It has been a long time since I donated to an AROS bounty, mostly imo, the economy has not been that good. But also I think there may be some basic problems with our bounty system.

Let's say we want someone to assemble a jigsaw puzzle.

After awhile someone takes the bounty, but soon realizes he doesn't have enough time to complete it. Since he won't get paid for completing half, he puts the pieces back in the box and walks away.

In other words, perhaps a bounty should have milestones, after a project reaches a certain agreed upon point, the developer might be eligible for a part of the bounty, so that work done is not lost. (Assuming another developer would want to build on the first devs work.)

But also, most modern projects are really too large for one developer. There should be a way to split the bounty so that if Huey does 50% of the puzzle, Louie does 30%, and Dewie does 20%, they could all be compensated fairly.

But of course, Dewie would say that his pieces were all one color and the hardest section, so he should get more.

I don't have any answers, but these issues should be looked at, if we really expect some of the more complex bounties to be completed.

Hey cavemann, I have to agree with you 100% but have you tried to contact Bill Buck from Power2People? I believe that's where all the bounties are managed.

http://www.power2people.org/contact

Hello amigamia,

Thanks for the positive response. I didn't think to contact P2P, since I am only asking questions and do not have any solutions.

However, you have a good idea. After all, P2P has been crowd-funding longer than most kickstarters. It would be interesting to get their input and suggestions. So I will shoot them a link to this thread in hope of a reply. My guess is that P2P would be ok with milestones depending on how the bounty is written.

Another problem with AROS bounties is that the developers work with ABIv1 and the rest of us use a distro based on ABIv0. So the devs consider a bounty complete when it works on their ABIv1 development systems, yet we cannot use their work until it gets back-ported to ABIv0, frustrating all parties.

Edit:
Yet another problem with bounties is actually writing them. People are always proposing bounties and willing to contribute. But when people discuss which items should be included, or how to word the bounty, the whole thing usually falls apart. Too bad there is no standard fill-in-the-blanks bounty format ;)

Wow that was quick, here is the reply I received from P2P.

Hello, thank you for your email. You raise three basic issues and I will try to address all of them.

1. It should be possible to split donations among multiple developers. => This is already possible and has happened multiple times. Most recently, the Port AROS to ASUS EeePC donations were split between Neil and Michal, for instance. They figured out the percentages among themselves which is the best approach since non-developers are unlikely to add anything productive to these type of discussions.

2. ABIv1 vs. ABIv0 => This is purely a bounty description issue. If you want ABIv0 to be supported, ensure that it is in the description and only donate to those projects that have it in there. There have been bounty projects that specifically referenced ABIv0 or ABIv1 or both before.

3. Splitting up large bounties to more manageable pieces. => Once again, this has already happened. For very large efforts, multiple related bounty projects have been created (usually with the wording "phase 1", "phase 2", etc.). The two completed Kickstart replacement bounties are a good example. Unfortunately, many projects are simply not very suitable for a modular milestone approach. Even if somebody writes 95% of the code to add certain new functionality, it might actually take another developer more time to carefully review, clean up, and ultimately fix this code so it works properly and reliably. Please note that it is impossible to know this before-hand or without a thorough investigation, which takes time and effort. Power2People relies on the AROS community to test and determine if all conditions of a bounty project have been. Thankfully, there are quite a few users around who are willing to test software. However, with many software projects, there is simply no working binary that end users could test until the project is almost finished. In these cases, reviewing whether a milestone was met or not would be up to other developers who are not as frequently found as end users and who might in many cases provide far more value to AROS by focusing on actual development as opposed to review other people's code. Another issue is that Power2People is far more comfortable paying out funds after a large group of people has tested piece of software compared to paying out money based on the opinion of a single developer willing to review source code.

I hope this is helpful to you. I understand that you may have preferred different answers but some things are just naturally complicated and difficult.

Best regards,
Andr? Siegel

Wow that was quick indeed. Very good points and well addressed by Andre'. I do however have a suggestion but it is more for other sites than P2P. Bounties should have more visibility and what I mean by that, they should be advertised clearly on web sites instead of just a link to p2p.

I am going to see if I can add a Project infusion to this site and post the bounties in a more explicit way. I'll see how it goes and maybe other AROS sites can follow. Just some help in marketing the bounties.