Am Donnerstag, 22. September 2011, 22:13:12 schrieb Cornelius Schumacher: > On Thursday 22 September 2011 10:29:16 John Tapsell wrote: > > I think it would really help with bug reports if we let people > > > > donate money towards getting bugs fixed. If there are lots of users > > annoyed by a bug, and there's no longer a maintainer for that code, > > then it makes a lot of sense to let those users contribute money > > towards the bug fix until the financial incentive is large enough for > > someone to step up and fix it. > > As Harri already pointed out Don Sanders tried that a couple of years ago. > It didn't work out.
What did he try? I think there things are mixed-up here. There is "fixing a bug" which is most of the time different from "add a feature". And there is "earn a living from bounties" and "get some extra money for what you do/would have done anyway". > There are mainly two reasons for that. The first is that serious software > development costs serious money. Getting enough money together from > individual donators to actually competitively pay for the development is > really difficult, because it's by far not done by two or three people > pledging 50 EUR. So making a living from this kind of system would require > massive numbers of supporters who continuously put in significant sums of > money. This is really hard to achieve and certainly won't work for many > developers. As stated above you seem to assume that if one cannot make a living from it there is no use of getting some money and further that this is only about adding large chunks of code rather than small bug fixes. If people tell others that it's easy enough to learn and start contributing, even if only with small patches, then that contradicts that one must earn a living from bounties which are exactly meant to add incentives to small contributions (and maybe even bigger ones). > The second reason is that money is a horrible motivator. Of course everybody > needs a certain amount of money to fulfill basic and some non-basic needs. > But what motivates us is not getting money, but doing something that > matters, doing something we are capable of doing well, having the freedom > to do it in a way we feel is good and right. > > Money introduces a dangerous dynamics into this. It values extrinsic rewards > over intrinsic rewards, opens the can of worms of fair distribution of > money, and it requires a new level of administration, which is hard to get > right. Ok, so if the money you earn while coding for KDE is from a company that's not dangerous. If you get money from google's summer of code (which is advertised by KDE) it is not dangerous. If you collect money through donations – as Sebastian did – it's not dangerous. But if you get a bounty for a bug fix it becomes dangerous? How come? If one claims that bounties on features may lead to unmaintained code – why still advertise GSoC? It does exactly that since years. On the other hand there are lots of examples where features ended-up unmaintained without their addition being paid for. So money is not the one cause and IMHO certainly not the main cause – and KDE advertises projects that pay money for features – not for a living but just some extra bugs. And of course, bug fixing is something completely different. Most of the time it alters some lines of code but does not add any feature to maintain. It fixes rather than adds. And maintaining working code seems less work than buggy code. > Doing free software because you love it works much better. Enjoy the > technical challenges, enjoy being part of a wonderful community, enjoy the > feeling of accomplishment, when you get something great done. This pays > much better than money for fixing bugs. It pays in happiness, but it also > pays materially. Successful involvement in free software projects is a > booster for many carreers. This will put you into a situation where you can > earn the serious money, and if you are a bit lucky even with something you > really love. I mostly agree with you but it's simply a fact that there are issues in KDE that "nobody" cares about about, loves, has time for or however you want to call it. Issues that nobody enjoys fixing or nobody has time for fixing etc. They get fixed if somebody feels like it and money can help to feel like doing otherwise unsexy work. And I would also claim that those currently contributing do not suddenly turn evil (I'm exaggerating) because they get some money for a bug fix. There won't be hundreds of people just getting on the KDE train to earn money. (And even if, every bug fixed is positive for KDE. Features might be different.) I would like to take a more positive approach. Those that contribute to KDE do so because they want to, because they are able to and because they like to, they decide which bugs and features to work on. Adding some money to that will still leave it up to the people to decide what they want to work on and how to share their money with their community. If your company pays you for development it might put up some goals, e.g. in kdepim feature x/bug z to meet the demands of client y. Your work is directed according to the company's goals. Users offering bounties try to give an incentive for things that "cannot" be loved and thus stay unfixed for a long time. If you accept the former you cannot abolish the latter. Especially when it comes to bug fixing! What's so dangerous about giving incentives to otherwise pretty unsexy work, i.e. bug fixing? So to sum-up. Bounties for bug fixes are different from bounties for adding features. Unmaintained code appears with or without bounties. KDE does take part and advertises projects that pay for features already. (KDE) devs are not evil and will not suddenly become unsocial and just lead by money because there are bounties set out for some bug fixes. There are already developers payed by donations/companies which are not as free in their contributions as they would be if there was no company paying them their living, yet there is no issue with them within the community. And regarding the article about the rats, I do not think it's valid because it would translate to "people add bugs to KDE in order to get money to fix them later". The code added/altered now, without any bounties, has the same "attractiveness" to do evil, i.e. introduce bugs or security issues, if not reviewed properly. Bounties would not change anything about the need to review code. And finally, I'm not claiming that there are no issues with bounties but one should take a positive approach and try to pick its positive bits rather than abolish the whole concept. Sven >> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<