I think martinfick offered great metaphor... [LWN.net]

I think martinfick offered great metaphor...

Posted Jan 20, 2012 13:34 UTC (Fri) by michaeljt (subscriber, #39183)
In reply to: I think martinfick offered great metaphor... by RobSeace
Parent article: LCA: Addressing the failure of open source

> There are various web sites that hook people up with freelance programmers to work on various projects: [...]

I will have a look at those when I have time to try to work out what a typical bug-fix (if there is such a thing) would cost. I would expect it to be quite high, since unless you find programmers already familiar with the project you want a fix for you have all the overhead of the programmer understanding what you want, understanding the problem and getting to grips with the code before they can even begin fixing, and the overhead of communicating with the project owners to get the fix accepted afterwards (always harder than one would naively expect, unless one has been in the shoes of the project owners before).

> But, you could probably find someone willing and more capable of doing it right on the project's official mailing list or something! And, let's be truthful here: most of the time, you aren't even going to need to pay them a cent to fix the bug, because they'll WANT to fix it...

I certainly know that we (VirtualBox) have lots and lots of open bug tickets that no one has ever even found time to look at. And I thought that we were better than average, but perhaps I am wrong there.

(Log in to post comments)

I think martinfick offered great metaphor...

Posted Jan 20, 2012 14:37 UTC (Fri) by michaeljt (subscriber, #39183) [Link]

> Ie: upstream wasn't interested in this fix (or doesn't see it as a "fix" for anything), so you're effectively taking on your own fork...

That means that you are either foregoing all future fixes to the software or that you are willing to pay for future maintenance of your branch - how much do your local plumbers charge?

> Sure... But, if someone posted that they considered one bug in particular to be VERY serious for them, and they were willing to pay you to fix it, don't you think that might spur you into tackling it?

Almost everyone considers their particular bug to be very serious, some even telling us that it makes the product unusable (even if no one else has ever hit it and told us) and that it is not acceptable that we don't fix it immediately. A few do suggest paying us, but we would have a conflict of interest if we took their money for a fix done on work time (maybe even out of work time?) and I am usually not convinced enough that someone would put forward enough money to make it worth sacrificing precious free time for the job - my feeling is that they tend to be thinking of one or two hundred Euros, which by the time you have fixed the problem and done proper quality assurance on the fix (not to mention administrative overhead for freelance work, as I'm not currently working freelance on other things) is too little money.

I think martinfick offered great metaphor...

Posted Jan 20, 2012 15:09 UTC (Fri) by RobSeace (subscriber, #4435) [Link]

> That means that you are either foregoing all future fixes to the software

Lots of people are perfectly content (even happier) running out-of-date software... Look at RHEL/CentOS... As long as you've got the important fixes, especially those that actually matter to you, many are happy to live without whatever new features (and most likely new bugs) newer versions bring... Yes, there may be some fixes in there, as well; if they're important enough, they can be ported into the local version, and you can surely find someone to do it...

> how much do your local plumbers charge?

A lot... But, the point is: much like plumbing problems, hopefully the need to do future software fixes will be a fairly rare thing... Yes, if you're using a bleeding-edge version of actively developed software, it's obviously going to continue having new bugs to deal with... But, an old stable product? Hopefully, you won't run into many very often, just as your pipes hopefully won't spring new leaks every week...

> Almost everyone considers their particular bug to be very serious

Don't I know it... But, far fewer truly care enough about it to put up (much) money to fix it... You're probably right that most of them wouldn't pay enough to even make it tempting... But, that in no way implies that someone couldn't pay to get it fixed if they truly wanted to and were willing to spend the money on it... It may not be cheap, but then neither will that plumber's visit or that auto mechanic's bill... And, unlike those, there's a really good chance you might just possibly find someone willing to fix it for you absolutely free, simply for the joy of coding...

I think martinfick offered great metaphor...

Posted Jan 20, 2012 15:43 UTC (Fri) by michaeljt (subscriber, #39183) [Link]

> But, that in no way implies that someone couldn't pay to get it fixed if they truly wanted to and were willing to spend the money on it...

People really willing to spend the money are often well-served by support contracts. But due to the way that they let/make private users pool their money together (game theory anyone?) I think that proprietary projects have the edge on affordable fixes for those people - and an incentive to actively look for and make those fixes. See also [1] though for my other ideas on the subject.

[1] http://lwn.net/Articles/476077/

I think martinfick offered great metaphor...

Posted Jan 20, 2012 16:08 UTC (Fri) by RobSeace (subscriber, #4435) [Link]

> I think that proprietary projects have the edge on affordable fixes for those people

Possibly, in some cases... But, there's also the possibility that the proprietary software company just doesn't want to fix your particular issue for whatever reason (they don't see it as a bug, or are worried fixing it might break compatability, or whatever)... In that case, you're completely out of luck, unless you can find someone good with a disassembler and hacking assembly code these days (and willing to break the law to do so, since that's probably illegal)... With open source, at least you have the option to go elsewhere for your fix, if you should deem it important enough to do so...

As for your other comment about the bug bounty, I think that's a great idea... It'd be cool to see a little "Donate towards getting this bug fixed!" button on every open bug tracking entry... I don't know how many would do it, but I suspect the popular ones would get quite a few people throwing a few bucks at it... Of course, it might also encourage less scrupulous programmers from adding in bugs deliberately only to claim the bounties on them later... But, hopefully that'd be seen through fairly quickly, especially in an open source project... More likely, I suspect it would encourage new contributors to join in trying to fix the popular bugs, which can only really be a good thing...