Haiku activity report - April 2020 | Haiku Project

USB boot disk is fine for most cases including post 2000 old PCs. Internet is still required to download Haiku image no matter will it be written on CD/DVD or USB drive. Downloaded image can be stored on HDD/SSD after download.

  • The requirement from the GNU GPL is that if we ship a CD or DVD, we must, for 5 years, be able to shipt the matching sources in the same way if hte person who received the disc asks for it. The simplest way to do this is to just include the sources on the same disc. I could do a two CD set instead, but it would cost more to manufacture and ship. Or I could indeed make sure I stay around for 5 years so I can burn a CD with the sources in the unlikely case that someone asks for it.
  • This affects our own release but also all mirrors of Haiku (they would all need to host the two files instead of one), as well as other people who decide to sell CDs or DVDs (there are several people doing this on eBay, I don’t know if they send some money back to Haiku).
  • I do not expect the DVDs to be the main way to spread out Haiku. To me, they are just a way to celebrate releases and have some physical marker for important milestones. They are also a way to get people to donate to Haiku (when we run a booth at open source shows, the combination of DVDs and piggy bank on the table allows us to receive up to 100$ donations a day). But most people mention that they won’t actually use a DVD to install.
  • You can make your own custom CD images with the sources removed if you want (the easiest way is to use one of the nightly builds images instead). But by doing so, you accept to handle the requirements of the GPL on your own way.

As the person handling the manufacturing and shipping of the DVDs I tool the course that was simplest for me. I ship the DVD, I know that I fullfilled my obligations, I can forget about it and move on (probably to trying to distribute the about 400 DVDs I still have in stock for beta1).

And even if your machine doesn’t support USB booting on its own, you can get around by using something like Plop Boot Manager. And it’s probably a good idea, because running from a read-only CD or DVD requires more RAM for the write overlay, which is likely to be a problem on old machines (I think we’ll be able to install with 256MB of RAM when booting from USB on beta2, possibly not when booting from CD or DVD).


A post was split to a new topic: Translation system question

I’ve got the email address of Georges-Edouard Berenger, the author of ProcessController. I’ve sent him an email already regarding re-licensing it on Sunday, but haven’t received the reply yet. Maybe someone from the Haiku dev team would try to convince him? I’ll provide the email address on request to anyone from the team who is willing to volunteer.

Georges replied and he doesn’t mind changing the license for ProcessController! He’s asking how should it be done? I thought of adding someone from the Haiku’s dev (@waddlesplash, @PulkoMandy, @kallisti5?) team to the email thread for assistance. Or would his email reply with formal consent to re-license the source code of ProcessController from GPL to MIT be enough?


Yes, that would be enough but we need agreement from all other contributors over the years as well. We can easily get it I think for the ones in Haiku, but are we sure no one else worked on it in between?

I’ve quickly looked through https://github.com/haiku/haiku/commits/master/src/apps/processcontroller and most of them seem to be core Haiku devs.

If you can, ask him to send the email to haiku-development@ so it is on the “permanent record”, if not that, at least to Haiku, Inc. (and at least CC me.)

We can handle validating licensing permissions for everyone since. :slight_smile:

Outdated maybe, but it works plus there are DVD re disks for cheap now. Ofcourse USB thumb drives are cheap too. But the hardware I deal with is old from pentium 2 to a core2duo.

Woah,there will be another beta several days later!I’m looking forward to try Beta 2.Maybe I will tell you my experiences testing Haiku on a virtual machine,because I don’t have a real computer for OS testing.Since I tried Haiku Beta 1,I was so deeply interested in such an outsider OS,just better than another two I tested(which are eComStation,the bad one,and Solaris,the regular one).Actually I am an “traveller” among OSes,from old to new,and from regular to unique.I heard that Haiku’s development was slow,but what’s the matter?It’s just careful and hard.And Haiku’s really a better OS than eComStation,So I hope Haiku will become better and better.

Is there a plan to stop adding features? It may be difficult to get rid of bugs otherwise.

1 Like

No. So we may be stuck in beta for a few decades still.


I’d be sad if this happens. I assume it would be harder to keep maintaining BeOS compatibility (the main objective of R1) without the API advancements as time passes. On top of that, because of the ‘beta’ status, Haiku would never get rid of the ‘unfinished OS’ label, and potential 3rd party developers wouldn’t bother writing or porting anything to it. That’s just my 2 cents.

Ok I was (of course? but apparently not) half-joking here.

There are few plans for big new features currently. The focus is on cleanup and stabilization, but we also continue to worry about hardware support (with things like NVMe) and continued support for the ever-evolving web in our web browser, for example. But even the simplest bug fix leads to regressions and more bugs to fix. In fact, that’s even more likely: the NVMe driver has no chance to break an existing system that don’t use NVMe.

In our R1 roadmap, the goal is not a 100% bugfree system, the goal is that what worked in BeOS also works well, and the extra features that were voted in the R1 features poll in 2010. Essentially, the idea is “a system I’d personally recommend to anyone”. Right now that is clearly not the case, with for example a quite unusable web browser.

If we were in a strict “no new features” mindset, things like nodejs wouldn’t have been ported, because that requires additions to Haiku itself. And it’s likely the same for other sufficiently complex applications. So, I don’t think the “we’re in feature freeze” mode is good for attracting developers, at least now. Quite the opposite, we have good relations with developers working on Haiku because they know they can report a bug or missing feature and it will be fixed quite quickly.

Also, note that we will have to break BeOS compatibility in at least one major way before the year 2038 bug, so “a few decades” wouldn’t even be possible. We’ll see what happens, still, but it’s likely we will need a few more beta versions to stabilize things. And remember that R1 essentially means “we’re killing BeOS compatibility” which isn’t necessarily a strong feature announcement. It means the start of a new cycles of large changes, new alpha versions, etc. And meanwhile, R1 will be in some kind of “long term support” mode, getting only bugfixes and security updates. I don’t know how long our users will be using it, given the current userbase behavior I’d expect they would jump quickly to R2alpha1 whenever it’s available?


If I grow up,will Haiku still in development?
(Actually I’m 14,I’ll grow up probably in 2030)

1 Like

Who knows? I hope we’ll manage to release our R1 version, and also that we will have started work on R2 (at least) and still be active by then :slight_smile:


Good to hear from you, Jenny. Like you, I think I will probably grow up by around 2030. Trouble is, I’m 71.

I certainly hope to see R1 in the next 12 months, along with a small number of things that will make other operating systems irrelevant for both me and many other people.

And maybe you will play a part in that.


“Haiku as a lifetime experience” would make a good tagline :wink:

1 Like