Post a Comment
I sure am. Haiku has been making brilliant progress, and it's been far to quiet about Haiku itself that last few months, I'm confident that a lot is happening under the hood.
Axel working full time on Haiku sounds amazing. He's a guy who walk the walk rather than talk the talk. My hopes are people will get encouraged and fund him to at least new year =).
I checked the status page recently and all "kits" has been updated to what I guess reflects some sort of "October status". Judging from that, we are very close to seeing some sort of Beta...
RE: Who isn't encouraged to donate now?
Axel working full time on Haiku sounds amazing. He's a guy who walk the walk rather than talk the talk. My hopes are people will get encouraged and fund him to at least new year =).
Yeah, my one week stint of trying out various flavors of BeOS ended a while ago, but I keep on hearing about this Axel guy as a coding machine.
He sounds like that SkyOS guy, and with Linux desktop efforts going nowhere these guys still have a chance of making an impact.
"and with Linux desktop efforts going nowhere"
I beg your pardon?
I keep trying absolutely everything. Smaller projects can be nice, even "sweet", but I don't see them becoming many people's main desktop OS for any foreseeable future.
As to linux I have been using it as my main (99%) OS for years. So have many other people. So it must be "ready for the desktop" somehow. Linux has always supported all my hardware, something that smaller projects have always failed to do, and by far.
And if you are suggesting that linux isn't improving all the time, and at a very fast speed for that matter, that is utter nonsense: it suggests me that you are talking about something you don't know.
RE[6]: Who isn't encouraged to donate now?
RE[4]: Who isn't encouraged to donate now?
I detect a pot calling whom they presume to be the kettle, black, with what you've just stated. Heck, why not just call all zealots incredibly intolerant and narrow-minded? Then again, why am I replying to a blatant troll that only bothers to insult whole projects that they don't believe will get anywhere, without adding anything substantiative about why they whine stupidly about someone they label as a zealot, while being an at least as obnoxious anti-zealot for the given topic?
Of course, what you think about Haiku and the developers and their chance of success (whatever that is defined as being) means less than squat, because at least they have a dream and are working towards it, learning useful things from the experience, and not badmouthing projects/people they don't believe will amount to a hill of beans, because they're too busy coding. They're walking the walk, and all you do is blabbermouth talk.
RE[5]: Who isn't encouraged to donate now?
Now, don't be hasty, Master Meriadoc. It seems a single person disagreed with you. The other ones appear to be replies to your "gah! -1!" All of you should take a step back and breath in, breath out. It's not the end of the world if people don't agree on the goals/usefulness/rightousness of a project. Time will tell. Anyway, we're all just trying to amuse ourselves while we're here on planet earth. There is no historical end-station or goal. Linux will peak, eventually, and possibly evolve beyond recognition. Open-source may not be the final destination after all, (open/closed source may become irrelevant), and computers most probably won't work the way they do now, forever.
RE[3]: Who isn't encouraged to donate now?
RE[5]: Who isn't encouraged to donate now?
And XFree getting ported makes Linux a serious desktop OS? Please! I suppose the fact that web servers exist for BeOS are enough to make it a server OS, and for me to criticize it for its small server market share?
XFree doesn't make a desktop. Linux didn't have a serious desktop product until GNOME 2.0 came out in 2002. Only in the last year or two have the major Linux vendors even paid more than cursory attention to desktop support. The biggest, RedHat, *still* doesn't care about the desktop. If market share is what you get hard over, realize that market share is only built where companies try to build market share, and only very recently have Linux companies marketed Linux as a desktop.
Rayiner, you just made my point, maybe inadvertantly.
XFree is(was) the basis of the desktop, unless you were targetting the framebuffer.
The biggest, RedHat, *still* doesn't care about the desktop. If market share is what you get hard over, realize that market share is only built where companies try to build market share, and only very recently have Linux companies marketed Linux as a desktop.
Yeah, no shit. The RedHat VP "Tieman" or something was just saying a year ago that windows is a better fit for the desktop for many tasks. RedHat or Novell still don't care about the desktop because they don't make any money off of it. Surprise, surprise.
But I'm glad to see you come around finally.
As this is an OS related news site AND this is the first paid coder to a really neat effort, I say it is worth to mention.
AND your assumption is wrong, they release previews and nightly built previews.
It is already possible to run Haiku (If you like you can try it using Qemu - Prepared images: http://www.schmidp.com/blog/haiku/finally_new_haiku_images..html)
Understandable if you are a new OSnews visitor.
In short, Haiku aims to be a BeOS (If you don't know what BeOS means, do a search in your preffered search engine) compatible OS with the same feel/Gui/API and extend it with some new power ...
All of the work is done in free time, and this news means that Axel is released from his heavily guarded basement to much more convienent place to code
Not to do discredit to the other coders (you are highly appriciated) but sometimes it looks like it's an AxelOS
'BE the difference that makes a difference' - JEWEL
How knows ...
but maybe he's already in talks with an obscure CPU-design team to make a very inexpensive, powersaving, heavily multitreathed aware, quad-core, multimedia Haiku CPU.
Yeah that's what we need, buzzwords.
Remember the time BeOS was a Buzzword enabled OS.
'BE the difference that makes a difference' - JEWEL
This is good news, but for some reason only US citizens can send in $ even though I live here half my life.
Funny you should say that
Actually I have been designing an obscure cpu that is 4way threaded per core, and uses many cores per MMU and best of all it doesn't suck on memory accesses, ie all load/stores look like 1 or a few instruction slots for all DRAM accesses while P4/Athlon can be anywhere from 1 to 1000 slots depending on locality ie cache misses.
Ofcourse the scheduler, communications and the object memory protection are in the hardware so there is less need for most of what passes for an OS. My dream desktop will be Tracker even if I don't get to see Haiku ported. The more interesting part is the native language and compiler, a combination of C,occam, Verilog for concurrent objects. All in development of course.
I could only dream of having full Haiku on board, but in fairness they should complete x86 1st and when we are all done, look for ports to better hardware. It will be available for those that are tired of pretending the x86 is the great arch it isn't!
transputerguy
threads (or CSP processes) are the only sane way to go to solve most comp architecture problems, but try telling that to the masses!
I have a conference paper for anyone thats really interested, write me at transputer2 at yahoo or look in the usual NG
John
This is good news, but for some reason only US citizens can send in $ even though I live here half my life.
You can use Paypal to get money to them. At least you could when I did it. You won't get any tax benefit though.
According to http://haiku-os.org/contribute.php?mode=donate "For citizens of other countries, Paypal is also available here:"
They should probably format that page a little better.
_V_
Actually I have been designing an obscure cpu that is 4way threaded per core, and uses many cores per MMU and best of all it doesn't suck on memory accesses, ie all load/stores look like 1 or a few instruction slots for all DRAM accesses while P4/Athlon can be anywhere from 1 to 1000 slots depending on locality ie cache misses.
I'm playing with VHDL as a hobby using Spartan 3 dev board. Which ASICs do you use, and how many gates/macroblocks it usually takes up to design such CPU ? I'm also interested in the paper. My email is spideros1 at wp dot pl.
I also hope that Haiku will support PCI BIOS for PCI access, unlike probably BeOS does, because I can't boot BeOS it on my laptop 
Hi spideros??, Jonus
I use Spartan3 FPGA board, the $100 unit from xilinx- digilent with the PS2 & VGA ports. The plan is to bring up this board 1st, later to move up to a V2Pro or V4 board with special RLDRAM for 3x the performance, and then worry about ASIC later.
Each processor element (PE) uses just 1 BlockRam and about 500 FF/Lut area cells so it cna be replicated a few times. It needs a bit more work for unfinished instructions. Each PE on simulated benchmarks compares with an xp2400 about 8-12x slower but it has no memory penalties for missed cache (since no data cache). The compiler needs to be alot more robust before more nos can be estimated.
One of these PEs should be enough for a Haiku like OS to fly though as they compare very well to x86 at a few 100MHz but support is included for concurrency & protection.
The MMU exists only as SW library so I don't yet know how much HW it will take, I guess about 2-4x as much as 1 PE. The available bandwidth of memory issues that the MMU provides is meant to keep from 4-20 PEs satisfied depending on type of code that runs and memory used.
The compiler is in continued development so gets in the way of HW development but usually contributes back in other ways besides executable code. Another shared idea is the scheduler structure is very similar to the tree structure the compiler uses, so completing one helps get the other done.
And so it goes
Hello John/transputerguy !
I'm always happy to see your posts, especially in a Haiku thread such as this. How's your design progressing? I should probably email you about that and about your paper.
Did you read about the OpenGrapics Development board? They won't be supercheap, but they'll have a large Xilinx FPGA for the graphics core and a Lattice FPGA for the PCI/etc interface. Those could be used for other things, you know.
Best regards,
Jonas Sundström. // jonas.kirilla
The OpenGrapics Development board has come up before, I am always interested in what others figure can be done in FPGA thats on a PC level.
Ofcourse once all the IP is developed, it would interesting to merge it all into a single FPGA although not sure if I will bother with PCI, I prefer fast serial links. I like the idea of sub credit card size PCB modules with FPGA + DRAM + something-else TRAM style.
regards
John jakson
transputer guy
I dont think so.
Axel is THE man behind the kernel.
All other's work depends on Axel's work. If kernel fails, its useless to code a perfect "app_server", for example.
It's right to invest that money in a full-time payment for Axel work. Haiku should receive now a development boost.
I don't think anybody would be discouraged by this. In fact, it might actually give them more encouragement. With Axel working full time, there will be a lot of interesting stuff happening, and activity tends to encourage more activity. One of these days, maybe I will even do some of the stuff that I am planning (if real life would quit getting in the way).
Sorry to post anonymous - I am not at home where my password is. :-)
No, we are not hiring. Sorry. Yes, axeld is really, really good. No, he did not write the whole OS himself; "JUST" most of the work on the kernel, BFS, the Mail Kit and maintaining and improving OpenDeskbar and OpenTracker.
Mario asks a very insightful question, one that I considered very strongly before agreeing to this. We talked at the admin/developer meetings before this happened and no one had any objections or, honestly, wanted the job at the time. The kernel and the app_server are the two pieces furthest from ready to be done and axeld will be spliting his time between them.
http://behaiku.altervista.org
This site hosts regularly new builds of Haiku.
Where can I find a ISO of haiku... it seems that it's only available a 7z image, is it a beos cdwriter format...???
http://behaiku.altervista.org/index.php?mod=Download
>>> http://behaiku.altervista.org
>>>
>>> This site hosts regularly new builds of Haiku.
Some Haiku's screenshots are available there.
http://bug-nordic.org/gallery.php?gal=haiku
"And once again, Linux fanboys have overtaken a forum which has nothing to do with Linux."
Totally wrong. I was just replying to the following sentence:
"and with Linux desktop efforts going nowhere"
And in my second post I said that I like Haiku, I really do. I find that there is no need in *any* thread to say: "the OS we are talking about ROCKS, whilst OS >name OS here< sucks."
People...i think we need some daily builds. We have only 1 site in the whole world with haiku images (http://behaiku.altervista.org/). That's frustrating...
Most posts didn't deserve to be modded down - it's just what people do with their powers :-)
Just don't take it personally, or see it as a sign for a mad community - it takes only 2 people to exclude a post out of default settings view. And even 5 people don't make a community.
Despite all the hot air and mighty furor over the Linux Desktop's supposed lack of progress or its continual evolution into a usable desktop, the parent's point has been lost. The point? Simple: Until a solution appears that can be easily and usefully installed to a computer with FULL hardware support no one cares. And until [INSERT FAVORITE OS HERE] is capable of doing something that the 'average' user is concerned about no one will care whether it overtakes or surpasses the [INSERT DOMINANT OPERATING SYSTEM HERE], as long as they are able to do whatever it is that has their fancy with that OS.
Right now, quite frankly the War for the Desktop has barely even begun. No one can say what will be the case in the next ten, twenty years. So at this point in time Haiku has just as good a chance to be come the Desktop enviroment of choice as anything else...whether that be SkyOS, KDE, GNOME, or whatever else might exist. It took us nearly twenty years or more to get Windows to the point where anyone would want to use it; it may take another twenty years to create something 'good enough to replace it.
Fact is, we won't know what the next 'killer app' is until its already become blasé...right now on the Desktop no one has that aspect working for them...
--bornagainpenguin
I just tried running an Haiku image on Qemu and i was really impressed. In less than 10 sec i was in front of a full graphical system, launch Tracker and Deskbar and you got a "full" desktop environment. Its not fast yet, but everything is almost there and i thought it would have been a lot more buggy than that. I one word, congratulations!!!
Haiku has a big advantage, the specs are written and the API is stable from day one.
Every dev goes in the same direction, they've got full specs for a true (proved to work) unbloated, modular, multimedia desktop!
Keep up the good work!!! Even if it take years to complete, there is nothing that could make it fail.
We (long time BeOS users) all know the people and sure know how much they have done so.. Yea!
What I am more impressed with is that someone got an image of the new OS and ran it on qemu. This is the way to deploy it to new users who want to try it instead of them burning a cd and running it live.
-jefro



