I’m sure that sourceforge has a marketplace. I’ve never used it but had quite a few mails about it. No idea if it actually does anything though.
You ask do you honestly have the kind of time it would take to do that?.
I think you have a point but it was only valid two decades ago and it mostly appeared to the things that were very different (i.e. build system etc). Faced with this problem the open source community created Debian and other systems like it. Debian basically provides a single consistent way to building every significant open source application every written.
You can compile ANY debian application using the command: debuild -us -uc -b and by that I mean EVERYTHING. It works for small apps like gedit, it works for big apps like Firefox and it even works for hardcore stuff like the entire Linux kernel.
If you know C very well, you can quickly get up to speed with most C applications (of course hardcore stuff like the kernel has a bit of an extra learning curve). But in general, if I run into an error I can just grep for the error text, start from there and work myself backwards through the IF statement.
What about the risks of fixing something the wrong way due to misunderstanding the architecture of the app? Well, remember that most of the time you don’t get direct GIT/SVN access instead you submit a patch to the bugzilla of that project and a maintainer will review and merge the patch. This adds a great quality threshold to the process.
It seems like it would be too hard and not worth it, but it reality I enjoy it and weird as it seems… it actually WORKS.
As many have said in different ways, the problem is a nightmare for requirements gathering. In the normal scenario, someone is the boss and decides when the requirements doc is finished and then it is sent to development. With a bounty system, you have the problem that no two users want the same exact patch and some requirements are mutually exclusive.
User 1 wants a patch with sub-features a,b,c and there is a d sub-feature User 1 doesn’t care one way or the other about.
User 2 wants a similar patch with sub-features a,c,d and definitely not b.
How do you combine their bounties?
Jeff, this time you miss it completely. The whole idea of open source is of people contributing time and effort for creating something good, to be proud.
even your own experience in donating to an open source project didnt have the expected result.
the rest I have to say is here:
http://design-to-last.com/index.php/Technical/paying-for-open-source.html
It’s like this Stallman and his freetards, the source code is freely available… yeah right!
The borders between what open source is and not have actually diminished a lot. And asp.net has a lot of open source I think!
codemonkey on March 27, 2009 06:51 AM
The word orthogonal is being used wrongly here. If two features were orthogonal then implementing one would have absolutely no influence on the implementability of the other.
That said, features conflict quite frequently.
Security vs. Usability is an extremely common example.
Modularity vs. Performance is another.
Extensibility vs. Consistency.
Standards-compliance vs. new useful non-standard features.
Privacy vs. targeted content.
Simplicity vs. Configurability.
Reliability vs. Performance.
Peak Usability vs. Discoverability/Learnability.
Platform Stability vs. user configurability.
Support for content with certain legal restrictions vs. availability with other legal restrictions (eg. GPLv3 patented content, codecs for media players, and even directly GPLv2 as a feature vs. non-GPLv2 as a feature – dual-licensing may not cut it for some cases).
Engineering is about tradeoffs, after all. Sure, in some sense in many cases you could do both, but advances in one consume the advances in the other (especially with respect to performance).
How would it be in the following scenario:
Feature A : somebody pays a programmer 100 US to do it
Feature B : somebody pays a programmer 200 US to do it
and it so happens Feature A and B are orthogonal so if you implement one you can’t implement the other. Should then the one who pays most decide which feature should be added? This I see as a major concern when money starts to play a more dominant role in steering how OSS should be done.
snipWhy isn’t there a service to aggregate and pool funds to sponsor programming particular features or bugfixes in open source software? /snip
There are … they’re called software companies.
I’ve recently brought on someone part-time to help with my personal contracts, and the next round of work I’ll be handing over is for this person to fix up some failing test cases in the open source projects we depend on, Pinax mostly. I think its a perfectly reasonable way to contribute and more of us with the means to do so should consider it.
Hey Jeff,
check out http://matrix.squiz.net/.
These guys do exactly what you are talking about.
The CMS is free to download and use. If you want a new feature, you pay them to build it and then it becomes part of the core system.
They encourage companies to ‘band together’ to get various features built cheaply.
It’s an awesome business model and one I am seriously considering for my CMS.
So start it yourself.
(F**k you.)
I read:
Sometimes, when people say this:
The source code is freely available. If you feel so strongly about this bugfix/tweak/feature/plugin, why don’t you code it yourself?
They’re really saying this:
F**k you.
I thought:
Well, that’s gonna play well on the comments page.
@Anders
It so happens Feature A and B are orthogonal so if you implement one you can’t implement the other. Should then the one who pays most decide which feature should be added?
- Give me a sensible example of this situation arising. To begin with, it’s pretty rare that you get orthoganal features at all. And when you do, it’s normally pretty easy to set up a control to select between the two functionality sets, or to produce two release variants. In open-source, time is the more problematically finite resource.
- Surely it’s better that slightly-too-much is devoted into getting a single feature rolled out, than not-quite-enough effort ending in 2 half-finished non-features.
you are an enigma. an absolute proven legend, who says stupid things to get clicks. Jeff, don’t sell yourself short, even if sleep is not available at the moment.
I think one problem is too many choices, too many open source projects. Which GUI should you contribute to, KDE or Gnome? Which editor, Vim or Xemacs? Which programming language, Perl or Python?
Sure you could say choose your favourite of each class. But there are so many open source projects, the list is never ending even if I choose just one of each category. Browser? Music player? Video player?
What if I donate $10 and nothing happens? Should I donate $20? $100? What happens if I run into anther bug?
I personally hate this If you want that fixed, fix it yourself attitude. According to this attitude I could as well write the whole application myself. Why are you writing an application and release it to the public, if you later on give a sh… on if it is useful for people or not? What kind of BS attitude is it to release software as take it as it is or leave it. OpenSource developers always say you don’t need commercial software any longer, OpenSource will always fill the the gap. I don’t think so. Commercial software is created by people who want to sell this beast, so they won’t say take it or leave it, because leave it is very bad for them. If I report a bug to them and this bug has already been reported by a couple of other users, they will fix it. They must fix it, or people will stop buying it and every buyer counts… maybe not if you are working for Microsoft (actually thousand buyers less won’t make a difference to them), but for most companies fixing something, that can be fixed in two or three hours is reasonable if this makes you sell 10 more copies (unless the price is too low).
I have found a bug in Firefox, a serious one. One that has been reported and voted for by over 100 people. Has it been fixed? No. I gave them sample source how it must be fixed. Have they used it? No. They wrote me back please add your code to our existing one, create a diff/patch and we will review it. That is like someone throwing 100 dollar on your table and you say Please take that bank note, fold it appropriately and put it into my wallet. If you are too lazy to do that yourself, you haven’t earned the 100 bucks in the first place. If I already present them with code they only need to copy/paste into their source code and that is already too much work, this is like saying We give a f… on our users, go elsewhere. Thanks for letting me know, I will.
I see a number of problems with this whole thing.
Just because there is some rent-a-coder deal set up to get hired hands from pool X to implement feature Y for project Z does not mean there’s anyone out there actually willing or able to wade waist deep into the guts of project Z from the outside. I’d expect much of the time that it’s not all that much easier for the pool X guy to write the feature or fix the bug than the person originally making noise about it. True, the person requesting the whatever might not know anything at all about programming, in which case that user would be totally helpless without going out to buy a book about programming or whatever, but in my experience there isn’t as big of a gap between a motivated rookie and an experienced but uninvolved pro as you might think.
The other side of this is that the best way to get something done is to hire the people who already work at the project, and know it from the inside. The issue with this is that your money can’t buy features unless you want to buy a LOT of features. I already use all the time I have. Money is nice, but I can’t in good conscience take your money and promise to deliver feature Y in exchange for it. Not unless you want to pay me enough to quit my day job while I work for you. If you do, let’s make a deal! That would be outstanding! The problem with this is that in the real world it just isn’t so easy to tell your employer you will be taking a week or a month or six months off while you work for hire on this little side project. In my case, I put the minimum price at one year. Less than that, it’s not worth quitting my job and then dealing with trying to find another one when this temporary gig is gone. I’ll even work cheap by programming standards, but so far nobody has been willing to pay $40,000 USD for their pet feature. Such is life.
And then a third issue is that the people in the middle of the thing are invested and involved, and they’re using the project as a conduit for realizing their own personal dreams. That’s a huge part of what motivates them to volunteer there in the first place. Much of the time, these loud requests come in out of nowhere demanding something that just isn’t a priority for the core developers, or which would take the entire project in a direction the core developers do not wish to go. In this case, write it yourself really means I hate this idea, and have no interest in it, but if you’re extremely determined, you will be in a better bargaining position to change my mind about this if you dump code in my lap.
Most of the time though, show me the code means nothing more than Yes, I know, and that pisses me off too, but there are only so many hours, and there is SO much work. Why don’t you do something useful to help instead of wasting all this energy telling me why you’re going back to Windows if I don’t comply with your wishes instantly? What? You’re not willing to make any effort at all? Well then, in that case, indeed, f**k you!
That’s the upshot of not getting paid, I guess. Sometimes it really does mean f**k you.
I remember this one guy who wanted to purchase a feature with a six-pack of beer. Sure. I’ll work 1-200 hours for a six-pack of beer. Jackass.
Yep, that’s exactly how most OSS feels to me. I’ve gotten the fix it yourself many times, and it definitely feels like F**k You to me.
Many (not all) OSS projects, even those backed or even developed by real companies think of themselves as projects, when they really should think of themselves as products.
If someone uses your project for something that you tell the world your project does and then encounters issues, just telling them F**k You results in leaving a rather bad taste in that user’s mouth.
I’ve basically given up on OSS software for now, due to this attitude.
There’s a risky way of popularizing sponsorship:
@author: Joe Programmer
@date: 1-Apr-2009
@sponsored-by: Joe Reasonable User
@amount: USD 500
@lines of code: 1234.
Risks:
copyright blurring and misunderstanding.
Daily dueling between sponsor and programmer for ownership.
Cures:
-
Everything open.
-
Community rating of the deal – could work effectively, like any village gossip council - unethical sponsors don’t like being named and identified as such in source code. Ego massage is an important part of their existence
-
Don’t ask a gentleman his salary, look up the source code!
That way, FOSS programmers can say at their next job interview - i wrote this module that you may be using already!! The small sum of payment might mean that you need to negotiate a lot for a good salary/package/rate, but if you are good enough to write FOSS code with the RTFMs and the STFWs, you should be able to handle corporate slime easily. (But this is an assumption.) -
ROI statistics for all sections of industry - what huge amount of quality content can be obtained for what little investment - compared to buying from closed-source, legally fortified, ethically void bunch of corporations. This is, was, and will be the tipping point for opensource, whether it be Linux or Java or C# or web apps in any of those platforms.
You must use diplomacy when posting a bug report or a feature request.
I have been using debian unstable for two years, and now I think I have posted at least one bug report to almost any free software I have used.
If you read the guides to posting bug reports, have a lot of patience, and don’t get mad at the first negative answer you get, you can usually obtain most of what you ask, with the minimum effort of sending a bug.