Big features like exposing multi-user via the GUI are planned for post-R1 IIRC, as the Haiku devs are trying to get to R1 within a reasonable timeframe.
1 Like
Understood, thanks for letting me know!
[citation needed]
Seriously?
This makes no sense. In what world do you expect developers to try things by live-patching and breaking the system they are developing on?
If you work on something, you donât want to immnediately install an untested version. You want to run it in a controlled environment. If itâs just one app, you build and run it, but you donât replace the system version with it. If itâs a library, you set up something with LD_LIBRARY_PATH to try it. If youâre porting software, you use haikuporter and get it into a package as soon as you can, so, in case you mess it up and files are all in the wrong place, itâs super easy to undo (just remove the package instead of spend hours tracking files around the whole /system directory). And if itâs a driver, you install it to a test partition and reboot to it for testing, or you use a VM.
A lot of people think packagefs is great because of this (my word is as good as yours on this).
2 Likes
Iâve been hearing this for, like, 25 years? It was already what people thought of BeOS. âIf only they had shipped it before Windows 95 was outâ. âIf only BeOS had good pronting support in 1997 so Apple would have bought them instead of NeXTâ. âIf only we had Haiku R1 ready when Windows Vista was released and lot of people were looking for alternatives to Windowsâ. âIf only Mark Shuttleworth had picked Haiku as a base for Ubuntu instead of Linuxâ.
That seems to be the eternal malediction of Haiku, being always just 10-15 years late to everything. Not because weâre slow, but because we started later.
So, two things to make of this:
- Either we are too slow, and the situation now is worse than it was in 2001 (situation in 2001: Haiku had nothing but a screensaver preferences panel initially, Linux was slowly getting USB support, Windows XP was released)
- Either we are moving faster than other systems, and eventually we will catch up, even if it needs a few more years of work.
Have you read the reviews for Haiku Beta 2? Look at the one from ArsTechnica. Clearly the reviews are not that positive. We have work to do, still 
I donât think that there is just a âfinal humpâ as you say. There are still years of work ahead of us.
That is a non-argument, the malware will come as the system becomes popular.
Thereâs some difficulty here in the sense that the developers have grown a common vision of what Haiku is and what it could become, and are rather conservative on how to advertise it. The reason is, we know we are not ready for end-users yet. So, any advertising should take that into account, and be honest about it: at the moment we are looking for early adopters, people with some technical background who will be understanding of the state of the OS. Itâs something to keep in mind of you are going to work on marketing/outreach (I think it is not a problem, itâs still very much possible to advertise to that audience).
However, it may result that the advertising channels are perhaps different ones. We have people relaying our news on Phoronix and Osnews, for example. We have articles on linuxfr.org for the french free software community. We have things going on Hackernews as well. Maybe Iâm old-fashioned but I think for that particular audience, this seems more efficient than Youtube? And as usual with our linmited available time, we try to focus on the most effective things. Spending half a day writing an article for linuxfr is something I can easily do (often I work on it by spans of 1 or 2 hours during lunchbreaks at work). Recording and editing a Youtube video is something I donât have the hardware for, and no interest or knowledge in doing. That being said, most of the talks I gave about Haiku have been recorded and are available elsewhere, we could already have a playlist to reach those, for example. Cheap way to get Youtube presence 
I have already said this in previous discussions, but I think approaching things from the âwe need fundingâ side is not quite right. In fact the Inc has some money available already. So, what we need is not funding, but ways to spend that money.
No.
The developers are volunteers and any kind of attempt to do this will be ignored by them. This is because we all work on Haiku for fun during our week-ends. In my case I can tell you exactly how it works: this week Iâve been writing high level C++ code at work for a new project. During the week-end I plan to work on low level driver things for a change. No chance for me to look into WebKit this week-end, for example.
That is why the roadmap has no deadlines set, for the most part, and there is no strong commitment from the developers to it. We are doing it in a best-effort way, and thatâs all you can get from us.
Of course, if there is a paid developer, this changes. But, thatâs already in place, with a very simple mechanism:
- The developer offers to work on something under a paid contract, and explains to the Inc a) why this thing should be worked on and b) how a paid contract enables them to do it more efficiently
- The inc, with good knowledge of the project goals, decides to accept the contract or not
This is very simple, as you can see, and has worked fairly well until now. Thereâs no need for anything more strict, as far as I can see.
What we need is more people applying at step 1. And thatâs where the outreach comes into play, but the inc members currently think that hiring people from inside the Haiku community is better than hiring complete outsiders, or setting up bounty programs. Hence they did not do a lot of outreach: the people in question already know that the option is available.
Now thatâs something which can be disagreed upon. In fact, if no one among the developers is interested in such contracts under the current conditions, it makes sense to look elsewhere. This is why there was an attempt with Outreeachy last year, for example. It could be some other things, but I doubt that a generic job offer to the outside world would work, for the simple reason that we donât have enough money to make a competitive offer.
So, we could fund it then? Yes, maybe. But it seems difficult to me to start a Patreon or Kickstarter where the plan is âwhen we have enough money, weâll hire some random developer(s) to work on this thingâ. I think it makes more sense to find an interested person first, see what their conditions would be, and only then, see about how to fund things to make it happen. In that case you can have a Patreon/Kickstarter/whatever thatâs more specifically focused and I think it would be a lot more effective.
4 Likes
Hi all,
First of all, I want to say thank you for continuing such an insightful and civil conversation/debate about what we should do next. Itâs good to share our opinions, even if you or I disagree with them.
In response to @PulkoMandyâs post:
Recording and editing a Youtube video is something I donât have the hardware for, and no interest or knowledge in doing. That being said, most of the talks I gave about Haiku have been recorded and are available elsewhere, we could already have a playlist to reach those, for example. Cheap way to get Youtube presence
We could just use a video editor like OpenShot or Canva and make a simple YouTube video without the need for cameras/editing etc. The playlist sounds like a good idea - whoever has access to the YouTube account could create that. We might also consider applying for channel verification as well as adding channel art (maybe the winning wallpaper), linking to our other social media channels and websites and adding a description of Haiku in the âAboutâ section of the channel page.
Of course, if there is a paid developer, this changes. But, thatâs already in place, with a very simple mechanism:
- The developer offers to work on something under a paid contract, and explains to the Inc a) why this thing should be worked on and b) how a paid contract enables them to do it more efficiently
- The inc, with good knowledge of the project goals, decides to accept the contract or not
This is very simple, as you can see, and has worked fairly well until now. Thereâs no need for anything more strict, as far as I can see.
What we need is more people applying at step 1.
I agree with this - no need to overcomplicate things! So the idea of a Mission Statement and Roadmap can be dropped too, since developers already have ideas for future releases.
You mentioned that the Inc already has money, except that we need to decide where the money should go. Maybe the Inc could hold a meeting to discuss this point? This thread grew out of the thread discussing the lack of HTML5 media support. Whilst some people werenât keen on the idea of bounties, maybe we could consider contracting the work on WebKit out to a Haiku developer (maybe @X512 if they are interested, who said they are working on WebKit API and interface support)?
IMO the Web Browser is probably one of the biggest pain-points of Haiku now from an end-userâs point of view. If we draw up a contract for that and others work on solving blockers to ready us for Beta 3 that might be a good course of action.
You misunderstood me.
Iâm not saying that Iâm against packagefs. Itâs not the goal of this thread. (Itâs a good thing and itâs here to stay anyway.)
Iâm saying that where you and me are convinced, some are not. (You have whole Hakilo thread and things that were split as example) I donât want people stopping at a first impression that everything is locked when it isnât true. Showing them how packagefs can help them and simplify their life is better than explaining again and again that it isnât a constraint. And I think that it is possible to convince these people with a series of small concrete examples easily understandable for a beginner.
Also, you donât need haikuporter to package the 3 or 4 wallpapers that you use or a sound scheme. You will hardly mess up your system with that kind of things. Those are not forcibly things that you want to share either but you can make easily a local package for them. We also need simple examples for that kind of use.
1 Like
Indeed, Packagefs is one of the great features of Haiku. Installing new applications and updating everything is easier in Haiku - much easier - that it is in any other system I have come across.
There are occasional problems with both the updater and the Depot, but these will no doubt be fixed.
To be very precise: what we said is it doesnât help if your goal is to have the existing developers spend more time on Haiku. The reason is that it doesnât enable us to stop or scale down our existing jobs.
On the other hand, there is the idea of attracting new developers through bounties. There has been some arguments mentionned against this as well: that people will come in for the money, dump a block of code on us, and go on to the next bounty maybe in some other project and never plan any maintenance. I donât know if thatâs the case. I never worked in any project with bounties. Perhaps we should ask the project who routinely do this what their experience is about it. Maybe we should try it and see what happens. Maybe we can attract some developers with bounties, and then retain them with a more permanent contract offer, as well.
Thereâs merit to the idea. But for it to happen we need people to start putting money on bountysource, and if the thing catches on, we can get the inc to join and donate as well.
They are working on it, I think. But youâre never sure since their discussions are private and there are no meeting notes for their mail discussions.
1 Like
If youâve been hearing it for twenty five years, then maybe itâs time to do something about it?
But seriously, all I am proposing is that the management of Haiku be placed on a more rational foundation, where there is transparency about decision-making and a structure that helps to turn decisions into action.
For instance, several people, including yourself, believe that more work needs to be done on Web+. Is this a general view amongst the developers? And is there also a general view that somebody should be paid to do it?
If so, then a decision needs to be made to find somebody, and the first task is to draw up a job specification. Then a way has to be found to fund the salary. We could just wait until somebody dies and leaves their fortune to Haiku, or we could be pro-active and go out and try to raise funds.
But to do that we need to be able to show that the organisation has a mission and a plan, is properly run, with officers who are accountable, and a membership that gets an annual report and accounts, and can elect or un-elect directors. Itâs these things that give donors confidence that their donations will be used properly.
But what do we have? A shadowy group that appears to be completely unaccountable, was never elected (as far as I can see) and which keeps half of Haikuâs assets in Bitcoin. Well, sometimes itâs half, and sometimes itâs more, and sometimes itâs less. At the end of 2017, Haikuâs Bitcoin holding was worth $42,000 and at the end of 2018, having received further donations, it was worth $14,000. None had been spent or cashed.
Iâm glad to say itâs gone up again since 2018, but I have yet to hear an explanation of why it hasnât been turned into hard cash. As a donor (admittedly a very small one) I find the decision to gamble with Haikuâs assets to be disturbing and dispiriting.
And before anybody says anything, not taking a decision is in fact a decision.
But I suppose itâs easier just to do nothing and let Haiku plod along for years
Yes, indeed. If nobody is employed to speed things up.
Haiku is amazing. Itâs potential is enormous. Letâs try to realise that potential, and make all the developersâ hard work worthwhile.
1 Like
To the developers, it doesnât matter if someone is paid or not. There is just a very large list of tickets which are triaged into milestones and also thnaks to user votes. We need to solve them all. If someone gets paid, we solve them faster. Doesnât matter which part they work on, as long as it fits in the roadmap. Of course, preferrably R1 things.
Thatâs your view. I agree with some of it only.
There should be annual reports and accounts, I think no one is going to argue with that. I donât think memberships and elections are required. The current solution is that the board of director co-opts other people and I think it works fine, because the board of directors is composed of people who are very trusted by the other developers.
Also, it has been mentionned several times, if you want an organization with memberships and elections, you can already get that. Join the Haiku Support Association which has been there since 1998 and also funding parts of Haiku (mostly the coding sprints and developer meetups). So, thatâs already done and you can completely ignore Haiku inc then.
Haiku inc is designed this way: if you donate to it, you have no control over what they do with the money. Thatâs not an error on our side but an intentional thing. The inc status are designed so that people making donations can in no way influence the direction of Haiku, and the developers are the only one responsible for defining where Haiku is going. As a donor, you may not like that, but I donât think itâs going to change for Haiku inc. So, in that case, you can use some other way to support Haiku with your money: bounties (where you very much have control on what you fund), joining the HSA, direct donations to developers, etc.
Thatâs your opinion, I would guess people who decided to donate bitcoins instead of standard money donât necessarily agree. Otherwise they would not have donated bitcoins.
1 Like
Well, at least we agree on some things.
I donât have any issue with a few trusted people having control of the purse strings. It seems eminently sensible. But Haiku needs more than that, and if a benevolent dictatorship is a good way to get things done, you would see more of them around. I havenât seen any. Have you?
I donât âwant an organization with memberships and electionsâ for itâs own sake, I want a responsive and accountable directorate, and what we have is neither.
I certainly donât expect to sway Haikuâs direction because I donate. That is never the way, and I donate to several organisations. But I do expect to know how the money is spent, and the plans for the future, and that is what the other recipients of my donations give me. Itâs not unreasonable to expect the same from Haiku.
I canât âignore Haiku, Incâ because Haiku, Inc. is where the money is.
The HSA appears to have very few members, and although it solicits for donations, it publishes neither accounts nor minutes of meetings. It runs an annual event in Germany, but I havenât seen any evidence (other than Codesprint) that it donates money to the Haiku project to pay for developers, which is what I want to do.
It may be that Codesprint is an effective way to support development, but there is no information about past Codesprint sessions, and what they achieved, on the HSA website
Finally, are you seriously suggesting that those who donated Bitcoins expect Haiku not to cash them? Why? What then is the point of them? The reason they donated Bitcoins is surely that they already had some.
In my opinion, if Haiku has survived and progressed that far in 20 years, itâs because it is a developer centric organization driven by consensus.
Drastically changing the way it is organized would most probably alienate the developers and destroy âthe Haiku Wayâ.
Therefore I think the Inc doesnât need more power over the haiku Project that it has now but needs more manpower to better organize and be more efficient, with clearer goals and more transparency.
That is why I proposed my humble assistance.
I personally think the changes put forward by @Sebrof are too far reaching and I am more aligned with @jt15sâ propositions and @PulkoMandy âs concerns.
In particular, I think being more present on youtube (by the way, does Haiku even have a youtube channel ?) would help attract regular users and young developers.
Iâm an old fart, but I imagine that being more active on twitter and other social networks would also be beneficial.
And once the Inc has set clearer goals on how and on what spend the money it has accumulated, we could start some kind of crowdfunding to attract enough donations to properly pay for the improvement of the OS and its ecosystems, especially on pain points (for me) like proper HTML5 support, video acceleration and multi-monitor support.
Hopefully the Board of Haiku, Inc will soon find the time to regroup and discuss these ideas.
And I donât think they need to onboard new members if they are not ready to, they could instead set up a discussion group to organize the proposals and actions of members of the community.
3 Likes
Haiku has done very well. I am not proposing a change to the way developers organise the development work, and any changes to the way other things are organised should still be subject to the control of the developers. The developers ARE Haiku.
We agree that clear goals and more transparency are needed. But we also need to ensure that the structure doesnât get in the way of achieving those goals. When Pulkomandy says he has to work around the Inc. it suggests that something isnât functioning as well as it should.
My posts have been made in an effort to spark discussion and find a way forward, in particular with an eye to increasing Haikuâs funding and therefore its ability to employ developers. How that is achieved is not important, so long as it works.
1 Like
Before I start, I want to point out that there is a similar discussion on the mailing lists as well regarding donations and where the money will go:
https://www.freelists.org/archive/haiku-inc/12-2020
There are a couple of interesting things there, including the fact that the Inc has people interested in joining as new members:
Yes, in the past months we have had people offer to help Haiku, Inc
and add new members to the board, we just need to try to finish some
of that up. It is sort of a chicken and egg problem because we only
have a few active members and we donât have a lot of time. Given the
money Haiku, Inc has we also need to make sure we vet people and donât
give a random person off the street access to all the funds.
We will try to sort this out and then if we have some fresh members
with more time we can more actively try to find a good developer.
This leads me to @xipehuzâs post:
Looking at the quote from the mailing lists, it seems that new members have to be âvettedâ just to be safe, which obviously takes a while, and I think that only current members of the Inc are allowed to decide who is allowed to join and whatnot. Have you asked on the mailing list about your willingness to help?
Thanks for the clarification - I agree with your point on finding information from other projects already using bounties and analysing the benefits and drawbacks. Iâve found the following information on what the Open Collective Project uses, which describes their models etc.:
We could try contacting them and asking what their experience with bounties has been like.
Looking through their last meeting minutes (on the 14th of September), it seems that all was discussed was a member stepping down and letting the Inc know of the Haiku Inc. owned items he had in his possession:
https://www.haiku-inc.org/docs/haiku-inc.meetings.2020-09-14-1900.txt
The last documented meeting before that was 2016 - Iâm not implying that the Inc is âslacking offâ or whatnot, but I personally think that maybe the Inc should make sure that all discussions are filed so that there is no false impression that the Inc is not doing much. It might be good also to make any other discussions more transparent, whether that be letting us, the community know what the Inc. is working on and has discussed here on the forums or providing summaries of any other discussions.
Another thing - perhaps the Inc. could adopt some sort of Aim describing its purpose just to make it clear to everyone what role the Inc plays.
Finally, I found this on the mailing lists:
This is actually easy to realise, all it takes a frozen-until-goal-reached
Patreon account and a small campaign, that wonât charge people until the
desired monthly salary amount has been reached.
A Patreon account for Haiku, Inc. was supposed to be opened a while ago, but
that has also stalled I guessâŚ
So it seems that Haiku Inc has decided on opening a Patreon account, but that has seemingly stalled.
If there are any board members reading this, could they perhaps join in the conversation and let us know whether they have agreed or disagreed with the topics raised here?
3 Likes
Yes, I have offered my assistance through the aforementioned mailing list at the end of october : https://www.freelists.org/post/haiku-inc/Offer-of-administrative-assistance.
Just to be clear, Iâm not asking to join the Board or change the governance of the project.
I juts want to help the administrative side of Haiku in any way, shape or form the Inc will deem appropriate.
2 Likes
There are meeting notes only for formal meetings (such as when a member join or leaves). Other than that, a lot of the work is done through internal mailing list discussions.
For the haiku project, our solution is to have the mailing list archives public, so everyone can see what was done. For the inc, that is not possible because they deal with personal details occasionally. For example, one developer had to stop a contract with the inc because of health issues. Not to mention bank account details, and so on.
That being said, a yearly report (at least) would be nice. And I think also a yearly meeting where the inc board members would explicitly renew (or stop) their membership.
That is part of the statement of incorporation, itâs a rather wide and generic statement, but the idea is to let things âflowâ and allow the board to change the way they do things. Again a consequence of having only a small team and trying to reduce administrative paperwork to a minimum:
3 Likes
Totally agree - and maybe this could be a time to allow new members to join as well.
Ah - I didnât know that existed. Should that perhaps be added to the Incâs website?
Yeah, I think the Inc really needs help with Admin stuff. Understandably, they are all quite busy both with real-life and other duties so it may take a while for them to come back and figure things out. Iâve copied the following response from @leavengood for those who havenât seen his reply yet on the mailing lists:
Maybe someone else will respond too, but just in case not: I think we
can definitely use some help, but it will take some time and effort to
figure out what we can do to bring people on board. Most especially
for me and things related to the books and Treasurer stuff.
Hoepfully the Inc gets around to figuring how they can take more people on board soon.
I got it from there (itâs the first thing in the âdocumentsâ page). But an HTML version would be nice.
Iâll try and make one when I have the time - I think running OCR on the PDF will do the trick.