I heard a developer say so on Reddit and I was wondering if this was really the case or not.
PulkoMandy,
I explored further the statistics provided.
Interestingly, the months of July and August appear to have seen the most level of activity. This could be a result of a few Google Summer of Code projects. There is also a high level of activity for January, October, and November - although now obvious reason comes to my mind which could explain this observation.
Anyways, as an end-user, I prefer a stable release to a buggy one even if I have to wait a bit longer…
Hi,
Some of our developers will spend all the time they can spare on Haiku, including days off work during the summer or new year period. I think this explains the january,july and august activity.
October/November is when we have the yearly coding sprint of the BeGeistert conference (http://begeistert.org), where we put several Haiku developers in a room for a complete week. It is a good opportunity to get a lot of work done, as well.
Overall when looking at the data by month of the year, our data set is still quite small (13 years of commit history, so only 13 data points for each month). Maybe the results will even out over time.
There were 11 articles posted to the homepage this month. Let’s see about other operating systems: ReactOS (https://reactos.org/) has 1 news item. AROS (http://aros.sourceforge.net/) last news is from january 2014. And this is just the homepage, not counting all the activity in the forums.
How many articles per month would we need for the website to be considered active? 50? 100? 32768?
On the code side, 2015 is not one of the most active years according to openhub (https://www.openhub.net/p/haiku/commits/summary), but they rank us as "very high activity". At 300 commits per month, I tend to agree. As with any open source project, progress is slow, people can't spend as much time as they would like on it and we would do better with more contributors. But it has always been this way in Haiku and many other projects, so that isn't, by any means, "dying".
http://pulkomandy.tk/stats-201504/activity.html shows a year-by-year view (towards the bottom). We had a peak of activity around 2009, but the levels after and before that are pretty much the same, about 3500 commits per year (a rather steady rate of 10 commits per day). And this is just counting the main Haiku repository, in the current year there has also been a lot of activity on haikuports and several 3rd party apps at HaikuArchives. There has also been a lot of work in upstreaming changes to, for example, Qt4.
There is one real problem: we haven't made a stable release in 2.5 years. But we are working towards that at our usual pace. You just have to be patient.
It seems fair for someone to feel that. If I recall correctly during the first years of Haiku developers were really optimistic about dates for a beta, I remember some kind of progress graph that was almost finished. Plus as it has been discussed the Package manager took away precious time. The website although it is functional doesn’t get updated as much as it should, I read more articles and news in other websites. So for someone that is not very involved is not difficult to assume is dead.
We are not.
The developer in question was me, in this comment. But I didn’t say it was dead, I said it was dying due to lack of developer time.
Ah, okay. Good
Thanks for pointing to the original post.
I found the context quite interesting - a mention of a Google SoC project for porting ZFS to Haiku along with the wish list for improving BFS and a possible approach for this by reusing concepts from ReiserFS.
Many of the items on the wish list appeared to be related to efficiency and implementation could be contemplated for R2.
Whether there is sufficient resources within the community to addresses the data integrity items for R1 is another matter. Even though not perfect, the current BFS appears to be sufficiently stable for most uses.
@waddlesplash… so as active as it ever has been = dying?
That seems to me like you are trollin’ pretty hard there. Haiku has never been super active after all… just slow and steady.
Retro-computing systems tend to die very slowly indeed. If you look at for example OS/2 (now under the name eComStation) there are still enthusiasts keeping things pootling along.
Haiku is a slightly unusual case because its developers started out believing they weren’t building a retro-computing system. If you look back to OpenBeOS in 2001 the idea is that it’ll take 18-36 months and then they’ll have a complete working BeOS clone and they’ll take on the world. And of course that’s not what happened.
Something that surprises retro-computing developers is that the goal posts still move. If you asked an OpenBeOS founder about laptops they might have said vaguely that it’d be nice to run OpenBeOS on a laptop but it’s not the focus. Today of course practically everyone runs Haiku on a laptop. So that contributes to the apparent decay. Something you didn’t consider in 2001 becomes quite significant in 2011 and is a huge embarrassment in 2021 when you still haven’t quite got around to it.
One thing that trips these projects up is that little gaps, things regular users learn to live without, add up until they become a huge chasm of functionality. For example Haiku doesn’t support a lot of built-in webcams, which is a bit annoying, and it doesn’t support USB3, which is a bit annoying, but on a laptop with only USB3 ports, this means there is no way to use a webcam at all.
Only very tiny projects, typically one man band things, just suddenly stop altogether. So that’s very unlikely to happen to Haiku. But yes you can expect a gradual decline, punctuated by seeming changes of fortune that don’t pan out to anything. After many years watching such projects that’s the trajectory they all follow, it’s pretty interesting.
I don’t mean to bug you about it but the way news and articles are handled are not targeted to the average user interested in Haiku, they are more targeted for programmers. They lack some human element, pictures of the people behind them, media from the conferences. Sure right now it would be overblown since the project is kinda short of developers and budget. I was just pointing out the fact that some people can get the idea that the project is dead very easily . Which of course is not.
It has been dead for years now. Around 2010 it really looked like it might make it, then they started working on the package manager etc, and now it looks like it will never happen. They had a chance when Microsoft told all the XP users they were kicking them off, then again when Widnows 8 came out, and everyone saw how ugly and blocky it was. They really needed to make their move at that time. Once Windows 10 is in full swing, Microsoft will do their best to never let anyone leave again. They are going to push people towards the cloud, and then make sure leaving would break your connection. So people wont ever leave.
Seeing Haiku make it was a dream, the idea and people behind it are the best, but sadly, its time for the fat lady to sing. Haiku should have stuck with making a rock solid, super simple, core OS, with no frills, it didn’t need a linux like package manager, anyone can run a executable file from the desktop. Its just like Microsft pushing their packaging system, Windows 95 was fine with installshield, if something didnt work, you just removed it, then they pushed their packaging system, and its been a pain in the rear. Yea its working good 20 years later, but it wasn’t needed in the first place. But of course, Microsoft has hundreds to thousands of people to get it out the door fast, Haiku does not
If they would have stuck with just the OS, just the core, with working media, etc. They could have been done and out there, let other people make a package manager if they wanted it. Let other people make anything they want to. Stick to making it work, making it modular and update-able, having basic drivers, and a fast strong file system, Until you have people using it every day, your falling more and more behind every day
I still wish it all the best, and would love to see it working, but its been 19 years since I first used BEOS, I have grey hair now, and grand children, and I’m pretty sure I will never see BeOS live again before I leave this world, in any way
And if it looks and works like Linux, I dont want to
As a Haiku user (a real one, not someone who keep waiting for it), the package management system is very useful to me. It does a great job at installing and keeping applications up to date.
We are in this strange situation with Haiku, where some people tell us "you should make it simpler, just the core OS, more like Windows 95", and others tell us "you are still chasing compatibility with a 1990s OS, you should go with modern things instead".
Haiku is available, just now. Grab a nightly and start playing with it, and see what you can get done. Yes, there are still rough edges, but we are already far beyond what BeOS could do. And this is why we could continue the work with things like wifi support, package management, and the new launch daemon: bringing modern features to the OS because some people couldn't live without them.
You are right on one thing: we should not have delayed the R1 version to after all this. We should have published it in 2010 or 2012. But we did release some versions back then, does it really matter if they are called "alpha"?
I doubt the "alpha" in the name is what made people run away from Haiku at the death of Windows XP. People just stayed on their XP as long as they could, then they moved to Windows 7. Why? Because they are Windows users. Why would they switch to Haiku? Or even Linux or Mac OS X? That is even more different from XP than Windows 7 is. So if you want to keep as much of the XP experience as possible, Windows 7 is/was the natural upgrade path. Or possibly ReactOS, but that didn't work, either.
Sadly I think in the same way.
Haiku missed its chance back in 2010 or 2012.
There was no Beta no R1… still Alpha all those years…
Sure the package manager is needed…
But I think it would be better to publish a Beta and a final release first and then after that present a package manager.
He is right… the right time to publish a beta or an final release was missed! If there was a beta back in time then it would gave Haiku much publicity…
After the Beta… the final release would give Haiku much more publicity…
and after that the package manager would be very welcomed those days…
sadly Haiku is fiddling around now to polish the package manager and the new boot installer.
Sometimes I think Axel do not want to get work finished!
The right time to publish a beta and\or a final release was missed! Even buggy Haiku should publish a release at the right time…
All developer should try to get the blocker bugs resolved… before adding new enhancements to the system… After that it would be great to work with a stable system…
It doesn’t matter that they were called alpha releases, it matters that they were treated like alpha releases. The ABIs exposed in those releases are abandoned, software which works there is not expected to work in the final Haiku R1, and bugs reported against them are ignored or closed, since the alpha is not in fact a maintained release that will see bugfixes, but instead something pushed out the door and immediately abandoned.
Because they were to be released with the “alpha” label, bugs you’ve elsewhere said wouldn’t be tolerable in a Haiku release were allowed to ship anyway. So as well as the name, you’ve got a serious drop in quality. The diamond-like polished jewel proposed in the early days of OpenBeOS might have justified a 2-3 year turn around for a release, but if you’re going to ship things like R1alpha4 then people will ask why you can’t have a six month cadence. And of course that’s exactly what was proposed (but not implemented) in the post mortem for R1alpha4.
in the beginning he team behind open beos, today haiku, will be a rebuild of beos. and after this is done (this are the plan as R1) new features and modernisations can be implemented. but in the tine of development many new things come to haiku, developers come with new ideas and so the release date switched away and away.
yes i think too that a r1 had to done befor implementing pm.
all thinks added ro haiku in the years are good stuff and work. does not misanderstoud (bad englush). they do many hard worknin there free time and sometimes sone one full time. good work.
i miss some one in the haiku ream who looked at this and deside to make this this and this and that later. a manager to coordinate the work and look at the road map. so not to much new things cones intobtge project before a release are done.
some times i have the feeling that all developer make only what he like and he complete thing is not important for him. haiku is not a collection of unktions and apps. it is a operating system.
i hate the touchpanel of m smart phone 
all over haiku is useable and for me stable. so i can use it.
hope for the future
ps: haiku does not can be a solution for a full windows user
It looks like this is as close to Haiku R1 as we are going to get for quite some time:
A fork of r1a4 that is being called a release.
I thinl i’ll wait for the real Haiku R1.
I agree with all the last posts. The PM was not a mistake but a bad timing desicion, and the solution is not fragment the OS as GNU Linux does. Right now PM is already way ahead to stop but as soon as it’s over the development must take the old route.
The other problem is that the promotion(if any) is being handled by developers and not by advertisers. Is an (GUI) advertising that only a developer would love. The beta-alpha is a very blurry concept but naming the release Beta would draw more attention from the public. Of course the release must be more stable.
The website can and should keep its current style but we need video and pictures, people don’t read as much as before. I would propose a BeOS / Haiku history video that would be released alongside with the next release and an improved website.
My other proposal is an open discussion (hopefully with videos) with the renegade developers that forked Haiku because there has been a lot of drama in the mail list and most people don’t know what exactly is going on (myself included).
There actually were quite lengthy discussions between the team members before starting the package manager work. It was not one dev waking up one morning and deciding “hey, let’s add thsi to Haiku!”.
What happened was a discussion between some developers during the BeGeistert conference, where we decided what should be included in R1. This was around 2010 if I remember correctly, at a time where we were highly optimistic (probably too much) on what could be achieved. During this discussion it was decided to add a way to update the system, so instead of having a "frozen" R1 release, there would be a release to which we could ship updates and bugfixes. This is part of the plan to make a release, a stable one to which we can push updates if problems are found. Once again, the name "alpha", "beta" or whatever is not what matters, and if we had done a release earlier and without this, we wouldn't have been able to provide better support for it than we did for alpha4.
Then, with the bad habit that Haiku devs have of trying to do things right, this simple "update the system" task escalated into the package management system we all know today. And it took about 5 years to get things working. Had we known that back in 2010 when discussing what to include in R1, maybe we would have made a different choice.
But, I still don't think we missed an opportunity with the EOL of Windows XP or "around 5 years ago". Yes, Haiku as it is today would have been nice 5 years ago. But Haiku as it was 5 years ago was not all that good. IT was missing wifi support and still running Firefox 2 as the main browser, and even in 2010, this was laughable. It's just that writing all the code for a complete OS with a team of 20 people or so takes a lot of time. The only thing we can do about that is continuing to work on it until we get more contributors. Or give up on it and start using Linux, but no one wants this to happen, right?