I*'ve been hearing from a number of people that one of the problems with this project is that there is no proper Official roadmap or focus.

What 3 areas do you think Haiku should specialise in,

Where exactly should Haiku should be going?

Here are my ideas:

HAIKU as a Power Platform

(in order of priority)
  1. Low Level Development Platform for Games in C and Electronics also boot from USB Server Management and Tools
  2. Light Efficient Personal Computing platform with a Powerful Extensible GUI for old and new hardware-> away from the corporate consumer and bloated adware market, but much easier than linux to use for domestic users.
  3. Multimedia Experience Platform <- True to what BeOS was originally intended for.Plus, if you can make an old netbook run as fast as a macbook nowadays, you might as well use it for the same things , like watching movies, listening to music etc, music production, and video editing.
Feel free to criticise my suggestions or say whether you agree with me.
1 Like

it’s a media platform in its bones, it’s the part of the platform other systems are still playing catch-up with (osx with syphon, *nix with gstreamer etc), if not in implementation then in ethos. where haiku works for development is either native as an extension of its design (not very portable) or as a result of posix compliance, where it’s already outclassed. play to your strengths!

I’m partial to #3 because I’m a BeOS enthusiast, but #2 is good also. But, I’m not sure how attractive a “media OS” would be these days since CPU speed, RAM amount, and FSB are insane compared to the late-90’s technology. Is there much of a demand for an alternative OS for media today, when we have multi-core i7 processors, massive gobs of RAM, and crazy-fast FSB?

The main thing what i like Haiku Os for is it consistency!
Main usage of GUI for configure it.
Also how good Programms work to gether (drag and drop support and so on)
The Media and other things are not at all at a leve where other OS are these days.

Let’s make it as tiny as possible. Everyone forgets about the people who live in poor countries and do not have access to lots of RAM.


Let’s face it, there was not even a demand for a “media OS” back in the BeOS days. If there was, Be, Inc would be a top 100 company now and we’d all be running BeOS R15.

Which leaves us with option #2. I would be cautious about emphasizing the retro aspect too much. You don’t want to fall into another tiny niche. But if we present it as “Runs well on old hardware. Runs insanely fast on new hardware.” then we may have a chance of attracting people’s attention.

1 Like


And let’s sell it in shops on a USB stick.

Image and video hosting by TinyPic


I just see haiku as a desktop OS done right, to me haiku embodies the “Do it once and do it right” philosophy. The design choices all seem logical to me, there is just the right amount of flexibility, and so many unique features where things have really been thought through with much greater attention to detail than other OSs.

Unfortunately where it falls over is (1) hardware support (2) not being finished (3) not having enough relevant applications.

no hardware spec makes up for the horrendous video workflows currently available. for consuming media, anything will do. producing it? we need this.

1 Like

For me is Haiku a OS (Kernel+Userland basics) (kernel/lib/server/kits) for the Desktop, this means not to port any stuff to haiku (java, python, qt, etc.)

Haiku development is vor a long time broken, in the fact haiku inc make not a good job, only the community give here best.

A goals for haiku is missing, a leader is missing since Michael P is go out.

in my Eyes, the best way to fork Haiku and re think about a OS, Kernel? Compiler and libs? etc.

I think we need a Biz for an OS, for me, we need a free modern OS for the Desktop/Laptops. Unlike M$ and here Dinos.

But we can no do anythink, but a stable base, 3rd Party Apps, like a Browser or Photoshop clone cam from here.

SteamOS are a good idear but this is not radical, he adapt debian with steam client and the same situation as M$, Ubuntu are the same as M$ and so on.

GNU OS, hmm is miss modern like Haiku(i mean beos), beos says, we make a multimedia os and it is simple to used and are posix ready, this mean we ar not a unix.

The fakt are, we need money and people he have the same vision and that is not more in haiku.

Making a OS thats mork with modern Hardware (x86_64) with a modern tool set (gcc6, golang, Vulkan) by a minimal Dektop stuff (Windowin,and App handling).

I personally never stopped saying “when it’s ready” (I’m sure you’ll find some quote to prove me wrong!). And I’m still working on making things happen.

Congratulations to SamuraiCrow for getting NoHaikuForMe to attack something else than Haiku, for once!



let’s face it, there was not even a demand for a “media OS” back in the BeOS days.

I think there was a place at that time for a compact, powerful OS. I remember that
one of the companies here (Toshiba?) declared they would deliver PCs with double boot,
windows and BeOS.
But this was killed by MS people threatening the company to terminate the discounts
on the OS. Exit BeOS on big players’PCs. So this means that even MS was afraid of BeOS.
And this was also confirmed by MS mail leaks.

Now as you said, “let’s face it”, there is no critical mass, and I wonder whether there
will ever be.
When the Haiku project started (2001?) formerly as open beos or something like that, I
thought it would go the Linux way, slowly picking up power. I thought (rather naively)
that it would take a year or 2. but I have the impression it is simply stuck.

I tried once to boot it on my PC. As it didn’t work, I asked on this forum what I should
do, and I was replied that it cannot boot on such a modern PC. I don’t know whether it’s
still the case, but if people cannot have a boot partition sowewhere, I wonded how it
will pickup.

There are a few technical choices that made sense at the beginning, especially binary
compatibility with executables. Now is it reasonable to invest that much effort
to make a new OS compatible with applications that were created… 15 years ago
for the latest ones… If these apps are not supported anymore, this doesn’t make sense.
And if their creators are still alive, they can recompile them anyway. I would reused
the APIs which were good (except the media kit) and I would go for something new.
With the latest GCC.
It used to be a new OS with no legacy, now why are we keeping artificial legacy?

Well, I’m aware that what I just said might have evolved, but let’s face it: many people
are tired to wait. Existing OSs are possibly not as good, but anyway better than nothing.
The fact that is there isn’t (and has never been) any roadmap is possibly one of the
problems. I remember that 15 years ago, all people asking when it would be delivered
were replied “nobody knows”. Now question: when will we have a functional version that
can be booted on a PC, call it R1 or V1.0, whatever? In the next 2 years as I thought
a long time ago? Or maybe not in 15 years from now? I hope I will be proved wrong,
but in the meantime, there is no reasonable hope, no promises, no hint whatsoever.

In the meantime…
Here is a pesonal development. You may recognize ST’s development board. It captures
audio for recognition.


1 Like

Why we like Haiku?
'cause it’s simple.
'cause it’s done by engineers, not by modern f***ts who read «HOW TO DEVELOP YOUR OS IN PHOTOSHOP».
'cause it’s working.
'cause it’s has good architecture.

We don’t need:

— golang, vulkan, kokokotech and anything f***ts-modern
— absolute x64 anywhere just for getting ability to use more memory than needed
— stupid fanboys who talking that s***t and did nothing.

Seriously, kids, what you lost here? Guys did system from scratch, but everyday one kid comes here and say «I want Windows» or «I want my GAY/Linux», or «Where’s my Apple Photoshop OS».
Wanna Windows? Install Windows.
Wanna BSD/UNIX? Install it too.
Wanna GAY/Linux? COCOCOCompile it.
Wanna $teamO$? Ask at another place.

All what Haiku — and heroic team behind it — needs is your help. Code, money, advertisement, bugreporting, help on forum for another guys at least. Not stupid wet wishes.

Please help project or at least do not interfere.

1 Like

Totally agree.

I do not agree. I think it is too extreme and simplistic approach… And what is this “we”?
…In other words, personally I prefer cooperation, not domination in the community.

1 Like

It may possibly take the path of Ubuntu? - Start the operating system for schools - if at multiple points at the same time to start - and to share experiences, - to tie knot in the “joint” movement stronger, look can break in a serious project?

Plus selling flash drives :), then you can look at the creation of the control center with paid services …? The operating system needed because the maximum investment, design, optimization …

This is Tiny.

Poor countries first need drinking water and electricity, but that is headed their way faster than can be imagined via graphene development. So are handheld computing via the same technology, that will be much more powerful than current computers and using little energy.

path of Ubuntu

There’s not enough pathetic. Try this


Of course, there’s nothing good which can be said about linux or smth based on it.
At Russia I was saw a lot of attempts of linux integration. And know what? Every damned time it falls.
And it’s good when bosses realizes the problem and stops «integration». Even worst when they trying to continue till the end — in the name of savings.
I can remember every place with tasks more important than website hosting, where bosses said «We don’t want Windows or something else, we want linux 'cause we heard that it’s free». Not Ma€, not one of UNIX/BSD, not QNX, RISC OS nor smth else. They wanted «free linux».

Result — few businesses gone due to competition fault — you can’t stay on market if instead of work you wasting time and money on Gayoogle/Linux eternal integration.
And at place where I studied is now teaching «how to install linux and php+mysql and create your website» instead of old tech lessons — which had been good, with calculations, simulations and diy solutions, but — oh shame! — on old Windows and UNIX systems.

How ironic — «full of free choices» linux giving you actually no choice — you must waste your time and money on it or pay to f***t «geek» (abbreviation of «Google Eggless Evangelist Kid»; someone who is actually near zero in tech theme, but trying to swim near and imitate knowledge) just 'cause instead of learning smth he learned how to advertise, install and maintain linux.

Another cool story about GNU/Rooster is about integration at schools (and government too).
Ever having ability to buy Windows cheap, even having ability to develop own (or continue to develop Soviet ones) systems, which — Russians have good skills — can be more than enough for all purposes it always comes to this. Everytime when someone at government wanna fast money, he comes and say: «hey, I heard that we can make our schools more modern and economize lot of money — we just need to integrate linux». Because there’s no real engineers at government, just businessmen (as one’s from O’Henry “Strictly Business”) — all ignore that linux actually have a dirty architecture, lot of bugs, more expensive summary costs and — final shot — done not by enthusiasts, but by main US companies, and the more of them are known as US government hands inside an IT. War between MSFT and, as example, RedHat is only in a geek heads; when there’s an question about your money or important data they working together without problems. Invisible hand of the market, huh?

So, everytime when non-engineer fools attempting to integrate low-quality system developed by potential enemy I see a chain reaction of a geeks who celebrate «Ura, we’re now become more modern, more open, more democratic, more happy!».
These f***ts don’t see that linux integration — ends it with a success or as a unsuccessful attempt — gives nothing good, only higher costs for budget and money theft — for a good purposes, of course.
No one cares about children, only money and their GNU/Rooster religion important to them. Even more, part of money will be used to pay reporters for needed news headlines — «XYZ switched to linux, see how good and modern there and how kids are smiling. Smile at camera, Peter, show how thou art happy».

So what you mean as «path of Ubuntu»? Path of government-level fraud, corruption and kickbacks? Path of low-quality old principles based data-stealing system from NSA? Or path of salesman-sect hybrid?

As I said, good linuxer — dead linuxer. Sometime you’ll understand it.

About souvenirs.
Selling T-shirts and USB drives never, never helps. It only good if NSA wanna pay you for work, but don’t want to affiliate any connections with a project, but if smth like this will happen with Haiku — I’ll go out from here as any human with brain.

So please again, keep your wet wishes away from here.

too extreme

If truth is too extreme for you — please don’t read. Who knows, maybe one day even letters will turn too sharp and hurt or even kill you.
Read something optimistic with low containing of truth — pics with cats, rainbows, unicorns etc.
Good luck.

To me armanx64’s rant borders on hate-speech and I hope in the future he’ll at least try to avoid the anti-gay rhetoric. I hope nobody mistakes his attitude as typical for the Haiku community. From my experience most of us welcome diversity at all levels.
Anybody can present their “wet dream” for Haiku, and maybe some people like to discuss those. Just don’t expect the Haiku project to shift their focus willy-nilly. You have to make a very persuasive and well researched argument to influence people.


1 Like

I’ve worked with AROS as a hobby for a while. It has old 90’s level stuff ported from AmigaOS mostly. It’s main advantage of the fact that it can run hosted on other platforms has been squandered to the point that it’s more common to run AROS hosted on top of the Linux kernel than it is on its own native kernel.

The relevance of Haiku is that it has SMP. The AROS kernel is stuck in the single threaded mode inherited from the Amiga. I wouldn’t mind seeing AROS run hosted on top of the Haiku stack. On the x86 platform, however, the 32-bit instruction set sucks. Even using 32-bit pointers within the x86-64 instruction set is preferred since it frees up registers for general purpose use. If we’re stuck on the x86 in 32-bit mode, we’re sunk as a low-level architecture.

The Amiga always used custom media hardware to accelerate itself and had 2 memory buses, one for the CPU, one for the multimedia chipset. The media platform ship sailed with the Mac and the downfall of Commodore. Now that even Mac is declining, another angle is needed. The modern Amiga-like platforms have a Lua-derived software VM called the Hollywood Multimedia Authoring Layer. See for more information on that.

Lightweight computing is cool on ARM machines. It doesn’t go so well on x86 because there is more software for lightweight Linux distributions. If we could target the RasPi 2 and 3, we could probably kick Raspbian to the curb.

I think the 3 angles presented by the original poster represent occupied territory by the competing software stacks.

One angle I’d prefer is something portable to many platforms like AROS but use existing driver stacks. That would make it possible to shift resources to cheaper hardware platforms like the ARM-based systems intended to run cheap Linux distributions. AROS has a skunk-works project called Arix that tries this approach. There is a policy of not announcing anything until it is ready so I don’t know how Arix is going.

Another option used by Amiga-like machines is FPGA-based hardware. This allows custom hardware to be developed around one set of specifications to ditch the need for drivers altogether. This preserves a monoculture but has only recently made the old Amiga-compatibles fast enough to run Haiku on their 680x0 instruction set systems. Only if the bugs are fixed and massive funds are raised, will a true ASIC SoC chip bring performance up to modern standards.

I hope my perspective as an AROS hobbyist brings you some insight into your own angles and approaches. My philosophy about BeOS and Haiku has been that they are the best operating systems with practically no software running on them. I hope that the emphasis of Haiku can shift off of the OS someday to bring more application development to light.