Open Source's Ego Problem

I agree with a lot of what you say, but the problem is that there is an awful lot of ignorance about OSS and Free software spoken. If left alone, these views are propogated to people who may not be as discriminating with their knowledge acquisition as they should be (when using the net). Leaving blatantly obvious FUD lying around can also cause harm to OSS/Free software.

For example, I often come across people who insist that Linux is no good for the desktop because (they say) that Xfree86 can only be changed by hand editing config files. I will sometimes write a brief (and polite) reply saying how it is done in Mandrake (click here, enter password click there, click there etc) and leave it at that, just to make sure that people are aware that inaccuracies exist in a lot of what people say. To be fair though, if someone insists that hand editing is the only way, it is best to walk away, although I am as guilty as anyone of indulging in a fight.

For another example, I posted a reply to comp.lang.python recently to someone who wanted some advice about the best IDE for RAD python development. I recommended Emacs, but also mentioned other IDE's like Vim, Boa Constructor. Somebody investigating Python from a business point of view may read this and instead of thinking "what a bunch of children", they may well come away with something more like "What a lot of choice! And what a helpful community to get support from".

I suppose the important thing to remember is not to take a personal preference as fact. Personally, I find it difficult to stand aside and let ignorance take centre stage, but I suppose the point is to always be polite, remember that my views are mostly only opinions, and that I can be as wrong as anyone can. And to keep some dignity.

Actually, none of this Foo vs. Bar is a problem if it is all backed with solid, factual information. I followed your article all the way up until the very last line, when you lost me. Open source development is fundamentally cooperative, far more cooperative than proprietary development.

If we can decrease the noise and show our cooperative side more, we'll be a doing a lot of good for the image of open source.

I think the noise level is acceptable; the only thing we need to consistently do is (for example) to support every statement in favor of one distro vs. another distro with some hard, cold facts. Give numerous solid reasons why one is better. If we can do this, we raise the standard of our own advocacy to one which is based on technical merits instead of 'marketing spin.' We also have the effect of giving something tangible to the people who are observing the conversation, as in your example above a company which is seeking to migrate from Windows into Linux.

Rather than trying to squelch the noise level, which often only creates more noise, I propose raising the signal level. Which often creates more signal.

Initially upon reading this, I thought it was going to be about Open Source software, not Open Source Linux Distributions. While I agree that there are a lot of camps around everyone's favorite distribution (with 123 separate "recognized" Linux distributions in circulation today), this really doesn't affect Open Source or the "ego" of Open Source at all. Please don't lump together the users and developers of Open Source (and Free Software) with the rubber stamp of "What distribution camp are you in?", because that's greatly misleading.

Let me augment this with where I think the real problem is in Open Source itself, distributions aside: Education.

Since there are more and more "general" users using Linux, and those who believe that Linux can "replace" their existing environment, Windows or Macintosh, they believe that it is inferior in their environment. The problem is education.

I find myself explaining daily that Linux was not designed or created to run as a desktop operating system (nor was it created to be an "alternative" to Windows), though it runs very well in that capacity, and I have been successfully running it on my desktop for 1/2 a decade, every day. When something doesn't work, or has a bug, REPORT IT! Don't just bitch about it and say that Linux sucks (mostly due to laziness on the user's part, failure to look at the documentation, failure to configure things properly, etc.) Also, since you have the source, fix it, or use some other package, or return it for the full purchase price. I get this all the time:

"Hey, I like your software, but can you add [insert feature] in the next release?"

"I'll add it to my list, but there are other more-important fixes that need to get added first. You have the source, feel free to send me a patch.."

"You suck, your software sucks."

This isn't just me, I see this on irc, mailing lists, and at LUGs all the time. We have many more "general" users who don't understand the roots of Linux today than we did a year ago and 10 years ago, they want to just configure it exactly like their Windows machine, even down to the Start Bar, colors, icons, etc. and they believed the marketing dreck they were thrown that "Linux is a better Windows". Here's where the problem lies.. educating users who are new to Open Source, Linux, and Free Software. Case in point, a quote I saw yesterday:

Then the UL people started talking about release schedules, and I asked a real question: How does this whole "release schedule" thing work into the traditional Open Source "it's ready when it's ready" concept? An obfuscated reply followed that didn't answer my question at all. I asked the same question again in slightly different words. This time the response was a little clearer: The Open Source "release early, release often" concept doesn't work in the world of corporate budgeting. Oh. Okay. Glad we got that straight.

Therein lies the real problem with egos, and it's not always the developers (though I've met some "abrasive" developers in my time).

As a developer, my largest problem is with the community at large not understanding how Open Source works, how software is written, and most-importantly, that we do this in our spare time, with no funding (and in my case, being out of work for almost 300 days, and supporting several dozen community projects out of pocket), no documentation from the vendors we interoperate with, and no official support. Many people assume that we, as developers, can just write up a driver for some new, unseen device, in a weekend, for a piece of hardware for which there is no documentation, no support from the manufacturer. This is a huge gap in the understanding of the Linux environment and the Linux community.

In my capacity, I offer the users two types of web-accessible CVS options, a public cvs server, a very robust bugtracker, several mailing lists, irc, and other methods to learn and contribute, and I still see people who refuse to listen, refuse to search for answers, refuse to learn.

"Just tell me exactly what to type, I don't have time to read the docs..."

"Someone just had this same problem a few days ago. Have you searched the mailing list?"

"C'mon, just tell me what to type!"

Oh, and I run Debian on my main development machine and on my server farm, 5 different versions of Redhat in VMWare for sandbox testing of builds and regression testing (along with 12 other Linux and Windows test images for build testing), NetBSD on my firewall, and many other distributions on various boxes and PDAs. Each one has their good sides, and each one has their bad sides. I use what works for what I need it to do, depending on task. I see too much of the "When all you have is a hammer.. everything looks like a nail.." in the distribution and user community.

Let's make a community effort to fix that part first..

Simple human nature. Tell me that Windows, OS/2 and Mac users/developers don't get into exactly the same issues. Word vs. WordPerfect, FrontPage vs. Dreamweaver, Dell vs. Toshiba vs. IBM laptops, etc. Whenever there are choices to be made, people will make those choices and then defend them and try to convince others to make the same choices. I don't think it is realistic to expect the entire open source community to rise above this basic characteristic of human nature.

I agree with most of the article, except for the last sentence on the "image of open source". Seriously, I'm not sure what the `mainstream' image of open source is now, nor do I really care. Probably most people are thinking that open source is something that comes in CDs, is cheap, has a cute animal as its mascot, proclaims itself to be the Holy Grail of computing, and is out to replace Windows, etc. Not many people are able to wrap their minds around the idea of modifying source code -- even though it's a basic tenet of both the FS and OSS philosophies!

To me, what's more important is ensuring that people find open source useful.

By the way, I say Emacs sucks rocks... Let the jihad begin! :-B

...except what you personally prefer.

That seems to be the general approach of some. This is basically a FUD approach, turning your credibility into negative influence. Sometimes it can make you look like an arrogant fool or a narrow minded idiot.

The worst example I can think of is Other Licenses This does not seem to be community building at all, more direction to the zealots.

As hacker pointed out, people are lazy and want simplicity and easy answers. If, for example, someone asks what text editor to use, don't try and indoctrinate into the one holy church of {emacs|xemacs|vi|vim|ed|nedit|jedit|slickedit|etc..}.

Instead, reccommend solutions that will give them the most immediate satisfactory results, solutions with the lowest learning curve. (I would personally humbly suggest not reccommending a tty terminal derived 1970's relic editor of any sort to a new user).

Later on, when they have mastered that, educated themselves some as to that tools deficiencies they will be better prepared to evaluate where to move to next. And having learned something, they might even be in a position to surrmount the learning curve of the new, more complex tool.

In the CVS example, I rarely point people right at the CVS command line client and ssh, and CVSROOT enviroment variables and pserver login, blah blah blah. That's just a recipe for failure.

For a first go, I point them at the daily tarball or the cvs web browse. Usually, they have enough to complain about in building the project and starting to make patches that the addeed complexity of cvs represents a unneeded barrier to entry.

So for a while, the price of getting a new commiter is to support integrating patches thier against the tarball. (yeah they have to learn diff and patch too). Eventually they will have mastered things to the point where they actually get tired of waiting for their patches to be integrated and downloading tarballs and they will start to fiddle with cvs themselves. Shortly thereafter they might ask for commit access. If you force them to jump right to this point right off, chances are it won't happen and you will lose that adoption.

Most software is too complicated and encumbered by UIs (or command line options) that have been conglomerated seemingly randomly over a number of revisions. This seems even more true for open source software. The window manager I use (it came with the desktop enviroment) has 11 tabs of freaking options! It sure makes it tough to reccommend it to anyone.

Factionalization, (gentoo vs debian vs redhat vs slackware vs ...) serves only to increase the preception of complexity.

Ok, so to type up my resume I need only to choose a processor, a chipset, a distro, a desktop enviroment, a window manager and an office suite, no problem. Oh yeah, and I also have to partition my hard drive, make boot floppies, configure my video card, mouse, keyboard, printer, sound, network, and fonts. Did i miss anything? Are there any more choices I need to make?

And for the most part it seems pointless. If, in reccommending solutions, we take the approach that is demonstrated in chipx86's example of choosing Linux offerings, none of the presented arguments actually makes a case for adoption of Linux! (wasn't that the goal).

The expedient answer is the best one.

Thanks for the replies, everyone. They have given me some additional points and perspectives to consider. I expected more of a negative attitude toward this article. However, your replies were anything but negative.

I am going to address some of your statements below.

Alleluia

You're right. If the opinions were backed up with solid data, the issue would be essentially non-existant. The problem is that many people either choose not to give out solid facts, or don't have any. For example, I'm regularly told that my project, GNUpdate, is a waste of time because apt-get already exists. I have an extremely difficult time explaning that GNUpdate is more than an updater, but some people choose not to listen to facts and instead form an opinion and stick by it. Many people, however, do get the concept, and many people also try to stay out of the fights and flame wars. This is not a problem with everybody who is involved with open source, but it is a problem amongst many of us.

hacker

I didn't intend for this to be simply about Linux distributions. I tried to briefly cover software (my example was text editors) without reiterating everything I had said before that. Distributions were a good example, I felt, because flame wars surrounding them are something I see every single day. This problem also exists in the form of Linux vs. BSD vs. Windows vs. MacOS, this program vs. that program, this concept/idea vs. that concept/idea, Open Source vs. non-Open Source, Open Source vs. Free Software, etc.

I also didn't intend to, as you say, lump the users and developers into distribution camps. I tried to make it clear that not everybody takes sides, but many do. It's something I see every day, both online and offline. I personally use RedHat, Debian, and Windows. I plan to use BSD. I know many people who use multiple operating systems and don't play the "This distribution is better than that one" game, but not everybody can wrap their heads around the concept that each serves a purpose, and that it's typically a matter of choice.

I agree with your comments that state that education is important to open source. This really applies to everything, and it's often those who are not educated who participate in the flame wars or spread the FUD.

"Just tell me exactly what to type, I don't have time to read the docs..."

"Someone just had this same problem a few days ago. Have you searched the mailing list?"

"C'mon, just tell me what to type!"

I feel your pain ;) We get this in #gaim on freenode (formally OpenProjects.net) all the time. We stick the FAQ in the topic, ask people to see the FAQ when they have a question, and they still ask us to step them through it by hand. Then they get all upset when we refuse to do so. Ah well, at least there's a good amount of helpful people in #gaim to balance that out.

rasmus

Yeah, it is human nature, and such things have been going on for a long time. However, I have personally talked to people who are turned down by the idea of open source because of ego problems, aggressively stated opinions, FUD, and a lack of facts. I try to explain that not everybody is so, well, pig-headed about their beliefs to the point where they find it more important to convert everybody to their beliefs than direct the person the way they want to go. Still, even though this happens amongst the die-hard users of applications, operating systems, dish washers, and squirt guns, it seems that the Open Source community is a little more known for expressing these things publically in an irritating manner.

I've read articles from a couple of people in newspapers about the problems I expressed in Open Source. I wish I could remember where they were. It's been so long. I think one or two of them were on Slashdot or kuro5hin at one point, but I'm not sure. The fact that an article writer new to Open Source found these problems and made them publically known doesn't help things. Perhaps if people could express their beliefs with facts or, failing that, simply not express their beliefs in such a forceful manner, such articles would never have to be written.

tk

In a way, I'm glad you don't care about the mainstream image of Open Source. I kind of liked it better myself when Linux, BSD, etc. are what the geeks used and not the businesses and schools. However, since things are now moving that way, and a lot of people want Open Source to be spread across all nearby planets, I felt it was important to address the issues I have noticed over the years.

The people I personally know who are new to Open Source pretty much believe that Open Source is Linux. I've tried to broaden their concepts of it. Unfortunately, those that I've talked to whom have gone online for opinions and support often come back saying that one or two people were really helpful, but most of the people were too busy expressing their beliefs and as a result weren't helpful at all.

To me, what's more important is ensuring that people find open source useful.

Couldn't agree more.

By the way, I say Emacs sucks rocks... Let the jihad begin! :-B

You want to take this outside?? ;)

IMHO flame wars is not evil in general.
It is a way to exchange information and some hidden tips about software and it's possible usage.

It is not a way for some manager to get information.

There is magazines and special software companies to recommend software and distributions. Also this service (of creating good software collection) is not free - it costs money (you do it yourself by learning in flame wars and after personal testings or you pay someone to select it for you).

Malx

Flame wars often work people up to the point where most parties leave angry. Maybe not always, but I see this quite often. When the parties involved are angrily trying to defend their beliefs, it's hard to share tips or knowledge. They just don't listen because of the state of mind they are in.

I see people who are considering making the jump to open source ask questions on IRC or in other public forums. Sometimes a flame war breaks out. Sure, maybe magazines are a good way to get information, but that's not the only method that is used.

I'm not sure what your point is regarding cost. I think I misunderstood what you were trying to say there. Can you try rephrasing that, please?

I hear all this talk about how people shouldn't talk about what open source software is better than another comparable one. That view is all twisted and warped out of line.

The simple fact is that a contribution to one project is as good as a contribution to all. Take RAIDFrame in NetBSD for example. Take the source, change some lines of code, and viola, RAIDFrame for OpenBSD. Take OpenSSL or OpenSSH as another similar example.

In fact, I've never been in a single argument where people have gotten so heated that they went away pissed off and wanting to break something. Sure, I'm bet you could find an example, but I don't think that is the proper representation for the bulk of developers or users. I've never heard of Linus getting all pissed off about FreeBSD simply because he thought his kernel was better.

"We" can say "Debian is better because it has apt-get" and support it with reasons that could save time and money, but someone that uses Mandrake Update with RPMs and is comfortable with the system configuration and has never had a problem or reason to change, is going to have a hard time *caring* about apt-get.

I like slackware over debian because I think it feels more like BSD. Trying to convince me to switching to debian because apt-get is easy to use, isn't a good case since what I care about is the more bsd-like configuration.

Rational people realize other people have differing opinions. Since people will always have differing opinions, we will always have lots of software forking, project copycats, and 500 different IRC bots. As long as people have curiosity about other programming languages, we will also have people making new things that will have security holes and new ideas.

Open Source == Programming Hobbyists. As hobbyists, they will do what they want, when they want to, and how they want to, even if it means having differing methodologies and ideas. It's not fair to make that sound like a crisis.

I hear all this talk about how people shouldn't talk about what open source software is better than another comparable one. That view is all twisted and warped out of line.

The simple fact is that a contribution to one project is as good as a contribution to all. Take RAIDFrame in NetBSD for example. Take the source, change some lines of code, and viola, RAIDFrame for OpenBSD. Take OpenSSL or OpenSSH as another similar example.

In fact, I've never been in a single argument where people have gotten so heated that they went away pissed off and wanting to break something. Sure, I'm bet you could find an example, but I don't think that is the proper representation for the bulk of developers or users. I've never heard of Linus getting all pissed off about FreeBSD simply because he thought his kernel was better.

"We" can say "Debian is better because it has apt-get" and support it with reasons that could save time and money, but someone that uses Mandrake Update with RPMs and is comfortable with the system configuration and has never had a problem or reason to change, is going to have a hard time *caring* about apt-get.

I like slackware over debian because I think it feels more like BSD. Trying to convince me to switching to debian because apt-get is easy to use, isn't a good case since what I care about is the more bsd-like configuration.

Rational people realize other people have differing opinions. Since people will always have differing opinions, we will always have lots of software forking, project copycats, and 500 different IRC bots. As long as people have curiosity about other programming languages, we will also have people making new things that will have security holes and new ideas.

Open Source == Programming Hobbyists. As hobbyists, they will do what they want, when they want to, and how they want to, even if it means having differing methodologies and ideas. It's not right to make that sound like a crisis.

May be I just meet different people. My expirience is about russian/ukrainian speaking community.
Indeed it is bad if people stay angry after flame wars. But here it is rarely a case.

As for people asking in forums - I whouldn't count on them. If they can't find more realiable source of information (magazine, sites, revies, even guru to help them) then they are not valuable for open source. I whould thing they even submit usefull bug reports. That means - it is their bad luck to ask there those questions ;-)
That is IMHO again.

In fact, I've never been in a single argument where people have gotten so heated that they went away pissed off and wanting to break something.

I guess different types of forums have different demographics. A quick example (or counter-example) of a noisy forum is, well, you-know-what. Maybe it hasn't caused people to start breaking things, but it's so noisy that it doesn't matter.

However, even within e.g. Usenet, there's considerable variation. comp.os.linux.misc seems to have a reasonable signal level. In contrast, comp.unix.advocacy has a lot more junk in it, probably due to the charter of the newsgroup.

It may be interesting to study the dynamics of several public forums, and form some conclusions.

As hobbyists, they will do what they want, when they want to, and how they want to, even if it means having differing methodologies and ideas.

That doesn't seem to be the problem chipx86 was talking about.

perhaps, part of the reason is that, most seasoned open source people prefer coding to talking...

"we're doing it for our own amusement, remember?
 who said that we wanna conquer the world?"

Yes, and that is always important. If you don't enjoy your own work, there's no reason to do it. I consider myself a seasoned open source developer, and I too prefer the coding over talking, but that doesn't eliminate talking.

Remember, though, that not everybody in the open source community are developers. Many of them want to see Linux take over the desktop market. Not all, and not everybody even believes it can, but it's still a dream of many people. Anybody who has said, "You should use Linux/BSD/etc instead of Windows" has contributed to this dream by trying to convince one more person to switch. So, the above still applies. Many people prefer the talking over coding, and that will always be true. Getting a message across about opinions, beliefs, etc. is only bad when it's done in a negative way. Flame wars, holy wars, etc. are negative, and often provides more anger than information. That's the point I was trying to make in my article.

Part of the misunderstanding I seem to getting is the expectation among members of the congregation in reaching outside the programming space. This urge to witness beyond the border of the believers and into the realm of the grandmother stuff has somewhat become an obsession. To the point of posing a Jihadic stance toward the unbelieving masses.

I too am faced with this reality of having to deal with something that can't be controlled. This thing called 'choice' among the masses appear so distant, yet remain in our midst. Could it be considered already a mystery? Maybe it is.

Nobody is expecting us to shake the sands off of our sandals when endusers don't put out the welcome sign. It all can probably be simple enough to mean the masses do not have to the time of giving something back, though. Why shake the sandals when the fact nothing is actually at stake. The risk involved in any direction is almost zero. So the question among us is probably the words "Why do it?"

Nonetheless, for whatever it's worth. The road to Open Source and Free Software will always be there for everyone to use, but only if they choose to.

Note: pardon the prophetic tone because I'm currently in free writing mode. Have fun.

we should try to _explain_ rather than _convert_. let's make sure that the way of open source isn't misunderstood... and those who destined to be with us will follow. after all, OSS isn't for everybody.

chipx86

yea, spreading out the correct words is our responsibility. i personally do the 'talking' by showing others my notebook running linux (debian, btw) without bashing any parties, when asked about linux or open source in general.

this subject is old and shouldnt be brought again .

I agree with much of what chipx86 said. Religious wars are bad. Religious wars divide communities. If someone asks about switching to Linux, then the answers should be about Linux in general. Of course, a poster can point out the advantages of a particular distribution but it should be a factual explanation. It does not help me to know that you think Debian is better than Mandrake if I don't know why you think so.

As hacker said earlier, it is important to educate the users, to set their level of expectations for your project appropriately. Over time, this is going to be incredibly frustrating for the developers because so many people do not search for information. You can choose to blow these people off with a curt "RTFM!" but I think this sets a bad tone for a project and in the end I think it limits the success of the project. I think it's better to educate users and to encourage the ones who are capable of it to at least answer questions and to write documentation. A crew of enthusiastic users who can answer questions can do a lot to reduce developer frustration.

The trade press and people in corporations are used to dealing with authoritarian organizations, that speak in clear, coherent messages because there is a central corporate marketing department or press officer that makes sure that everything that is said is consistent. Some open source programmers are so lame that they actively attempt to ape the style of corporations, putting out press releases full of corporate jargon as if this is going to impress someone. In case you haven't noticed, respect for traditional corporate norms has recently taken a nose dive; this is a bad time to try to look like some clueless dot-commer marketroid.

The free software/open source community has many voices, because it is composed of many independent people with different points of view. The kinds of bitter arguments they conduct in public are in many cases no different thant the kinds of arguments that occur internally in the better proprietary software companies (though there are also plenty of companies in which frank speech gets you in trouble internally as well). Because of this open discussion, a trade press reporter has to do substantially more work to adequately cover the GNU/Linux/BSD/Mozilla/etc. world; he or she can't simply rewrite official press releases. A reporter that gives a damn can actually get at something resembling truth because of this. She can learn that feature X is the way it is because this was actually thought out and debated, or else determine that no one ever considered the possibility, by reading the development lists. This keeps us honest, and that's a good thing.

Most people aren't used to this kind of openness and it makes them nervous. They call for a dictator, they call for speaking with one voice, they call for the community to assemble itself into a Borg. I see this several times a day: a call for some big company to just force Linux to look more like Microsoft. They fear letting the world see our internal debates, because we might look bad. My advice is just to get over it. Yes, your flames will be out there for the world to see, but so will all of your constructive work, so those concerned over image should just make sure that there is more of the latter than the former. As for vi vs Emacs, KDE vs Gnome, the arguments are no big deal and to a certain extent they are constructive.

The reaction to users who demand they get easy help or demand stuff gets fixed for them can be very friendly and very easy:

"Just tell me where to fax the support contract"

I've had good experience explaining users that if they wanted to demand anything, they should pay me for it. If they didn't want to pay (the normal case) they should be happy with whatever help they get.

By staying friendly and explaining the "harsh reality" I've educated quite a few new users ;)