Why The Community Of XXX Failed/Fails

I just read the current topic on Slashdot — Why OpenSolaris Failed To Build a Community –, and originally, I wanted to post a short reply only. But as I look at other failed .. failing communities, such as Mambo, Asterisk, Trixbox, this became a bit longer. Therefore I prefer to put it as a post here.

The company I am working at has clear guidelines which free software to use: it needs to have lots of installs, “a community” and — even better — the application is backed by a widely known company.

So, what’s been on their plate? First of all, Joomla!. That’s a CMS forked a while ago from Mambo. There’s a company which backs Mambo, and there’s a huge userbase for Joomla!, so there’s also a lot of forums (read “community”), so you get lots of hits when you google for Joomla!. Actually, Joomla! 1.0.x was (also) known for the security holes it featured.

Second, Asterisk is on the plate. It’s backed by Digium, which even some dead-tree business magazines learned about. There are “thousands, millions” of installs of Asterisk, so there’s no chance Digium would sell its customers crap. Not to mention CallWeaver or the reasons for the fork.

Now, I learned we were about to make use of Trixbox, a custom-made Asterisk Linux distro. We were about to use it for the same reasons as the two others. What I learned next was that Trixbox was a fork of Asterisk@Home, then that the company behind it managed to turn that free software package into some kind of trojan horse. However, that series got me thinking.
 

A great share of Joomla’s community apparently consists of people who sell software — add-ons to core Joomla!. Like on a bare-bone Windows, Joomla! had not that much of features as one might have hoped. And apparently the greater share of Joomla! developers hoped to make a buck, so the smaller share actually contributed free add-ons (free as in GPL). Which (as I see it) effectively did not help the quality of code of such add-ons. So, the community of Joomla! mostly consisted of developers who wanted to make a buck, and of users who had their issues.

By looking into the news of CallWeaver, I learned [Update: but am unable to provide evidence for, currently, though yet got a denial of Asterisk Project Lead Russell Bryant for the following note. Therefore I'm striking it out:] they forked from Asterisk for such issues as that Digium did not accept patches to Asterisk which enabled Asterisk users to make use of voice telephony cards other than those of Digium. [Update: Anyway, read the introduction to CallWeaver on voip-info.org and figure yourself.] In other words, Digium was after making money not of the software — which is free (as in GPL) — but by tying that software to hardware — the hardware they produce and sell. A classic approach of lock-in. Effectively, that restricted the users of that great piece of free software instead of achieving that what free software tends to achieve, to serve the user.

Fonality is the company in the back of Trixbox. Probably everyone has heard about that company yet. I admit, here I didn’t dig in that much, but what I got from comments on the trojan horse issue the developers made out of Trixbox is that a group of folks entered the Asterisk@Home community and turned it to the worse. The trojan horse case were only one aspect of that, the commenter wrote. A praised alternative to Trixbox is PBX In A Flash. Since a trojan horse probably is not what the user wants, here too the product is something directed contrary to the user.
 

What I see in each of these three cases is a group of people who try to immediately rip a buck out of the software, either by selling add-ons or by locking users in to hardware they sell. Effectively, that is contrary to what the user expects from a fine piece of software; so a fork might be inevitable. Maybe that’s even a rule, simply because developers of free software often are users too; and as the sources are free, they are free to change the poor experience they get by crippled software.

Actually I get alert when I learn a free software project gets backed by a company — since there’s simply no need for money to be thrown upon free software projects: If people find a need to improve anything by software or to improve an existing piece of software, and the software is free, and they have the time, they simply apply that fix to the software. No need for throwing money then. So what alternative reasons may be there to throw money onto a free software project? — Get money out of it…?

To make the point: I think, people who want to make money out of free products often simpy don’t get the idea. To cripple a free thing does nothing but causing a fork, sooner or later. Techdirt has explained over and over again, that you need to and can embrace the free goods in a way that serves not only you but also your customers.

My opinion on the initial measurements — community, company back-up, lots of installs — do not barely say anything about the quality of software you get. But if you dig into the suggestions and opinions of the community, especially on the developers’ side, you might get a close and true look into what’s really going on there. However, people still buy snakeoil. (But, this is a personal note, it’s good to see that people who sell snakeoil blunder into buying snakeoil as well. I just saw.) [t]
 

Update: As I had to fix that bit Russell objected against, I took the chance to smooth the post a bit. Ruled out some grammar and copypaste mistakes, emphasized what’s my opinion where it might have looked like a claim of facts. Also, I added some more links to make things become more clear and to provide evidence where possible.