Does Support for Open Source Applications Really Suck? And Why Not?

Having had to pay money for technical support from the Big Names, I'm more than happy to state that the technical support you can get from Google and from those unpaid programmers is far and away better. You don't regularly get direct access to the folks that wrote the code when something goes desperately wrong, and open-source programmers generally have their pride on the line so there's a huge motivator for improving their baby.

The only thing I can say in favour of the big boys is that they more consistently respond quickly. Aside from that, forums full of geeks are extremely helpful, more so than some barely adequately trained non-geek.

I do occasionally speak English, just not today.

As always, the answer is that it depends. "You can fix the code yourself" is of absolutely zero use not only to non-programmers but usually to programmers in different specialties or even those unfamiliar with the complex internals and standards of any non-trivial program. Even those who do fix the code themselves often find that they're stuck fixing it over and over again because their patches are ignored or rejected out of pure caprice. Stephanie, who is not a programmer, has come up with a wonderful theory about operant conditioning, but many open-source programmers do a pretty good job of demonstrating how relying on pride and ego satisfaction as motivators can lead to some pretty bad behavior too. Want evidence? Read LKML, any day of the week. You'll see a lot less altruism or even pride in one's craft than good old-fashioned dominance displays.

In the end it's nothing to do with open source. Free user communities exist for many closed-source programs, and many companies offer paid support for open source. Open-source programmers don't have a monopoly on creativity or passion for their work any more than closed-source programmers have a monopoloy on maturity and discipline. It takes a lot more passion to start an actual business than to start a project on Sourceforge. Claiming that attitude or motivation is an inherent advantage of open source is ridiculous. It's just a way of distributing software, with some advantages and some disadvantages. It's not a panacea, and shouldn't be a religion.

Jeff, what makes you think you have any idea what I do for a living?

Jeff, I've seen the "fix it yourself" arguments posted in forums myself. It's my opinion that a lot of people who are currently lead developers for particular open-source projects really should not be in that position (see Bug 316654 â no ability to configure the different screensavers, where the developer states “I don't have any plans to support this. My view is that any screensaver theme
that requires configuration is inherently broken.”).

When I first saw that, after trying to find out why I could no longer change screensaver settings in Ubuntu, my first thought was that the developer should never have been put in to that position, no matter how great his programming skills were.

But that's one of the things that drives open-source. When enough of the right people become frustrated with a particular project, projects fork. New projects arise. People find things that work for them, and they share this knowledge with others.

It takes a lot more passion to start an actual business than to start a project on Sourceforge.

How so?

It takes a lot more passion capital to start an actual business than to start a project on Sourceforge.

Fixed that for you.

I have more I want to say about Jeff's post, but I'm out of time at the moment.

I don't know, Stephanie, maybe because you do have an "about" page on your site in which you mention that you're a writer but not that you're a programmer. Maybe you're both, in which case good for you, but when somebody loudly advertises one profession but doesn't even mention another, I rather naturally don't assume knowledge that's not in evidence.

As for you, Jason, your inappropriate repetition of a common meme doesn't add any value. Anybody who thinks that capital is the only difference between a startup and an open-source project just doesn't know about adding value. Most open-source projects represent, at least initially, someone's spare time. If it works, great; if it doesn't, oh well. Starting a company means that somebody in the peak of their earning years, with significant opportunities elsewhere (becase it's nearly impossible to get funding if you're not such a person) gave them up to spend years working long hours for little or no pay. It means kowtowing to the barons of finance, which nobody *wants* to do but some do because they believe enough in their project to swallow their pride and devote more than their spare time to it. Statistically it's a risk to one's marriage, retirement fund, and health. People who try it and fail suffer much worse emotional damage that those who start Sourceforge projects and hardly notice that they've failed. That's about as close to a quantitative measure of passion and commitment as you'll fine. BTW, I've never started a company myself. I couldn't make that much of a sacrifice, particularly where my family is involved. I could, on the other hand, take any of several bits of code and slap them on Sourceforge in about five minutes.

Also, without people who start companies there wouldn't be an open-source community to speak of. Remember, most open-source programming is done by full-time paid professionals. Linux might have started in someone's basement, for example, but the roster of current developers shows a lot of Red Hat, HP, IBM, even Oracle. They draw a salary for this, and that salary comes out of working for somebody who built a company for them to work at. IBM service contracts and Oracle licenses pay for Linux development.

P.S. I work for an open-source company, writing and debugging code for open-source kernels and filesystems every day. Where does that fit into the "open-source hero, commercial zero" stereotype?

Dan J, thank you for an unusually thoughtful response. Yes, projects can fork, and that's one of the strengths of open source. Keeping the code alive and free even when its originator loses interest, or develops interests contrary to those of the community, is pretty much what open source is all about and why I believe in it. That doesn't seem to be what Stephanie's article is about, though. Being able to fork a project addresses the extreme cases - the long term, the total breakdown, the errors or disconnects too big to ignore. She was writing about the small thing - the every-day effort to get one bit of help on one person's problem.

I just don't think open source helps much with respect to support. Some open-source projects do better than their closed-source counterparts in that regard, many do worse, and in practically no cases do the differences seem attributable to Stephanie's theory. Where open-source support is good, it's usually due to a *team* of dedicated developers. Relying on pride and ego gratification as a motivator makes it *easier* for a single developer to leave, or for that matter to create a highly disruptive fork, but the community goes on. The open-source projects with poor support also tend to be the ones where "pride as motivator" has gone too far, where developers have been driven away by a few egocentric nutcases at the project's center (cough, Gentoo, cough). In both cases, the difference is not the type of motivation but the ease of participation. Perhaps that's an open-source strength, undermining my earlier point that claiming such an inherent advantage is ridiculous, but it's nothing to do with Stephanie's theory.

All right, Jeff. Now, a few more assumptions that you're making.

(1) No one at brunch claimed any support model was inherent to open source or commercial software. The fact that they are not was actually Greg's point in telling the story. I just didn't mention it because my point was about something else.

(2) This isn't about programmers. It's about users.

(3) One does not have to be an open source evangelist to be annoyed by people making false claims about why open source is bad. One simply has to be annoyed by false claims.

(4) You have a very narrow view of what a business venture is. Venture capitalized start-ups are not the standard model. Most businesses, in fact, are started on the side of regular jobs as projects that may or may not pan out.

In the same five minutes it takes you to slap together enough source code to merit a page on Sourceforge, I can paint a sign that says "iced tea 25c" and attach it to a table that I've dragged to the side of the street, plus mix up a batch of powdered Nestea in my cupboard presently. Your mishmash of code is as much an open source project as my iced tea stand is a business. We're talking at different wavelengths. If we want to talk about an open source project like, say, OpenOffice, as opposed to, say, Microsoft Office, then we're talking the same language. But to compare a viable financially driven business like Microsoft with your mediocre effort to load Sourceforge down with flotsam projects is laughable at best.

I have personally witnessed a pair of brothers start up one computer-related open-source-using company, bring it to the point of being viable (and quite well-known to the internet community), then sell it, only to create another, bring it to IPO, then sell it, and start a third, all within a span of ten years. The only reason this worked is because they had money to begin with -- without that capital, they would have had a big idea and no way to implement. The plural of anecdote, however, is not data, so I'd like to see some real studies backing up your assumptions.

Regardless, as Stephanie said, the article is not about building an open source project as opposed to building a business. The article is wholly focused on the support aspect of both models. Bear in mind that Canonical offers paid support for Ubuntu, despite it being open source. They even build patches for specific companies to address issues, at a cost. If you absolutely must have a paid support model, having the source code can only help the user-driven free support community. Try building your own patches for Microsoft issues with legacy Windows OSs. By comparison, the path to having issues addressed even with legacy BSD is problem free, given the source code is in fact available and you can pay a propeller-head (like me) to fix the bug if you absolutely need to -- and usually at a much lower premium than you'd get asking Microsoft to implement such a bug fix. Beyond that, the bug fix benefits everyone, while Microsoft would only give the bug fix to the company that pays for it once the support model has run out.

One simply has to be annoyed by false claims.

Equally true for claims in favor of open source.

Venture capitalized start-ups are not the standard model.

Got proof of that claim? In the business we're talking about, venture-capitalized startups are very much the standard model. I'll give you a chance to show your proof, then I'll show mine.

If we want to talk about an open source project like, say, OpenOffice, as opposed to, say, Microsoft Office, then we're talking the same language.

You mean the OpenOffice for which both development and support are staffed by Sun, out of revenue from not-so-open projects? Kind of makes my point.

But to compare a viable financially driven business like Microsoft with your mediocre effort to load Sourceforge down with flotsam projects

In your haste to insult me, you've pretty much admitted my other point that Sourceforge is full of flotsam. Thank you.

Bear in mind that Canonical offers paid support for Ubuntu, despite it being open source.

For the third time, a point I've already made myself. I guess the reaction has more to do with the fact that I don't posture and preen as a staunch advocate and defender of noble open source against the abyss that is the only alternative, less to do with the actual content of what I'm saying. Next time I'll bring my cape and tights so I can be accepted by the clique.

Try building your own patches for Microsoft issues with legacy Windows OSs.

Actually I have built patches for commercial OSes, notably Solaris long before it was open, but I'm not the best example. The real question is whether Joe User can fix his broken GNU software any more easily than he can fix his broken Microsoft software, and in fact his chances of doing either are exactly equal.

No, Jeff. The reaction to you in this thread has everything to do with you standing up and posturing about how wrong everyone else is about things we weren't talking about. Nobody's asking you to put on tights (seriously), but a little self-awareness might be nice.

There is a serious misunderstanding going on here about Open Source that I'll put an end to right now. Open Source does not mean free (as in free beer). It is rather hard to believe that people who claim to know what they are talking about and make strong arguments against Open Source models don't get this. Why would you do that?

Open Source is a model for how code is developed and deployed, and to a lesser extent how support is provided. It is not a model for how people's time is paid for.

You can take any open source code and sell it, you can change it and sell it, you can set up a company that charges to provide support, and you can have a company or institution that devotes paid FTE's to its development. This is all part of the Open Source model.

Or you can write a totally kick-ass utility that is closed source that does nothing other than tweak a person's system and install some opens source code so that something else happens. Like Crossover.

Arguments like "oh, they're paying to write this code" or "oh, they provide support for a fee" are utterly irrelevant to this discussion.

BTW, I'm not defending sun. I like Write, but the rest of their software is relatively inferior. Gimp kills Calc, for instance. Sun has made major and important contributions, but we need to watch them. And we are watching them.

I find it interesting that of all the arguments against Stephanie's model, every one is a sophistic side trip into lala land. Can't you prop-symps do better than that?

Yes, I'm an open source advocate, but not for shallow reasons like how much cheaper it is to implement. I work in an environment where support for software actually does come up, and in a number of cases, we standardized on a closed piece of software for the company solely because of the support factor, to my chagrin. To me, free-as-in-speech is infinitely better than free-as-in-beer, though both are good. The problem with free-as-in-beer is that once consumed, it's gone. Likewise with freeware software -- if the source isn't out there, even if it's free to download, it's not free to alter or update for bugs or new OS support.

In your haste to insult me, you've pretty much admitted my other point that Sourceforge is full of flotsam.

If you honestly found insulting my pointing out your strawman comparison of a five-minute Sourceforge "effort" with a legitimate business was laughable, then I apologize, though I'd advise you to realize that attacking the message is not attacking the messenger. In the meantime, the business world is littered with iced tea stand level businesses, they're just not viewable in a centralized database like Sourceforge is. How many local businesses that went under, do you hear about on a daily basis? How about local businesses that weren't local to you? There's a built-in selection bias between what you hear about (the successful, or the mediocre local) and what you don't (everything else) when it comes to the business world. On Sourceforge, the only way to filter what's successful and what's actually useful and continually updated, is to trust their "activity" metrics and filter accordingly. Otherwise you get the firehose of every project ever created.

Actually I have built patches for commercial OSes, notably Solaris long before it was open, but I'm not the best example.

Because you had access to the source code, I'm assuming as part of your contract with them. It may have been commercial, but the source code was available to you at a premium, so you COULD make those changes. That's more open than MS, and led almost directly to Solaris being wholly open eventually -- what part of any MS OS source code do you have access to? Regardless of how incredible of a programmer you are, unless you're planning on dissecting and reverse-engineering the entirety of Windows using assembler, you're not going to have the luxury of editing the code and creating a patch for any of MS' products.

Gimp kills Calc, for instance.

Um... Greg... I don't think they're in direct competition. Maybe you mean Excel kills Calc (which it does, for at least a few reasons, though Calc is all I use personally and I'm finding those reasons are becoming more and more shallow); or that GIMP kills Draw? Though even Draw is more like a cross between a vector drawing program and a poor man's Visio.

PS, I agree with the need to watch Sun, especially now that they're owned by Oracle who, while sympathetic to the OSS cause, isn't nearly as known an entity as Sun was. This underlines the need for forks, and thankfully in the case of OpenOffice, there's at least one excellent one.

Stephanie: still no evidence of your claims, I see.

Greg: I sure hope that wasn't directed at me, because it's a perfect example of Stephanie's "things we weren't even talking about" if so. I've been pretty darn clear about open source as a distribution model, accessibility to source being orthogonal to pricing, etc. The term "solipsistic side trip to la-la land" best applies to the unscientific misapplication of some randomly chosen psychological theories of motivation to software development.

Jason: no, I wasn't insulted by anything about a five-minute demo. I was insulted because that's clearly not what I was referring to. I've already written software which is available one click away on my site - my protocol validator, stack ripper, backup utility, other stuff of even less consequence. It's no Apache or Linux, certainly, but it's more than five minutes' worth of effort and it's already open source. Its existence alone makes me far more qualified than some to comment on open-source programmers' motivations.

All: I have better things to do. You know, bugs and enhancements in open-source software, that kind of thing. "Conversations" with open-source zealots are about as enjoyable and productive as those with evangelists of the more familiar type. Maybe someone who actually knows something about either software or business will show up in this thread, but I no longer care.

Sorry, that was stupid. I meant Gnumeric, not Gimp.

"Itâs better to have a warped sense of humor than a warped sense of seriousness."

Jeff, I don't dance for trolls. Their music sucks, and they don't keep their promises. You can respond to Jason just as well as you can to me if you have any real interest in defending your off-the-cuff claims. That is, if you're not going to take your toys and go home because we people are ganging up on you because you threaten our mass delusions.

Seriously, dude. Bingo.

Jeff, I'm assuming this whole polemic and all the hostility you think you're seeing from me, stems from my saying that passion about a pet project spurs someone to support that project better than money (a fact that I have seen in action far more often than I care to admit in my career, but again, anecdotes do not make data). I suspect you're seeing a level of hostility in my words that I don't intend to convey, though your answers betray a good deal of it toward me and my "religion", as you put it.

I do not disagree that there are people in charge of projects that shouldn't -- Dan's example is one that has bothered me as well, in that configurability is a feature, not a bug. I also do not disagree that starting a business while trying to maintain a relationship and a home life and a current job can be a huge investment, potentially straining to the point where rivets pop out all over the place. Now that I've made my concessions to you, I'm willing to overlook the numerous shots you've taken at my open source advocacy, if you'll just explain to me how any hobbyist's Sourceforge page is different from any hobbyist's iced tea stand. The point stands unmarred from the larger scale you want to put on it -- if you were to start a huge project like Apache, and try to solo it (or even put together a core of programmers to do it, much like a startup with its core of founders), investing all your spare time and sacrificing time at your real job and with your relationship and other obligations, I can virtually guarantee the same strains will appear. So wherein lies the difference? Please, enlighten us.

I've already written software which is available one click away on my site - my protocol validator, stack ripper, backup utility, other stuff of even less consequence. It's no Apache or Linux, certainly, but it's more than five minutes' worth of effort and it's already open source.

My entire point was regarding your "Compare with a mom and pop convenience store that's been around for about as long as your code. The size of the project and its longevity are of little import -- both examples scale up equally. I'm sure each of those tools, with the amount of time and effort you put into them, took a modicum of passion, passion that could equally have been spent building a business of some sort. Since you made the example that anyone could throw together something to put onto Sourceforge and not notice that it was a failed endeavour, I felt the iced tea stand example was an apt example of a minimal-effort business with little downside, that could fail (e.g. by no buyers showing up), to correlate.

I really do hope you answer my example, since you claim to know more about both open source and business than I. If I am incorrect, please correct me. It's edifying to me to learn that which I do not know. That's why I hang out on science blogs despite being a computer guy.

Apparently I dance to trolls too much. Sorry. I'm still sort of finding my way around the blogospheres.

No worries, Jason. Just don't let them get you frustrated, make your points with or without their help, and you're fine. If I had a point to make in this discussion, aside from the one I made on my blog, I might behave differently myself. As it is, I find swatting trolls rather rewarding.

Answering the first question (for myself and others) and ignoring the other stuff:

I prefer free (libre) software. It is what I familiarize myself with, and what I help others with. And so for this anecdote, support for free software is significantly better than that for proprietary software.

I get pretty good support for most anything, but that is an advantage of having friends who can offer it and who appreciate all the shit I do for them. But I definitely prefer open source apps and so do those who keep my computers functional - much easier to tweak shit.