a joke, isn’t it?
eT
No, it’s not open source. BEOS is completely closed source, and they seem to be proud of this fact. Whether they release a kernel patch or the whole OS, it’s still mainly proprietry at the moment.
Thesis: BFS can be replaced by Reiserfs or XFS.
At first I thought this is a joke, too.
But… I “heard” Reiserfs is developed into a more database-like filesystem with user-definable attributes, shipping with the next version. I do not know much about XFS, but it is damn fast! So is Reiserfs, and it is very stable, too.
What I am wondering about: how will Linux take advantage of a database-like filesystem using attributes without kernel-support? (I remember Eugenia talking about the impossibility of using BFS`s capabilities in Linux.) Will there be Just Another Filesystem Daemon?
Conclusion: sit and wait… I will not call anybody´s statement a lie or joke until I do not know better.