What is stopping you from contributing towards the development of Haiku?

Haiku Inc, is copyright holder (and thus the author for many purposes) for large portions of the Haiku codebase. As such it can re-license all that code at a whim. It also controls the Haiku trademarks, and it can terminate the existing grants for those trademarks at a whim. So, if Haiku, Inc. so chooses it can oblige everybody else to abandon the existing project and begin over under a new name, forking the codebase. It would need only the thinnest of excuses to do so, certainly there would be no need to consult with the many non-board member contributors.

I think it’s disingenuous to suggest that this amounts to “absolutely no decision power”.

1 - I got married.

2 - Lot of bills to pay…

3 - Search for my grandfather at Familysearch.org

[quote=Akuji]I would like to see the financial report for 2013 and actual ‘money’ counter on the homepage before I continue to donate: last time it was updated 1 of December so I don’t even know what the situation is right now.

To those dissatisfied commenters from RU-net: you can always fork Haiku and continue developing it the way you like ;)[/quote]

Yes … the financial reports. That’s one task I absolutely refuse to do.

As for updating the money counter on the homepage – sorry, I’ve been meaning to do that but have been caught up working on my new (and first!) house. There will also be another article announcing another month for Adrien too.

First off, don’t get me wrong, I love Haiku and I do appreciate the long hours you guys spend on it. I’m an old BeOS user so Haiku is its own killer app to me.

Several years ago I decided to give Haiku a real shot at use and possibly contribute since I’m a software engineer by trade.

At that point, you were about to release one of the stable Alpha releases but I can’t remember which. After installation, I was stunned at how well it had progressed and I began using it on a daily basis. Other than having to downgrade my video card and a few issues here and there, it was wonderful!

I set up a mirror at home and started to become familiar with the codebase. I’ve never used jam before, so that was a learning curve, but it was something I was willing to learn.

Life and work schedules prevented me from spending as much time contributing as I would like so it was a few months before I could return and in that time, things broke in the nightlies. This made developing difficult, so I waited for it to stabilize.

It never became stable. Sure, things would be fixed, but then others would break. I waited for newer releases, but those are few and far between so I slowly forgot about it. Now I just check back every six months or so to see how it’s going.

As an outsider who has experience with both commercial and open source projects, I’d like to give my suggestions.

Work on more consistent stability in nightly builds. If a user or developer has an issue with the main Alpha release and needs just a few fixes, they are currently subjected to all changes in the nightly builds.

I understand occasional breakage in nightlies, but it’s like you’re all developing in the stable branch. I can tell you from experience that this is a horrible idea on several levels. It makes getting big fixes to users incredibly hard without also slipping in changes they don’t need. It also makes major releases far harder than it needs to be since you’re testing completely from scratch each time. Only the directly effected developers should be looking at such unstable commits and checking them into stable once it is fairly tested by a few people.

Stable Alpha releases need to be far more common. You advise against running nightlies, but Alphas are so far apart that users and potential developers have little choice but to run one. Keeping the codebase more stable would reduce your release workload to acceptable levels. If releases take a huge workload then you’re doing something wrong upstream.

Much of the work done seems focused on new non-core features rather than stability of core features. I’m not sure how much developer time was spent on package management, but from the outside it seemed like washing a car with a flat tire.

Again, I’m not trying to troll you or be overly harsh, but from the outside there seem to be severe issues with the development process that are hurting users and developers alike.

Thanks for reading.

One thing that is stopping me is the lack of the proper resources(I.e a laptop).

I love haiku as I loved beos original, I have donated but the lack of wifi driver is the only reason I can’t use on my laptop as everyday OS and Ive kindof given up sorry :frowning:

And please don’t take that as a negative on the project im sure eventually it will be included, but its now alpha 4 and ive used since the start and still don’t have support for my hardware to be useable in an everyday enviroment. I think driver issues are very important and should be high on the agenda

“Contributing” can mean a lot of different things, besides financial support of the project. Some people contribute their time testing the software and filing bug reports, some contribute by purchasing hardware to test Haiku compatability, some write code, and some contribute money.
Since I know almost nothing about code, my contribution is spending money to buy hardware, assembling systems and testing Haiku on those systems, then filing bug reports when things don’t work. My bug reports are probably not as good as they could be, since I’m still new to this sort of thing. Hopefully, they will get better as I gain experience and feedback from the developers.
I’m contemplating starting a monthly financial contribution to the project soon, to keep what few coding people we have continuing work on the project.

Nice way to say “f_ck off” to the guys, that brought you KeymapSwitcher, big chunk of Qt4 port, VST plugin support and much more (it’s 3deyes and others from RU-net). Over-engineering (or should I say over-complicating) stuff like read-only FS in package management branch is exactly the reason some people begin to think about abandoning Haiku & it’s Quality Police for good and start their own non-compatible fork. Stick to GCC2 compiler and 32-bit as defaults for some more years and you’ll see the last few of Haiku supporters go. Newsflash, no one takes Haiku project seriously in OSS community right now for this very reason - binary compatibility with dead software (of long-dead OS) w/o source code available to recompile in modern environment. Actually, for them you guys are the laughing stock right behind GNU/Stallman and things like that.

Looks like all that backwards compatibility stuff going on for years and years left Haiku in the similar trap Amiga community found itself stuck in couple decades ago.

I agree that alienating the RU-net guys is a really bad idea.

But I don’t think that gcc2 is really a problem as everyone is running a hybrid, and if you want software that doesnt build with gcc2 (e.g. qt4 and loads of others) you can just use gcc4, so it really isn’t standing in your way. I also dont think haiku is a laughing stock, it seems very well respected by all the Linux and BSD guys I know and as far as I’ve seen online. And let’s be honest, Linux and BSD are systems that were started to be source compatible with other OSs, so it isn’t exactly a new thing to be backwards compatible.

One of Haiku’s supposed selling points is a stable ABI, write a program for Haiku today, ship it to all Haiku users for years. But the ABI stability promise only exists for GCC2 on IA32, if you need a more modern compiler or would like software for an architecture that’s not obsolete, or use any features of Haiku’s API that aren’t in Be’s documentation you do not get this guarantee. So it’s practically worthless.

Although Haiku developers have vaguely talked about it, there is no thunking layer for the core components. This is the only practical way to support two ABIs simultaneously in your OS. It’s how Windows does it for example. You can choose to forego thunking entirely, but then you’re stuck with one ABI and to change it you must have a flag day, as a Linux distribution would do. So Haiku is left with a situation where you’re supposed to be able to choose “either” ABI in a hybrid build but in fact you are locked to the one selected for the core components for any plug-ins, add-ons, etc. And that’s why you end up running something like VLC to watch videos instead of Haiku’s media player…

Today’s BSDs are the continuation of the Berkeley (ie University of California) branch of Unix. They weren’t started to be “compatible” with anything, they are Unix, which is why it’s ironic if someone claims one of them isn’t UNIX® because it didn’t pay money to a trademark holder to get certified.

I agree that perhaps “laughing stock” is too strong, but operating systems exist to achieve one goal, to run the users applications on their hardware. Haiku’s hardware support is very weak, Haiku’s application support is very weak, and so the product of these two factors is that Haiku isn’t a very good operating system.

I’m in a similar situation in regards to the lack of a working wifi driver. I want to help in what little way I currently can, so (after some issues) I installed Haiku onto my HP Mini 311. I come to find out that my wifi isn’t working, but I’ve come across that issue with various linux distros in the past so I thought I’ll just plug it in directly and find the drivers to make it work. I find this guide https://www.haiku-os.org/guides/daily-tasks/wireless and follow it exactly but when I reach the point where I need to type in

ifconfig /dev/net/iprowifi3945/0 list
in the terminal, I get an error along the lines of “/dev/net/iprowifi3945 is not a WLAN device”; so I guess that puts an end to me being able to help for now. It’s a shame, Haiku is by far the fastest OS I’ve used on my netbook…which is fairly dated by today’s standards.

And to everyone who has posted in this thread so far, thank you for doing so, it has definitely been enlightening to someone that is fairly new to Haiku. It’s also nice to hear such a wide array of opinions (even if some are negative). I know it’s easier said than done, but I’m hoping through this dialogue all sides can reach some sort of a meaningful agreement on what needs to be done to propel Haiku towards a better future. I think we all know how special Haiku is, and it would be unfortunate if it’s full potential never becomes realized.

My attitude on contributing has changed since I posted this topic. I’m hoping to be able to donate some money soon in the hopes that the few people that are currently working on Haiku will be able to continue to do so. And with this year’s GSoC coming up, I hope a whole slew of bright & innovative students decide to participate and are able to help Haiku progress even further.

What worries me the most is seeing things like this:
http://www.haiku-os.org/community/gsoc/2014/ideas

Specifically at the bottom of the page… “Evaluate Qt as a potential Haiku R2 API”

If I put time in to help with the OS, I wouldn’t want to see that direction ever come to fruition. QT is great as a multipurpose toolkit, but I don’t think it’s a good idea to implement the os relying on it. Yes the interface layers/apis need some love. Though this would completely turn me off to the whole OS in general.

That’s just the most recent one to me. Outside of that, it’s just a time crunch. Though, I’d be willing to throw in on a few areas that need more stabilization and love as well. There is a lot to do left, and stabilization is a key requirement.

As soon as I can get some more time for myself instead of work, I would love and will try to contribute as I feel for the devs right now. They’ve got their collective hands full and could use some help.

You should see the thread on the development mailing list that first suggested that idea! There were some very strong feelings expressed on the matter, with 1/3 of the developers in favour and 2/3 of them against.

It’s not a selling point I much care about, and I do build software on haiku. I don’t think anyone has ever said so either, the software is alpha, it’s constantly changing. It will be more important after R1 I guess.

Media player works fine here.

[quote=NoHaikuForMe]
Today’s BSDs are the continuation of the Berkeley (ie University of California) branch of Unix. They weren’t started to be “compatible” with anything, they are Unix, which is why it’s ironic if someone claims one of them isn’t UNIX® because it didn’t pay money to a trademark holder to get certified.[/quote]

Linux was started to be a free implementation of Unix, and to be compatible. BSD Unix does have differences to other Unixs (off the top of my head, BSD init doesnt have runlevels, for example) but is largely compatible and evolved along with (and originally from) Unix. I think saying compatibility wasn’t important to the guys that originally tweaked their unix source and re-released is wrong though… they still wanted it to do everything the original did.

[quote=NoHaikuForMe]
I agree that perhaps “laughing stock” is too strong, but operating systems exist to achieve one goal, to run the users applications on their hardware. Haiku’s hardware support is very weak, Haiku’s application support is very weak, and so the product of these two factors is that Haiku isn’t a very good operating system.[/quote]

It works great on my hardware, and I like building software for it because it is in many respects better designed than alternatives. It isn’t a laughing stock, because people appreciate its aims and what it is trying to achieve and has done already (present company excluded), even if they know it isn’t ready for them yet because of the reasons you point out. But hardware support and few applications are weaknesses of any new OS, and were once true of any old OS too.

@richienyhus

Thanks for that note, I’ll have to check that out hopefully later today. Just seems to me, that embedding QT is a huge mistake. I’ve built embedded operating systems in the past for various projects and I can’t even fathom wanting to make any OS rely on a library like QT.

Outside of that I need to free up some time because I do want to see this get further and get better. Worse comes to worse I’ll start on extending the layout engine a bit and get some more common controls built out. People like QT because it makes their interface jobs a lot easier, though when the OS Gives you a plethora of controls prebuilt and ready, no reason to rely on another interfacing library to do it. It should be the other way around with the qt library. That’s actually one of the points of the QT library anyway; to abstract away different os interfaces and help you port your app to multiple operating systems. It’s not a system level api, it’s userspace only as well AFAIR.

---- I just read the dev mailing list on the subject… Either way, its r2 and discussion - so no real impact anyway atm. I would say though, if it’s really meant to complete the port of Qt, I would clarify that in the title for the GSoC ideas page. Otherwise people (me or like me) may be confused as to the meaning behind the actual intent.

Only i can do is spread the “word” with my blog http://haiku-colombia.blogspot.com.ar/ and on facebook.

I know i had the same fealing. I have some how begin to think it would perhaps be good to test :). One that have done this are BlackBerry and it’s working greate :). BB 10 ar more or less QNX with QT 4.5 and a xml based gui on top.

Хвйку – это вообще проект для фана. Если фановый проект не приносит фана я вообще не понимаю зачем он нужен :slight_smile: И эти чуваки будут говорить мне как общаться с людьми :slight_smile: Как ревьюить патчи. Умные чуваки ушли из проекта и завели семьи, осталась одна неквалифициированная шваль с синдромом вахтера. А когда ты ничего не умеешь, то тебе остается только строить бюрократическую систему. Сделать продукт эти люди не в состоянии, они никогда не примут ничего нового. Вместо того, что бы пользовать гитхабом они будут в коммитите решать, какой инструмент для коде ревью использоват :slight_smile: Ну бред же натуральный.

Они будут патчить этот Веб+ до посинения. Еще 10 лет назад я за 3 месяца портировал KHTML/KJS форканутый от Nokia Research Center. Браузер конечно сложная задача, но судя по квалификации я не вижу, что эти люди смогут захендлить эту задачу. Умные чуваки есть, поляк например, которые скедюлер делает и другие, но в целом Хайку прогнила до мозга и костей. Коки был клевый чувак, его уход был первый звоночек.


Haiku is fun project for fun. If fun project doesn’t generate fun is just makes no sense. This guys are going to teach me how to manage open source projects :slight_smile: haha. My projects are in top github projects, if you don’t want me to play with you not a big deal, dudes. Smart guys leave the project to be with their families and low educated unqualified dump persons left to lead the stuff. And when you can produce nothing you starting to build bureaucracy. This people are unable to build qualitative software fast and couldn’t create nothing new. Everything was already invented. Instead of using best tools on the planet, using github and other great software they are going to spent monthes to decide wich SCM and which code review tool to use. It is really seem very funny from outside the project.

I’m leading a successfull company (Synrc Reseach Center) for several years that brings complex software to open source, and who you are to teach me how to cooperate with people. I don’t want to cooperate with you. I just wanted to publish an artice and Haiku Chat Application. Nothing more :slight_smile: Lets community decide how it is good or bad. This article was written according to Technical Announcements culture with Synopsis, Summaries and Technical details. Here is the proof:

http://synrc.com/client/chat/haiku/chat.htm

If you don’t like me or my comments about you and your process why not just remove my comment. Why was needed to remove technical Article? :slight_smile: I don’t know but I’m pretty sure you have million reasons to explain this. No Caya abusive or other, just technical details and explanations. Seems instead of reading technical article Haiku community lack of flame in ML and Forums.

10 years ago I was ported KHTML/KJS natively to Haiku in 3 monthes using Nokia Reseach Center GTK+ fork. I know that porting WebKit is a big deal but I don’t see you could bring qualitative software and manage this :slight_smile: There are some clever new guys in kernel teams, but I feel really sad about what Haiku is goin to be. I remeber Koki, he was a very good old-man, and his leaving was a first bell.

Here is my contribution:
Two system sounds.
Find the proposed (by me) start-up sound here: http://soundcloud.com/momadezuber/startup
And the shutdown sound here: http://soundcloud.com/momadezuber/shutdown

All in ogg format. Free for all.