I have a fantasy that 20 years from now, Haiku will be the operating system of choice for hackers with design sensibilities (NB: I didn't say graphic design) who want a powerful, flexible, programmable OS that just works and requires minimal configuration.
This niche is currently filled by OSX, but I think it's slowly falling out of favor because of Apple's increasing focus on the end user and "iOSifying" everything.
Of course Haiku is still some ways from it, but I've been following its development for a while now and I feel like it's on the right track. If one or two manufacturers end up providing good hardware for Haiku to run on when it's a bit more mature, it'd be wonderful and then my vision would happen.
There are worse foundations and inspirations to have than BeOS - a man can dream :)
FWIW Debian stable (esp. with KDE IMO) is now very close to this. I have no patience at all for a fiddly daily machine, and for the most part I've been pretty happy with it.
There are probably other Linux distributions (Mint?) that can fill this niche.
I only mention this because, if people are looking for something right now, setting aside a little bit of time to try out Debian stable or Mint might be worthwhile.
But yeah. I'd love to see Haiku become a strong alternative OS. I've been keeping an eye on it ever since it was announced. (I still have the BeOS R5 disc that I installed on a PPC machine years ago.)
Desktops like KDE can preview images in a directory. But BeOS was on another level with this stuff.
Imagine if (1) your filesystem was a high-performance nosql database; (2) your desktop was just a thin layer around this, optimised for navigating around and querying the stores. But NoSQL has never since been so smooth and beautiful to interact with.
The mail system: a daemon drops files in a directory and makes a ding sound. The user then just browses the directory containing the mail messages. (The file explorer detected the metadata and displays this as columns.) When you double-click on an icon, you get a special-case form that displays the contents of the data. http://archive.arstechnica.com/reviews/3q99/bemail/bemail-sc...
If a unix were to do this stuff, it'd stop being unix: metadata on files is incompatible with the everything-is-a-stream-of-data philosophy. ryanweal writes below, "think everything is an icon instead of everything is a file". Brilliant. Haiku team should take "everything is an icon" as their motto.
The people that used BeOS and look forward to Haiku are going to disagree with you on that. I'm guessing that they look at Linux distros as a hodge-podge of various libraries without any cohesive philosophy.
Yeah, that's my position. I use OSX at work and Arch at home; and while I love Linux, it is quite messy. (Which works to its advantage in many ways- it wouldn't be that powerful and flexible if it weren't)
It reminds me of BSD in some ways: small, focused, albeit on the desktop instead of the server. I love using it, and have a VM running alpha 4 on every computer I own!
Once the Gentoo Grinch grows a heart and make an installer your problem will also be solved. I can get any package and any library built the way I want by typing `emerge`.
Gentoo certainly gives you the tools to make a great system, but it needs far more than just an installer to get there. The whole distribution needs a lot more curation. It takes far too much research to figure out which options are potentially useful and which are just relics or break too many other packages or include a feature that is impossible to put to use.
I guess I'm not surprised. 100% discussion about whether Haiku deserves to exist, zero discussion about the post. Can anyone talk about the significance of the removal of the hard processor limit?
Everyone is an OS consumer. Far fewer are running Haiku on a 9+ core server. Write what you know.
To be fair, the scheduling might matter if a reader considered writing a library for Haiku. Can't game the system by surrendering a lock before the slice ends anymore. :-B
It's good to remove hardcoded limits, but everything I see in this post looks like it was implemented in Linux and FreeBSD years ago. I'm sure this has been discussed somewhere, but it's not that clear why Haiku has its own kernel at all.
Because it can. Also, because doing things differently is fun. It's API is something to behold, and besides, would you say the same thing about Plan 9, or BSD, etc?
Since when does everything have to fill some niche? What about just trying to make a good open-source desktop OS, for fun, because they're developers?
I see this quite often on articles about Haiku. Everyone jumps on them like it's such a bad thing that they're working on something and it has no relationship to Linux.
I'm not trying to jump on them, I'd love to be able to spend my time on that. It's just that it seems to have significant mindshare, and funding, and usually that means its answering the requirements of somebody.
Maybe they're just here to help diversify the OS world, which in itself is a nice goal.
They have donations, but nobody has ever "funded" the project. Google's thrown a few $k their way ($5k this year) but I can't recall any other big name.
The mission has always been to reimplement BeOS R5 as open-source with some modernization.
I've actually been using Haiku as my primary work OS for a while now. I don't need much software beyond a web browser and a text editor (increasingly true for many people I imagine) and Haiku offers some benefits over Linux in simple shell features like stack & tile window management. Check out the User Guide[0] for a sense of the unique features.
Could you be talked into doing a little writeup on what you've got running on it, any problems you've run into, that sort of thing? I think I'd like to try dual-booting or a VM, but I'm years out-of-date on the Haiku world.
I agree with thaumaturgy, I'd certainly like to hear from someone using it as a daily driver. I think around Alpha 1 or 2 I installed it on a spare box for fun, but never really tried much with it.
No, it has a built-in browser. Not sure what engine is behind it but it reminds me of the Konqueror days. There is no Firefox build that is reasonably current that I have seen yet (hopefully I'm wrong, please someone prove me wrong).
Nobody is saying it's unreasonable amount or shouldn't have been spent, and I consider the money's origin utterly irrelevant. But it is not a trivial amount of money. Unless he moved recently, I believe Ingo Weinhold lives in Germany. His income for this year just from Haiku work exceeds the median German household income. I'm sure he's not working at market rate, but he could be said to be making a living off it.
My one and only point is that you can't say Haiku is just a hobbyist OS. That's simply not a reasonable answer to the question of what its purpose is. It's just dismissive and insulting to the person who asked the question.
I think it is trying to be a simple and elegant desktop system. It requires nearly zero configuration and has a consistent look and performs well.
I get the feeling it aims to be an open source implementation of what the original Mac was meant to be (pre-OSX)... boot-to-desktop with no terminal layer spawning first.
I have been interested in Haiku for awhile but could not get it to install. Last week I finally realized that I could run it in a VirtualBox instance so I started trying it out. I like it... very simple UI, very clean, and very fast.
It comes with many Unix-like shell utilities including scp and ssh. git is installed. There is a graphical text editor and something resembling a package manager that will allow you to fetch vim.
The web browser is a bit lacking/weird/not-sure-how-to-describe, but reading through the installation manual it appears that many Wifi chipsets are supported via BSD drivers.
Press shift when booting if you can't get anywhere, and definitely read the manual. There are few necessary concepts to learn...
Really reminds me of Mac OS 7-9, but much more stable and a better user interface. I'm considering using it for back-end web development (as most of my work is server-based). I'm not concerned about Flash anymore so that is a non-issue.
I've been using Haiku since around 2009, and while I've no experience with BeOS itself (I'm far too young), I adore it. It's such a well thought out and designed system, right down to the APIs. Heck at this point, I'm tempted to use it for all my dev work (server side), as I enjoy using it so much. It lacks the quirks linux has across distros (somewhat an unfair comparison, as haiku has one distro and that's it), while being super quick and a joy to use. It's truly a desktop operating system, which I think is a good thing. It's like an unpolished, open source Mac OS in that way. The multicore and scheduler work being done is going to be handy for taking advantage of new computers.
Now building a desktop OS is a an odd niche. At this point, all of us live on the web, and the desktop matters little, as we have the move towards mobile OSes, and the integration of mobile features into our operating systems. If we ever see a move back towards more "traditional" desktop OS paradigms, perhaps Haiku will see more use, but for now it's community is small, passionate, and fun. I don't think it will ever be big, but one can dream right? :)
The issues of priority and starvation are large issues. I see the OP implemented another ad-hoc behavior, which I predict will move the issue to another place but not eradicate it.
I believe priorities are a sucky way to express the developers desire to perform some tasks in a low-latency manner, while others are ok to delay for a larger time. Maybe latency would be a better thread parameter than priority. It has the advantage of being a hard metric, one you can easily measure and adjust your algorithm against.
But latency is slippery. Do you mean time-to-first-execution, or time-to-completion? Completion of what? The OS would have to be aware of some units of processing (message digestion time?). This is not a bad idea. I'm all in favor of the OS having a more complete model of what the developer is trying to accomplish.
It's a nice UNIX with a cohesive environment and design (both in terms of software and UI) philosophy. Other than than there aren't a lot of "advantages" because its alpha software.
Maybe check out some BeOS R5 videos, they were doing some really cool stuff before they went under!
This niche is currently filled by OSX, but I think it's slowly falling out of favor because of Apple's increasing focus on the end user and "iOSifying" everything.
Of course Haiku is still some ways from it, but I've been following its development for a while now and I feel like it's on the right track. If one or two manufacturers end up providing good hardware for Haiku to run on when it's a bit more mature, it'd be wonderful and then my vision would happen.
There are worse foundations and inspirations to have than BeOS - a man can dream :)