Mellel and MacIntel -- Many... Well, some... Well, at least two of you asked whether Mellel will run on the new MacIntels -- the future Macs that will run with Intel's processors. The answer is clear: Mellel already does. With pre-release machines using this processor, Mellel is already running quite nicely. We estimate that once those machines are out, we'll have a Mellel version ready within a few days.

A few words about Mellel 2.1 -- Yes, we're already working on it. One thing we've promised will be in this version was already announced: Mellel 2.1 will adopt the open and free XML standard.

This move means two things: First, Mellel files will be accessible as plain text, for future use or archiving. Second, the open format will allow those who are interested to easily convert, reuse or manipulate Mellel files (with or without Mellel).

All this is very important, but also under the hood and, frankly, quite boring to those of you who aren't interested in those options. To nip this boredom in the bud, we'll add quite a few new features. I really can't tell what those will be, but I think you'll find them quite interesting.

Ah, and to answer your question: Mellel 2.1 will be a free update to all registered users of Mellel.

Just in case you've missed those... -- Just in case you've made the super-human effort and managed to avoid our constant marketing nags (i.e., this tri-monthly newsletter) -- take a look at what I think are our best offers:

• Mellel 5 Pack: Just US $69 for five licenses. That's less than US $14 per copy. If you need more than one, but not that many, it's the thing for you.
• Buying in volume: When five isn't enough, there are excellent offers for volume licensing. Although Mellel's price went up with Mellel 2.0, our volume licensing actually went down. You can get a copy for as little as US $11 per copy. It's worth a look.
  
  
• Mellel 5 Pack
  Volume licensing
  
  

What's going on

The thing about Wikipedia -- Wikipedia, the free encyclopaedia, is in deep, deep trouble. Recently, a former USA Today editor, John Seigenthaler Sr., complained that the free encyclopaedia falsely claimed in an article about him that he was involved in the assassinations of US President John F. Kennedy and US Senator Robert F. Kennedy. The anonymous accusation remained unedited for 132 days, until it was removed.

The incident led Wikipedia to change its policy, preventing unsubscribed editors from creating new articles. Wikipedia's founder, Jimmy Wales, worked hard to control the damage by redefining Wikipedia not as an encyclopaedia but as a simple source for "background information... [that's] a source for initial background information more reliable [than] random Web pages on the Internet."

But the new limitations and redefinition do not solve the problem. In fact, by masking the real issue, they will probably aggravate the situation further. Wales portrayed Wikipedia as a source for sketchy background information and nothing more. But this is not how the public, the press, and certainly not Wikipedia's volunteer editors see it. If it calls itself an encyclopaedia, has articles, and compares itself to Encyclopaedia Britannica, then it is an encyclopaedia, and should be judged on its merit as such.

Wikipedia claims that there are three very good things about it: it's updated often and quickly, it's comprehensive in terms of the number of articles and their length, and it's constantly improving.

The basic premise of Wikipedia is the "Linus Law," so named after Linus Torvalds, the creator of Linux: "Given enough eyeballs, all bugs are shallow.'' That is, just as an open software project gains in quality as more beta-testers test it, so will the quality of the articles in Wikipedia get gradually better and better in time, as more and more editors contribute to them. "Now suppose this project lasts for many years," fantasise the Wikipedians, "how could articles not be burnished to a scintillating luster?"

The problem is that "Linus Law" only applies to software source code, and even there its veracity is dubious. Most open source projects are managed and controlled by a fairly small group of programmers, that write most of the code and fix most of the bugs. Most beta-testers do just that -- test. The code is written by a select minority.

Wikipedia doesn't work this way. It's non-hierarchal nature means that in principle every edit is as good as any other edit. In other words, it function as a committee without a chairperson. Usually with committees, the overall quality of the product tends to drop the larger the committee and the longer the deliberations.

Without anyone managing the discussion, the committee also tends to disintegrate into small groups of quibbling parties that can never reach a conclusion. Wikipedia tries to stem this drop and even reverse it by encouraging a non-formal hierarchy dominated by "gardeners": a group of very active editors and contributors. That is, form within the non-functioning committee an informal sub-committee that does function. The problem is that this group is selected not on the basis of merit and quality of writing, but on the basis of being active and dedicated to the cause. But spare time and writing ability are not the same thing, and enthusiasm cannot replace a sense of direction and vision.

And then there's the size thing. Lacking a chairperson or a clear direction Wikipedians opted for the easiest solution of adopting the faith that bigger is better. The more articles there are and the longer they are, the better. Wikipedia always takes care to mention the number of articles on its home page, reaching significant numbers of edits and articles are celebrated, and there are constant "crusades" to eradicate (i.e., add more content) to 'stub' articles.

Currently, Wikipedia holds more than 850,000 articles, more than ten times the number of articles in Encyclopaedia Britannica. This means that either the editors of Britannica got it horribly wrong, or that Wikipedia is a huge compilation of rubbish articles and non-encyclopaedic factoids. Get to Wikipedia and press the "Random Article" a dozen of times and you'll see for yourself which is which.

But having a bad signal-to-noise ratio and badly written articles isn't the problem, it's just the symptom. The real problem is that the collective equitycracy concept simply doesn't work. Some measure of meritocracy is urgently needed.

• Wikipedia
• The USA Today article that created the current controversy
• Business Week interview with Jimmy Wales
• Why Wikipedia is go great