When do tou think it’ll come out?
When all critical bugs for Beta 5 have been fixed and when the devs think there are enough changes that it’s worth a new beta.
You can check the open bugs here: Custom Query – Haiku
As the release of Beta 4 was not even a full year ago,I think that it will take at least some months for Beta 5 to arrive.
Tuesday maybe?
We haven’t started cutting the release yet. I’d guess a couple of months still.
We don’t have a fixed schedule
1 Like
Well this part not fully right - some handful of those used to remain and recategorized under next beta version - as them to be resolved requires more effort or time required AS IF the decision about new beta release already done
&
Yepp - finally that’s it. They are right, basically it worths to install a new beta who use the Haiku as daily driver and wants reliability and stable new features/softwares.
Until then you can have always the up-to-date Haiku if you take risks – installing nightly.
This way you can reduce the needless of ticking feelings of new releases from outside world.
I agree some people wants periodic new releases they used to be expected as familiar stuff from Ubuntu and its derivatives in open source world.
Well, personally I would say new Haiku Beta releases would be more awaited if some target/goal would be named and worked on, but this way is also fun.
For Beta5 until now I’ve red more bugfixes than inventions, but it can come from that I do not use actually the softwares those GSOC students or most of developers work on.
For Beta4 the big hits were
networking availability and stability
concluded in all efforts of developers. That was accelerated to be released in new beta – thanks to several portings of drivers, new layers, and sotwares and developing tcp/ip itself. Nowadays they must be very hobbyist and fan of a device/software if finds a computer/application to useful/good without reliable access of the net (and its services).
So access and widely usable browser is a must.
Of course, if I think better or more deeply, Beta5 maybe - besides many bugfixes - will be Big Hit for developers actually … not for end users.
Also many forum posts about several developments these days…
Italian group created a ‘native’ Haiku IDE - development environment
One people write a software that would finally use more seriously the BFS filesystem’s database-like capability with searching / indexing file attributes- this way storing data and their connection among them (all regular daily things like books, recipes, etc. not rocket science)
Developers dispute on new drivers’ design how and what extent they use the same file attributes in driver tree build/ driver registration and so on.
GSOC2023 .NET development porting – also for devs primarily.
When someone browse the HaikuDepot they can find many SW ported that just do nothing that enables much more games to run emulated. Ok mostly for 8-bit home computers and 16-bit stuff and consoles but way more than other like general virtual machines. Just right now a so called uxn appeared as ported to Haiku. You can write small size condensed apps in an assembly - like language. So Beta5 might be for devs mostly.
Except if one another platform image those under development - Risc-V, ARM - can completed to supported level. WOW, that would be really big hit !
So, dear @AetherealMeowstic
think about words of Gandalf what he said about arriving of wizards, and you have the quite right answer - in this case too. 
EDit #1 : fixing typos - mostly. Adding words - rarely.
edIT #2 : supplemental addition ;)) of left out thing(s).
1 Like
What’s the release plan for Beta 5? How does the system stability look nowadays? @waddlesplash, do you have anything on your list significant to work on before Beta 5? A “State of Things on the Road to R1” blog post would be nice and gain a lot of interest.
1 Like
well, how about this way ：
developer do developping.
application programmer do programming.
normal guy use daily builds and share anything of Haiku 。
when the right time, it will be r1b5, even R1 or R2.
Also don’t forget:
Users with lots of money donate to hire more full time developers to boost development speed.
3 Likes
There’s an R1/beta5 milestone with a bunch of tickets in it. I have a few other minor items on my own list but they’re not quite as important, I think.
Not sure why you say such a blog post would “gain a lot of interest,” though. The state of things is just “we have a lot of bugs to fix”, mostly…
6 Likes
Thanks for the reply; I am aware that there is a Trac milestone, but most of those milestone tickets usually get removed from the milestone or assigned to the next one before a release. 
1 Like
I can’t login to trac,but that ryzen spinlock issue is probably the same bug the bites linux users, by the look of the addresses involved.
Ticket 18444 in trac
https://bugzilla.kernel.org/show_bug.cgi?id=196683
My take, is that there’s a interupt that’s clearing/overwriting /moving the cache data and the pointer breaks.
But I’m not a cpu microcode developer. Windows doesn’t seem to have this issue, so it’s something in the wat Linux etc work .
Not most of them, but a few remaining ones at the end (always the same from previous releases )
For me the main blocking thing is updates of libraries. I have already started the work on migrating to ffmpeg 6. I also plan to look into ICU. There are already newer versions available in haikuports, and, in the case of ICU, it is required by current WebKit versions. So, if we update WebKit in the current situation, we would have to put two different ICU versions in the release 
Unfortunately, there is a bug in ICU integration that prevents Haiku from booting with newer versions (probably something like this: we use ICU to implement the C library locale functions, but sometimes ICU tries to use the C library functions for some things. And it’s possible to end up with an infinite recursion.
I also have some less important but still nice to have features, for example I want to finish restoring native button look in WebKit (the code got lost during one of the previous updates, I have already fixed checkboxes and radiobuttons but not yet buttons).
If you have questions about why one of the other tickets is currently in the beta5 milestone, feel free to ask (here, not on Trac to avoid offtopic discussions in ticket comments). Maybe some of them shouldn’t be there and can be moved out (I have not looked recently, I do move things out of beta5 from time to time but maybe it’s a good idea to have an open discussion about it).
I think at this point we should not be adding anything more to the milestone, however. Unless there is a really annoying hardware compatibility bug with an easy fix.
11 Likes
I think that beta5 should be abandoned and only rolling release should be done. Maybe a little better testing is needed before publishing the iso image (or only monthly, better verified releases?). Maybe this way Haiku programmers would save time and effort? Still, I don’t think anyone uses the official beta releases for any serious work that isn’t related to testing or developing the system itself.
…But maybe I’m wrong, what do you think?
1 Like
doesn’t seem smart to me to abandon useable releases 
4 Likes
You know how assumptions go right?
I really like the idea of “rolling release”.
Actually I’ve been thinking about it for a few days now, and I was going to comment it in the Spanish Telegram group.
Why are there repositories for each beta version, or stable version? Why are there repositories for native programs and for ports separately?
I think everything could be simplified with two repositories, “testing” and “current”. Many Linux distributions do it this way. Even Microsoft has a similar channel for Windows beta releases.
Something similar happens to me with HaikuPorts. The first thing I thought when I saw it for the first time, was that it would be something similar to FreeBSD ports, but it wasn’t. Wouldn’t it be easier to have one repository for binary software and another for sources? As it happens in FreeBSD or Gentoo Linux, it could be simplified a lot, using a program that automates the download of the sources, the configuration, the compilation and the installation.
I use 64 bit beta4 (with UEFI) on my main desktop PC and use it daily.
We have that, they are called “nightly” and “r1beta4”.
We have that, it is called HaikuPorter.
7 Likes
Given that Haiku is run often on low-end and older devices, it is recommended to avoid compilation as much as possible. No amount of automation will make source-based packages equal to or less intensive than installing binary packages, which is why both FreeBSD and more recently Gentoo have binary packages available. Particularly notable in Gentoo’s case, having previously been a source-based distribution with only a few binary packages (mostly web browsers) provided.
Then there is the question of managing multiple repositories. Should the source repo be almost completely unmanaged like Arch’s AUR or managed to the same standards as the binary repo such as Gentoo’s GURU?
The first option will lead to many lower-quality packages and a larger possibility of malware slipping through; while less of a concern for a relatively niche OS, Haiku does run as a single-user OS (despite being multi-user underneath) which allows both unrestricted user data access and package management. Theoretical malware will have no trouble messing with everything in /home and doing whatever with the package selection. All these can also happen with badly made packages, to; perhaps even more as people would prolly be less likely to question why there are package change confirmation dialogs appearing all of a sudden, after installing a package.
The second route will lead to higher-quality packages, without many of the issues of the first option. However, the problem is a matter of people. It will be effectively doubling the workload of existing HaikuPorts contributors. The time and effort spent could instead be used for maintenance and getting more packages into HaikuPorts. Additionally, the primary rationale for source-based packages is being able to make changes to the software being built yourself. However, Haiku is a desktop-oriented operating system. Its intended audience (desktop users) doesn’t generally know about customising builds and shouldn’t be expected to. Now if more people are willing to step up and manage a source-based repo, that’d be great! However again, would their time and effort be better spent on improving HaikuPorts as it is?
Assumptions and their environment are investigated and verified. Tats are “go right”.
Have you tried using “nightly” instead?
Be that as it may, Beta4 is not a stable release, and neither will Beta5 be.
Those “useable releases” are beta quality software, the same as “nightly”. Haiku does not have a stable finished version. Don’t let the name “release” fool you.