Disappointed with Haiku

I’ve heard of the Haiku project for years on slashdot, but only finally downloaded and tested it out within the past year. My first impression is this OS might be the answer to all my prayers. Wow! Finally, a fast, light weight, nimble, usable OS which is designed intelligently. Unfortunately the damn thing is full of bugs. It’s not usable as anything resembling a production system in its current state. Disappointing.

I downloaded the latest version of Haiku the other day along with the source code, installed both in a VM, and tried to compile the source. I was assured that the default Haiku system already has everything needed to compile the source, but this is a lie. Upon running the compile script the first thing it does is start trying to downloaded dependencies. Well, my dev machine is not connected to the internet nor will it ever be; MAJOR FAIL. Few things infuriate me more than the common stupid ASSumption by developers that the internet will always be available. The required dependencies should be kept in the source code archive, or at the very least, in a separate archive which can be downloaded from the same place as the source archive. The compile script doesn’t even realize it has already downloaded certain files and it tries to download them again. Stupid, stupid, stupid. To top it off, after having installed these dependencies (after a few days of waiting for internet access), the source code won’t fully build. There were 3-4 modules with broken build, one of them being the debugger.

Between this sort of frustration and the fact that the damn OS has not seen a real release in almost 20 YEARS, how the hell can anyone ever be expected to take this project seriously?

I’ve been reading the threads with people arguing back and forth over the future of Haiku. It’s clear that this project has no leadership whatsoever, no realistic plan or roadmap. Adrien aka PulkoMandy, like so many other open source developers, seems to be too busy thinking about himself and his own desires and not the needs of the community. If he was really thinking about the community then he would immediately pause all new feature additions, start fixing bugs, and keep doing that exclusively until ALL outstanding bugs are fixed, then release a beta to serve as a stable platform for developers and users. Unfortunately I don’t see this happening, because the developers don’t really seem to be listening to the community. I would like to use Haiku as the basis of a very special project, but unfortunately, the OS is useless to me at this moment. This could also be why for example certain major app developers (I’m thinking of one in particular) have left the community recently, which is likely the beginning of a trend. Feel free to email me when you people get your s–t together. I would love to give Haiku a real evaluation and begin work with the code base, IF there ever is some kind of release.

Who are you actually?


I’m exactly who I said I was: a person who discovered Haiku within the past year, and has been frustrated by multiple attempts to use it. Who are you, actually, and what does your post contribute towards improving the dumpster fire that is Haiku?

Yes, but as you write it you are just as well set up? What kind of project would you have thought of that would be especially good for the community?

I can understand that you are frustrated, We all have our wishes and thoughts that are not on the same level with everyone else.

Unfortunately, pulkomandy is repeatedly put down as “the evil”, while he is just one of the few who ever dealt with the forum and the users. He tries to mediate and to reflect the state of development.
Thank you pulkomandy for your help so much times.

Haiku is not a project that has planned and pushed by one person alone. It is a community project, born from the desire to get beos alive. Every developer has his or her own ideas and wishes for the future. Therefore, there is no “head” who can feel personally attacked or persistence of a boss who drives the others into a thought-out line.

If it is a problem for you to walk your way, make your contribution, or realize your ideas in this environment? Then you should think about turning your back and taking other paths.

Again and again people come here in the community, which I have actually met for the most part as very friendly and helpful, and make everything bad. especially because their own concerns are not met, but they see so much potential in the system (why do most of their contributions make this system always bad?). Also I have had many claims in the beginning, at times of beos, and this in the forum announced, but then quickly realized that I do not get on with it. In the end, I discovered the programming language yab for me and then helped myself at the end. I have learned that the things that I feel as “very importand” are not necessarily the same thoughts of others. now i have been part of the community for 18 years. I have found a field of activity for me, write instructions and program (mostly for myself) in yab.

1 Like

Actually, I can’t think of any OS besides the big three that is in a better state than Haiku. Look at AROS, KolibriOS, ReactOS…

I do however agree about the assumption that an internet connection is always available. That assumption shouldn’t be there as it can easily make a freshly installed system impossible to use.


At beos, the release candidates were packed with programs that could be selected during the installation process or later installed by the cd.

I think that has been written more often, it will later be the same with a release candidate at haiku!? At the moment you do not need anything that is not absolutely necessary, because it is a pure developer environment. That you can use it is already great, but not self-evident.

It’s not nice that everything has to be somehow connected to the internet, but at the same time it is the fastest and easiest way to provide software without having to tie up a whole package at every “nightly” build.

Never forget:

  • Haiku is in the alpha state
  • usable like a beta version
  • in development

A fresly installed Haiku system don’t need internet connection.
We don’t make that assumption.

But we do it when it comes to build an Haiku system.
Which is not, really not, the way we promote to install Haiku for end users.
Really really not.

If someone can show us where did we say otherwise, we’ll be happy to fix it.


Lack of the “head” is the problem for the whole Haiku project, if our community ever wanted to see the R1 in their lifetimes. If the current status of the project is what everyone wants, then yes, let’s not make a big drama out of it, just all walk our ways, many of the community members from the 2000s already gone.

On the other hand, if one is not happy with current Haiku leadership, then fork it, build a better bigger developer team and write better code (if you can). Thankfully, Haiku is under MIT license.

You can’t install any applications in Haiku without internet. This makes a fresh Haiku install unusable. While it’s possible to save an individual package locally, it’s not like in Windows where everything an application needs is usually included. You have to also get all the dependencies.

1 Like

[Sounding like Steve Balmer]

Dependencies! Dependencies! Dependencies!


But, in reality, because there is no actual leadership and no “one direction” that everyone follows, it is quite likely progress will continue to be a case of so many steps forward and so many steps back. It’s just the way open source projects work. But at least there is more forward momentum than… say… Minix3.

There once was a “head” to the Haiku project… but he left. I think once he accomplished his goal of getting the ball really rolling, he realized there was nothing more he could do and the project had taken on its own life, beyond his control, and so he had no more purpose. So why stay? His vision is why we have a Haiku today. So we should be ever grateful to him. But, the flaws in the scattershot method of development is why many complain, when they see how much time has passed and where Haiku is today, in relation to that amount of time.

When Be, Inc. started BeOS, how long did it take them to get to R5? How much money did they burn through, to get to that point? Could Haiku have gotten to “R1” as quickly, if they had a single “fearless leader” and had more financial resources, and everyone followed the lead of one man? Unknown. Because that road was never taken. Or maybe it never could have been taken, because that’s not how open-source projects work.

We may never get what any of us “want”, but is “getting there” half the fun or worth the journey? We don’t have much choice do we? If you can’t captain your own vessel, you can only go where others take you.

1 Like

Be Inc. was founded in 1990. BeOS R5 was launched in early 2000. So, a bit less than ten years (this is the 17th year of Haiku development). Their burn rate will be have been of the order of $10M per year, lower at first and ramping up. $100M total to get to BeOS R5 is in the right ballpark.

We don’t have a clear idea of Haiku’s income because Haiku Incorporated is a bad joke of an organisation. The last time it was able to cobble together an actual financial report Haiku was bringing in about $20k per year. I would assume it’s now far less, but who knows.

But Be Inc. spent a lot of money on other things, not only the salaries of dozens of staff in the San Francisco Bay Area, premises, and so on, but also money for the BeBox development platform. This sort of platform is intended to drive developer uptake, you could think of it as a sort of “loss leader” - but because Be had already begun to attract the type of rabid fans who stick with Haiku today it found that many purchases were by fans who thought they were supporting Be Inc. by taking up precious engineering resources and capital expenditure.

1 Like

You name some specific people in your post, so I’m interested in what your real name is. Otherwise Haiku probably doesn’t need you. Please also stop using caps if you can…

Ok, haiku development started in 2001, today is 2018. How do you get to 20 years?

1 Like

Yay, personal attacks. Thanks for your constructive feedback.

You know what? I don’t take Haiku seriously. I do this for fun. I don’t expect anyone to take the project seriously. It is developed by a handful of random people, using a 20 years old compiler, on their free time (which is limited) and with the silly goal of replicating a 20 years old operating system, which is completely irrelevant today. As one of our contributors once said, “Haiku has no future”. And this is the nice thing about it. It means we can do things for fun.

Still, let’s look at my activity in the past weeks. I helped one of our users getting his touchpad to work. I spent hours investigating problems in the web browser and trying to get website to render properly and without crashes. Clearly, I’m the one who keeps adding crazy features everywhere and I’m not listening to users.


Haiku has everything it needs to be installed and running (the OS). Any additional software is available with internet connection (the Packages). Any update is possible with internet connection. And this situation is not different from almost all other OSes existing today. Windows, macOS, almost all Linux distributions, all sorts of BSD, all different OSes based on Illuminos kernel (Solaris) of and almost all kinds of less known OSes require internet connection for updating the system and an online software repository.

Most Haiku users never compiled Haiku from sources, unless they found a bug in the source and know how to fix it. Your attitude, @gladius, doesn’t seem to be the constructive towards fixing the bugs, rather complaining about existing bugs. In this case, you really do not want to compile Haiku from itself. Rather you can file a bug report, if you want to see some bug fixed. Unfortunately, reporting a bug also requires internet access.


Haiku makes you angry, impatient and lash out at developers. There are lots of avenues for support for Haiku, but it looks like you’ve already gone beyond that point. So it seems like it really isn’t for you :frowning:

If you want to be contacted when Haiku R1 Beta or Haiku R1 will be released, please specify your email.

1 Like

I imagine they got to “almost 20 YEARS” by rounding up the same as anybody else. It really has been a very long time.

Actually, most of the income is from Google funding our efforts with GSoC and GCI, so I don’t think it did get much lower. You can check the yearly budget up to 2014 here: http://haiku-inc.org/donations-analysis.php, and indeed it is around $20K per year.

Calling the organization “a joke” is a bit harsh for the people trying to keep it running. We get even less contributions there than for the development of the code, and as a result some people end up being sysadmins, contributing to core part of the code (such as UEFI support or Radeon drivers), AND doing the paperwork for the inc. All of this on their free time and without getting paid. Yes, there are problems with the way things are done. But still, they are doing their best, and while there is a lot of people complaining, there is still few of them offering to help. So if you can write a financial report and have time to dig into this, maybe join the Inc and help get things back up. It will free up time for devs to contribute in other places where they can do what they know best :slight_smile:


That’s just the way Haiku/Porter team decided to go forward, the Linux way. While I respect their decision, I personally don’t get too precious when it comes to disk space (these days there is plenty) and more in favor of macOS way, with their “fat packages” or app bundles or however it is called today.

I believe Haiku package management system has the support for such packages with “all included”. You need a separate server and another repo with a new set of completely independent “big” packages, that need no dependencies. So it is possible, just not the preferable method of doing things, according to TPTB.