Anyway at the end, bounties or not, the problem is still the same, you need to attract more devs and make them stay.
IMHO, the best way is to attract more users, make them feel that the project is evolving, so they stay and think Haiku is worth contributing to.
Make their contribution a pleasant experience so they want to renew it.
Then you will have more people used to code on Haiku.
I think it is quite common for open-source people to be very wary of structures that remind them of large companies and the nonsense that is often found there.
But it doesnāt have to be that way. Even very small organisations usually have some sort of structure. And usually there is somebody apparently at the top of the pyramid. You could call him/her the co-ordinator or the convener or anything you like, but always remember that the incumbent is appointed by the members and is answerable to them for his or her actions.
In the case of Haiku the members would predominately be the developers, so they would still exercise ultimate control. And that is how it should be. Also, I donāt think there need to be any change at all in how the developers organise themselves. Itās the other areas that I think need changing.
I do think there is a bit of a vacuum at the moment, and that everybody who contributes to Haiku should give some thought to how things could be managed better.
This is obviously not a complete list of the things which Haiku gained thanks to bounties (I googled it quickly and canāt promise itās 100% correct): Webkit port, SATA drivers, FreeBSD Wifi Port, Gutenprint Port, Wifi Encryption, Gnash etc.
Based on @waddlesplashās comment, weād have to raise at least $400k to hire one developer for two years. Haiku Inc. has only a quarter of that money accumulated through the last what, 5-10 years? I donāt wear pink sunglasses to believe the Haiku community could achieve that.
I agree. We need more users. And for that, we need a better browser.
Web+ has come on by leaps and bounds. But there is more to do.
Waddlesplash was talking about Silicon Valley levels of remuneration. Open-source can never compete with that, but thankfully there are developers who donāt work in SV, donāt expect to earn $1 million in five years, and are still reasonably competent.
We just have to find them!
1 Like
Based on @waddlesplashās comment, weād have to raise at least $400k to hire one developer for two years. Haiku Inc. has only a quarter of that money accumulated through the last what, 5-10 years? I donāt wear pink sunglasses to believe the Haiku community could achieve that.
That would be hiring random devs for silicon valley prices, this would be hiring specific developers that are willing to be payed less because they care about the project. THe money would be sufficient for about two years now, maybe a bit less, but the recuring donations donāt bring enough money in to make it sustainable to have one developer on permanently.
Webkit port, SATA drivers, FreeBSD Wifi Port, Gutenprint Port, Wifi Encryption, Gnash etc.
I donāt think any of those were added because of bounties, I think webkit was added because of a hired developer, not a bounty. If you are refering to GSOC projects that too has nothing to do with bounties, itās more like an internship.
1 Like
From what I read, the Webkit contract was partly funded through the Haikuware bounty.
The most recent WebKit contract, from 2013-2014 (I think?) lasted over a year and was paid for with almost entirely Inc. donations, until the money ran out. (It was on the order of some EUR 2000 a month, relatively cheap for a full-time developer!)
Iām referring to the original contract mentioned here: https://www.haiku-os.org/blog/leavengood/2007-08-20_webkit_port_complications/
Going back to @Sebrofās flowchart, Haiku doesnāt run like a company and hence most of the legal and finance matters are handled by Haiku Inc. HR and Marketing isnāt really a thing for an open-source project either, although people are encouraged to talk about and promote Haiku. Itās also important to note that Haiku isnāt a product sold for profit, rather itās an ongoing project which anyone can contribute.
Decision making, I believe in the past has been through polls and forms filled out by developers and members of the Haiku community, and IMO that sounds like a fair way to do things. Itās not really the supposed ābad decision-makingā thatās causing Haiku to move forward at a slower rate but the amount of work needed that is way more than the number of contributors working on tasks. Our most active contributors already have a lot of matters on their hands to sort out, and the building number of tasks Haiku should work on is constantly growing.
I recall there was a discussion on here that talked about perhaps giving Haiku Inc a bigger role in the development process (which does make some sense, since most of the members of Haiku Inc are Haiku developers or people actively involved with the Project, so there already is some back and forth between the Inc and the Project).
In terms of funding for bounties, if that is the way to go, we could run a Kickstarter or GoFundMe campaign of some sort - in these sorts of campaigns we already have an advantage since the OS is already substantially developed, and hence we have something to present to potential donors. We could also try running a Patreon page, which will give people an opportunity to set up a recurring donation each month. One thing to consider is what rewards or perks we will set (maybe including Patreon supporters in the OS credits?). There are various bounty websites that operate on the traditional bounty model, and some operate on a model similar to GoFundMe and other crowd-funding sites. Iāve listed some sites below:
Issuehunt integrates with Github, which means itās more convenient for Haiku:
OpenBugBounty is a free bug bounty website, which focuses on web infrastructure (i.e. websites):
Bountysource SALT is targeted towards paying for open-source developers:
https://salt.bountysource.com/
And then of course there is Bountysource, run under the traditional bounty model:
https://www.bountysource.com/
Open Collective is quite like Patreon, except there is more customisable options (organisations can build their own pages to accept donations etc.):
Another good idea is to add a shortcut on the Haiku Desktop reading āDonate to Haikuā linking to the Haiku Donations page. We could also have a screen on the Installer or Haiku Welcome pop-up that reads something like this:
Haiku is open-source software and relies on donations and help from users like you. If you can, please donate to the Haiku Project, or consider contributing - you donāt have to be a programmer to contribute to Haiku!
ā¦And then linking to the Donations page as well as the āGet Involvedā page.
There are also open-source funding programs we could try applying for, such as the one Google runs - we might have an advantage here since weāve been affiliated with GSOC since it started:
https://opensource.google/docs/growing/funding/
However, whether others will want Google to be affiliated with Haiku remains to be seen.
On the security side of things, Mozilla runs a grant program to fund security audits of open-source projects - there are only two requirements which Haiku easily meets, and more details are available here:
For access to donations programs (i.e. Give to Bing etc.) there is Benevity. They have a system for non-profits so maybe Haiku can apply:
https://www.benevity.com/nonprofits
Just curious - what is Haikuware and what happened to it? The website at haikuware.com is blank and doesnāt have any content.
Thanks!
haikuware was a site like where it was possible to download software before the package management was introduced. A kind of Bebits replacement though it never has been as popular and was only for Haiku and Zeta. On last news, Bebits and Haikuware were run by same owner.
Despite the fact that it wasnāt run by Haiku team, it was expected to become the main repository for packages (Haikuport was only mean to be here as an example.). Fate chose otherwise and the guy never tried. He gave up on both sites not long after.
I think the main problem is the perception that haiku is going slowly. I donāt think itās true. We have developed a complete operating system, with relatively good hardware support, a decent web browser, etc, in 20 years, with a surprisingly small team of part-time volunteers.
It took a lot more efforts and resources from other companies to achieve something similar.
The main problem is we started late, in 2001, when Windows and MacOS had already existed for 15 years, and Linux for 10. Of course we have a lot to catch up, and all 3 of them have much larger teams.
The bounties vs contracts thing was discussed at length. The situation will not change here: existing developers are not that much interested in bounties, for a simple reason: they wonāt allow us to free more time for Haiku. We would be paid to do what we already do for free. Therefore, nothing is won. What we need is enough money to be able to leave our existing jobs (or work part time) and still be able to pay food, housing, internet access, computers, etc.
So, yes, there were bounties in the past, but it did not enable that. They were not on people outside the community either. So, itās giving money to the developers and making no impact.
This is the reason why so far, Haiku inc has not engaged into bounties. If, as an user, you think thatās stupid from them, fine. The good news is, you donāt need the inc to start a bounty. You go to bountysource.com, find the Haiku page, which is synchronized with our bugtracker, and you start a bounty there with a few dollars/euros. You can advertise it here on the forum, you can post about it on social media, and you can even tell the inc about it and see if they will consider setting up something (a typical thing would be āevery donation to this bounty will be doubled by haiku inc (up to some limit)ā or a scheme like that). Then weāll see if a developer takes up the offer.
So far, there was a total of $20 put on Bountysource. These were donated by an Haiku developer. Eventually the associated ticket was closed by another developer, who didnāt bother to claim the money.
So, a lot of talk, but you should all put your money where your mouth is or something. Bounties will only happen if people start putting money in, the infrastructure is all set at bountysource, and thatās all you need. It saves time for the poor people at Haiku inc who have other things to do, why should it be on their shoulders to organize this?
2 Likes
This all relative.
Considering the size of its volunteers base, youāre right, Haiku is progressing at a good pace.
But if you consider haikuās ambition to become a viable alternative as a desktop OS, then from the point of view of its future user base it is not progressing fast enough, hence the frustration.
Indeed that is very true when you compare Haiku to the likes of ReactOS or RiscOS.
This is a very interesting to know.
I guess if the core contributors think bounties are not a good way to onboard or incentivize new developers, we should defer to their long experience.
I totally respect your point of view and will not push forward on that front as you are way more knowledgeable than I am.
The simple answer is : because the Inc has money sitting in the bank and I donāt.
But as I wrote in the paragraph above, if bounties have been tried in the past and did not prove to be a good solution, then we should not repeat the āmistakesā of the past and try and find other ways of achieving our goals.
It came to my mind that to attract new devs, bounties could be put on relatively simple tasks, not 3D or browser these requiring a deeper knowledge.
The problem is that if even it was working, it would produce a certain amount of patches that would need reviewing.
That is very true, we should not use bounties if they are proven ineffective/counterproductive. My previous posts were intended to give a rundown of all the available options for Haiku (not accounting for the current situation since I donāt know too much about the current situation of Haiku Inc), and obviously, any option would need to be discussed in-depth first.
I did also mention other possible alternatives to bounties such as BountySourceās SALT which helps ākeep the lights onā for open source developers, so to speak, as well as maybe starting a Patreon or Koffi page. For big goals we could consider setting up a GoFundMe or Kickstarter to better help us towards that goal.
I donāt want to put any unnecessary burdens on the Inc, but the Inc is best suited to set these sorts of things up as they have designated accounts for finances etc, and especially in relation to open source funds usually it is up to the Inc to apply for those.
Another thing we can do is add something to the FAQs dispelling the notion that Haiku is going at a very slow pace, possibly a paraphrasing of @PulkoMandyās prior post about that.
A lot of people are scared by packagefs. They tend to think that you canāt modify anything because what is packaged is read-only.
Maybe the use of the several non-packaged directories could be explained better so people donāt stop on that.
Also making a local package when you donāt have dependencies is quite easy.
In fact, it even makes things easier if you want to try something but, unfortunately all help pages are explaining how to make a recipe to build gas factories.
1 Like
I should like to widen this dicussion slightly.
Haiku is entering a period of transition. For the past twenty-odd years it has existed in relative obscurity, known to few people outside its own community.
However, that is about to change. People are waking up to the way in which the existing big technology companies are exploiting their customers, just as they have become alive to environmental issues. They want something that isnāt MS/Apple, and is simpler and easier to use than Linux. There is therefore a huge potential user base for Haiku. And Haiku is almost ready for them.
If the release of Haiku goes well we need to be ready for an explosion of interest. First there will be positive articles in the computer magazines, and then in the mainstream press. Large numbers of people will be trying to download Haiku; large numbers will be flooding the forums with questions; large numbers will be finding bugs in both the system itself and in native applications.
Is Haiku ready for this?
During the development phase, the developers found a way of working together that suited them, and any suggestions for changes to their processes should come from them, not from outsiders. The role of non-developers is to look at the other issues, and see whether improvements can be made there, in areas where developers donāt have expertise.
In the short term, we need to get over the final hump, and to do so it would be helpful to have more money so that people can be paid to tackle specific issues. After release, we need to have a funding system in place to ensure we have enough revenue to service thousands and thousands of users.
This is why there has to be a change in how the non-development aspects of Haiku are organised. A structure is needed that can respond to the challenges that the future will bring. Haiku, Inc., could become the basis of that structure, but not necessarily.
Haiku is, and will remain, a free and open-source operating system. It will always rely primarily on volunteers, and those volunteers will mainly be developers. The new structure should ensure that developers have the biggest voice in how Haiku is run.
.
4 Likes
You are very right - Haiku could be positioned as a privacy-focused OS for users who are escaping the far reaches of the tech giants. Itās not that hard either, since Haiku is not āJust Another Linux/BSD Distroā and architecturally Haiku is distinct from any other operating system (except maybe BeOS). This puts us in quite a good place since most viruses/ransomware are incompatible with Haiku.
Except, I donāt think there would be a massive explosion of interest unless we all pitch in and help each other out - whether thatās updating websites, helping with documentation and manning the forum threads. Thereās also the job of producing content for our social media channels, especially for our YouTube channel which is very underutilized, as well as contacting technology websites and magazines (as well as anyone else who might be interested). Apart from those, there is also the task of helping with development, as well as porting apps at Haikuports (which I will point out that anyone can get started with thanks to the available newbie-friendly documentation). It is with all these parts of our community that Haiku can really realise its potential. In terms of server load, we can put out a call for mirrors but Iām pretty sure our servers should be able to handle most of the traffic.
However, to fully capture this new audience, new features also need to be researched, developed and implemented to further enhance the usability and privacy of Haiku - such as a user accounts system and āprivacy indicatorsā to show when a userās clipboard, microphone and location are being used. These will obviously take some time to implement.
In terms of a funding system, I have suggested some ways to obtain more funding already. We could possibly come together and try to discuss a funding strategy to form the backbone of our system in another post on the forums.
In terms of a new structure being needed, we could try to maximise the potential of Haiku Inc - something like adopting a Mission Statement or Charter as well as implementing a Roadmap or āPlan for 2025ā or something like that could help. The Roadmap neednāt been too long or fancy, and it could just be an official adoption of the release milestone descriptions in Trac. The Inc could then be responsible for trying to fulfil those goals and providing the Project with the help and resources to allow Haiku to work towards those goals. I donāt think a complete restructuring of Haiku Inc is needed - most organisations have a President, Vice-President and Treasurer. However, maybe the board members could adopt roles on specific parts of the Project and report back on them - maybe one Board Member could be responsible for PR, one for Development, one for the Community etc. However, the current Board Members are active Haiku contributors (and are quite busy) and Iām not too sure if they would be willing to take up another responsibility.