Just Another Friendly Introduction

Time-appropriate Greetings, Everyone!

Just figured I would start a thread to say hello, and try and mark a personal milestone in getting involved in an open source community I’ve only been passively watching for quite some time :slight_smile:

That said, I wanted to open up a general discussion regarding the current state of the community. There’s plenty of threads I’ve seen regarding reaching Beta status, requested software, etc. but nothing really putting a finger on the pulse of the people

Has your impression been a growth in users since you first joined? A slump? Is it your opinion that the community guidelines are being well-enough adhered to, or could moderation use a little boost?

Would just like to hear some thoughts on the matter, and looking forward to diving in and really doing some good in this, an exciting new endeavor!

All the best! :slight_smile:

Welcome officerwafl!

I guess over the lifetime of the Haiku project, it’s fair to say that the community has shrunken over the years. More and more people found a new OS after Be Inc. went under and it became clear that Haiku wasn’t ready in time to be its successor (in a productive manner).
Add to that the ever improving Linux desktop and the result is what we have today. That said, it doesn’t mean that some people won’t come (back) and spent more and more time in Haiku, should it become viable again.

What I do experience is that there is a steady trickle of people checking Haiku out. Some leave again immediately, but some also stay and join the fun. I think esp. after the latest article on Haiku’s vector format, there’s been a slight bump in newcomers. Many express their surprise how usable Haiku actually is. We do suffer from a lack of enduser applications, however. IMO, while having a beta would be very welcome from a PR perspective and to gain more users and possibly devs, getting more (native) applications would be more important to me… The old hen and egg conandrum…

Welcome to the Haiku family!

Her’s a little thing I wrote on my blog in 2014 that I think is still relevant. http://clasquin-johnson.co.za/michel/haiku/blog/2014/why-you-should-be-writing.html

Haiku is in alpha stage and not a complete releaded system, so the developers does not spend time in much user support.

Yes here is the forum and a mailing list, you can find tutorials and help from many people, but there is no User assigned person i the haiku team.

The beta version is called many times, but never released last years and I doubt that the next time will come.

I does not want to make Haiku and the Community bad, i see it from a realistic perspective. Many have migrated or haiku for the time being put aside, because the beta version so far has not released.

I also read a lot of discussions about the community and also this is
discussed again and again about the “most wanted” programs. Many of the new ones arrive and see what they can do with Haiku and fail on things like an office application. Then come the requirements for the Haiku team that these must provide for a solution and forget that this is not their task. Here, finally, a system is developed not an end user system interface.

More should be done to enable third-party providers to create such desired programs.

Welcome to the commnity :grin:

1 Like

Hello and Welcome !!! :slight_smile:

I’m following the project since the BeOS time. I dont know if the amount of members are growning up or not, but in any case, I always found great people here, and always open to help to everyone.

Welcome aboard! :smiley:

Excellent post, and terrific blog overall, I must say! :smiley: thanks for sharing!

I was originally under the impression that more devs working the core OS was sorely needed, due to still being in Alpha status. The more and more I lurk around, it seems userland developers are really what’s needed.

In that vein, I think I’ll focus more of my efforts and getting some good stuff ported in!

1 Like

Thanks for the warm welcome! :smile:

Thanks very much for the reply, and your insight!

The userbase is in decline, although maybe it is starting to plateau out as the numbers dwindle - the core (devs) team, some oldtimers from BeOS era, donators, few newcomers who for this or that reason stick around. Most newcomers probably come and go as they see that the last release is ancient, and see all the pleading for new one in otherwise often eerily quiet forums. What will probably happen is that at some point a business entity will see potential in all the work done and use Haiku codebase for something, like Google used Linux for Android. And as the commercial product will maybe gain popularity some will come Haiku’s way but it will remain in the shadow of its commercial progeny. Some Haiku devs get hired by the business entity to work on the thing they already know, and know best. And when devs go… Newcomers will mostly flock to the commercially developed product. Haiku will become quiet, yet maybe not dead. Occasional updates to codebase might still appear. Some donators and oldtimers will go, some will remain, and complain, and tell stories about the old days.
Sorry for typos and being blunt but it’s 2AM and I’m tired :stuck_out_tongue:

Oh, yes…Welcome.

G.

Do you have actual stats, or is that just some random guess?

“Probably”? Thanks for that glimps through your crystal ball. :grinning:

Some people are viewers, others are participants…
I’m sure that the Haiku will be what we (or someone else) do about it.
Any way, future not written yet, I think, I hope.
Personally, I hope that Haiku grow as open source community. I do not belive, that some big business company will be interested in Haiku. And smaller companies, that may be interested in Haiku, still have to rely on the open source community.

…So welcome!
…to our campfire.

1 Like

Personally, I think we’re close to reaching a point where Haiku will thrive. User aren’t declining, they’re waiting.

2 Likes

I have to disagree at this point. Having a daily build available, the lack of an “Alpha5, 6, etc” is not really relevant. How much different can be the next “Beta release”, from the “nightly build” from the previous day?

Haiku is evolving day by day. The today builds are a lot better that the previous year, and so on… even if there are not a new Alpha or Beta label attached to them… :slight_smile:

But this is the warning associated with the nightly builds.

These images contain the latest enhancements, bug fixes, and hardware
 support. Please keep in mind that there may also be regressions and broken 
features in these builds which have not yet been identified or resolved. 
USE AT YOUR OWN RISK!

I think the Haiku circle is probably not as big as it was or appeared to be around when I tried alpha 2, but to me, that’s actually a good thing, as I think the community is closer together and more active today… much more so than just a surge of interest is.

As for Beta, it has been mentioned time and again… and to be honest I don’t think it’ll happen anytime this year. Maybe next year, but I’m not as excited for it as I once was. And since I’m not on the dev team, (I guess to me it appears very hiearchial), I know I’m not helping get there, but I do want to see Haiku succeed and grow.

The Nightly builds, however, are the forefront of things. And aside from the occasional afflicting glitch, they’re a joy to download. Since Haiku keeps adding fixes and features, why can’t the project go with a rolling release model? I haven’t liked Microsoft for the longest time (with very good reason) but the new release rings in 10 make it exciting and feels like you’re a part of something. Debian and Chrome have followed this for a long time (Firefox for a few years) and it’s working. To me, as a developer myself struggling to get the Poem project out of the mess its in, I think rolling releases “unstuck” a project and make its users feel warm inside… if that makes sense. The admired release cycle of yore is giving way to the new. So as for Beta, there is my (most likely unnecessary) opinion on it: Haiku should remove the beta warning and switch to a rolling model.

And anyway, there’s a very big helping of opinion. Maybe I said way too much… and if I did, my apologies.

Oh, and welcome to Haiku!

Debian is doing rolling releases now?

Well, on one side we get complaint that we should do rolling releases, and on the other we get complaints that there isn’t a stable version of Haiku to build on and we keep breaking people apps.

The (current) plan is to release beta1, then enter a kind of “rolling” cycle from there, with updates until we get to R1. But meanwhile, some of the Haiku devs want to work on things that break everything (complete redesign of the media kit or the interface kit, for example). We can’t put this in the rolling release, until the rewrites are ready and well tested, of course. So, there need to be a separate branch.

“creating a branch in git is easy”, I hear you say. Yes, it is. But, our problem is not on the git side. It is in the infrastructure for providing package repos for each of these two branches, so people can get updates for either. This turned out a bit more complex than expected but things are moving forward.

Will the release happen this or next year? I don’t know. But I’m not in a hurry.

Once we have the beta1 out, we will still have big yellow “this is beta” warnings, for the beta branch, and big RED “this is not even alpha” warnings for the R2 one.

2 Likes

I freely confess my earlier example was a bit exaggerated to make a point. However, yes, SPI has been doing this for years now. There are, of course, the stable Debian releases (Buzz, Rex, Bo, Hamm, Slink, Potato, Woody, Sarge, Etch, Lenny, Squeeze, Wheezy, Jessie) that come out every ~ 2-3 years on the traditional software schedule, but these are equivalent to Canonical’s Ubuntu long-term support (LTS) releases or Firefox ESR releases, focusing on security patches and updates only.

However, the second choice most non-production clients use is testing, which is a rolling distribution which gently keeps up with the next target, without being unstable or prone to choatic changes. When 9.0 is released, testing will freeze and resume after the stable release is out, and so on. And it’s very easy to transition to it. In fact, if you don’t believe me, a simple triplet of sed -i ‘s/jessie/testing/g’ /etc/apt/sources.list && apt update && apt dist-upgrade -y is all it takes to jump from stable to a rolling release.

This said, I concede to saying Debian uses and has used both models. Of course, there’s also their equivalent to Haiku Nightly builds which is always codenamed sid, or “unstable” – which true to its name, is something they encourage only developers or bug testers to run.

As for Microsoft, they’ve changed quite a bit with 10, adopting the same hybrid SRLC stages/RR model. Users got 10.10240 (“Threshold”, or just “10”) as a final build, and the Anniversary Update (Redstone) recently, which is basically is like a time warp of going from Cheetah to Puma on the Mac. But 10 is also a rolling release if one wants to take the risk. Users can join Slow or Fast Release rings, which put 10 on par with Chrome or other RR cycles.

So all this said, what I was advocating is that Haiku could switch to a similar model, since the project has been in alpha status for years now. That’s just the way everything is, and Haiku is following a thought out, solid plan in the background. I think those who’ve used Haiku for a while (first used a2) understand this well, but newbies or observers don’t. And it’s only a suggestion, but I really and truly believe a compromise at least from the old software release lifecycle that BeOS followed would help the project out tremendously.

But regardless, the rolling release suggestion was/is nothing more than my opinion, and I know it probably doesn’t fit the grand plan or roadmap for Haiku. If Beta and getting to R1 is the way to go, then I’m more than willing to keep waiting for that glorious day.

Well, what we are aiming for is somewhat similar to Debian, but simpler because we have less resources.
Our nightlies are similar to “unstable” as you noticed. Things break from time to time, but generally it does ok.
We can’t really afford the testing/stable thing for the OS itself, so we will have a single branch here, with regular updates (just like you can update the nightlies currently). There will be one or more betas from this branch, then a final R1, then possibly point releases, until everyone gets bored about keeping that branch alive.
Meanwhile, the nightly builds will live on, and start to get R2 features.

The only thing really needed to get beta1 out is this: we need a way to update, both for the nightlies/unstable, and for the beta1/testing-stable. Currently we have only one package repository, so this is not possible. Work is ongoing to fix that and allow our infrastructure to host multiple repos. This has been ridiculously slow because we have only a few C++ coder, and not much people to do the sysadmin and python dev required here. I’m learning Python, but…

4 Likes