“SUNNYVALE, Calif., August 28, 2008 TWIKI.NET, the leading provider of open source enterprise collaboration solutions, announced a new release of open source TWiki.” is how the latest from the propaganda-machines at TWiki.net starts. Although not explicitly stated, it reaffirms the suggestion that open source TWiki.org is owned and created by TWiki.net. This is, however, far from the truth, as there has been no significant code contribution coming from the company whatsoever.
This recent outburst of unfair PR (but also requests from certain people) has prompted me to put online an english translation of a blog post I wrote on the subject early july of this year (on my own blog and on livre.nl). I am still an avid user of TWiki, because its potential is great. This means I find bugs, from usage on my own sites or one of the many hosted instances I mantain for my customers. I stopped active participation in the community, but so far did report any serious bugs. I think it is time to put a halt to that too: if TWiki.net wants me to work on what they consider their software, well: let them pay me!
So, here is the translated blog post, for the international audience to enjoy (it led to some nasty business where an innocent bystander was accused of being my sock-puppet, so it is some controversial material kids :).
Open source and commerce: how it can fail miserably
“Aforementioned projects, such as PostgreSQL and XWiki, have shown to benefit from an ‘enterprise’-version. I don’t see any reason why TWiki would be different.”, I wrote almost a year ago on Dutch open-source news site livre.nl. That optimism has been replaced by skepticism. Since TWiki.net entered the stage, TWiki went into a deep crisis.
TWiki.NET (the commercial off-shoot) rolled over the opensource project like a tsunami. TWiki.NET would strengthen the commuity, but forgot to respect existing structures. Without hesitation their marketing-department came with slogans that claimed work of the many volunteers that contributed over the years as their own.
The latest release is version 4.2, which is half a year old. In the mean time several bugs have been discovered, some of which threaten the security of TWiki installations. The long-awaited patch-release, 4.2,1, is seriously stalled [at the time of writing this article originally, this was true, in the mean-time two new patch releases have been released]. The best example of why not to use TWiki is the project website itself: the site is slow, confusing and, according to many, simply ugly [since writing this article, a new server has been installed]. Core developers are quitting. The community is crumbling.
When you start to dig in TWiki’s history, by looking in the project-wiki on twiki.org and by talking to the people who have been involved for years, you discover that all of the above is not new. The project has been limping on like this for years. Only a small number of programmers and contributors who developed the TWiki code and know it inside-out lasted longer than, say, 2 years. Others couldn’t reconcile themselves with the power-structure. The few ‘old-timers’ can tell endless stories about the people that came, got stupified and left.
What is missing is an inspiring leader. What TWiki has is a self-appointed ‘benevolent dictator’: founder Peter Thoeny. In itself, there is nothing wrong with having a benevolent dictator. Many open-source projects know such a person: linux has Linux Torvalds, ubuntu has Mark Shuttleworth, PostgreSQL has Tom Lane, perl has Larry Wall, and so forth. The significant difference is that they earned respect, they inspired people and thus earned the title. I am of the opinion that when you have to assert yourself ‘i am the benevolent dictator’, as Thoeny does, the predicate ‘benevolent’ can not apply anymore.
Now that next to leading the opensource project Thoeny has an enormous financial interest in TWiki.NET, the borders between both are fading. An extra handicap is that Thoeny is the only link between open-source community and the commercial entity. Looking at Thoeny’s public statements on (what is in his perception) his opposition, one fears for the image the employees of TWiki.NET have developed of the dedicated members of the open-source community. False accusations and blatant lies paint a bleak image of these persons.
What does the future look like for TWiki? TWiki’s codebase is complex. New developers will have to walk a steep curve when the old-timers really quit the game. Discussions about the power-structure are ongoing, for many of the old-timers a proper resolution of that dillema is an important condition for getting back to business. The discourse is, however, progressing slowly and has to be framed in velvet words, because there is no room for critical remarks (did someone mention dictatorial?).
Of course, TWiki.NET can pay or even hire developers. Create a fancy corporate version, of which a usable but less feature-rich version will be available as open-source. The MySQL model. There’s nothing wrong with that, but for me personally it would be a lot less interesting to actively participate. Who knows, from time to time the f-word is mentioned: a fork would be a way out of the current impasse. I am reluctant to acknowledge this option, because not often does a fork lead to a happy end.
Fortunately TWiki is not unique (although potentially it could be). There are other solutions for the problems that TWiki is an answer to. Solutions maybe more powerful in certain areas, solutions that are certainly more user-friendly. I would be dissapointed if TWiki doesn’t develop to fulfill the promise, but I have the freedom to use alternatives. There are people that have specialised exclusively on TWiki. They will have a tough nut to swallow when TWiki does go bust.
I think that currently the most probable outcome is the MySQL model, where the existing community is replaced by a paid community. Of course, there is again the possibility, like my original article on livre.nl, that i’m all wrong and things will work out in the next months. Unfortunately, i’m skeptical.