Bad, vile and meaningless: McHenry from Alan's clob

I'm going to ride on a high horse for this blog entry.

LWN carries Robert McHenry's response to Aaron Krowne's article "Fud-based encyclopedia". That article in question is an attempt to respond to McHenry's earlier article.

For better of worse, humans are community creatures. They are even slightly irrational when their community is threatened -- just look at the attack against McHenry, particularly read this awful piece prepared by Krowne. There is really no reason for the attacking tone towards what was reasonable criticism. It had a well-argued point.

Let's start with the following gem of wisdom that was snipped off the Wikipedia's editing policy document:

However, one of the great advantages of the Wiki system is that
incomplete or poorly written first drafts of articles can evolve
into polished, presentable masterpieces through the process of
collaborative editing. This gives our approach an advantage over
other ways of producing similar end-products. Hence, the submission
of rough drafts should also be encouraged as much as possible.

This is essentially the same mindset that open-source software gets written in -- just make it do something, someone else will come and pick up the pieces. It's also the same mindset with the guy who wrote readdisk (see previous entry) and with me. Sure, doing it was easy, but probably someone should have sat down with the program and actually written it up properly: symbolic links, hard links, proper permissions, the whole shebang. For the good of society as whole, better have one guy spend pains to make a great app than have plenty of morons to make small random changes. Or is it really so? Here is the crux of the matter from mr. McHenry:

In other words, the process allows Wikipedia to approach the truth
asymptotically. The basis for the assertion that this is advantageous
vis-à-vis the traditional method of editing an encyclopedia remains,
however, unclear.

I believe this is greatly related to the same issue that I ponder greatly, too. Why is it okay to spend, say, 5 minutes of your time to tug some random fact in the world a bit better, even if you perceive it all to be a huge mess in general. Shouldn't you try to tidy it up? Take control? I can only surmise that it comes down to each individual person's nature, which is of course a general cop-out that doesn't yield to any further insight, thus useless.

I think the easy-go-lucky structure of open-source development, of blog writing, even when it produces the quantities of garbage it does, is perhaps still salvaged by the process involved. Here's the key point one: it takes only a little time to do each day. Which means it allows this big community to randomly tug matters towards their own ends, and therefore perhaps generally improve the whole over great time spans.


Consider that it took me just an hour on two consequtive days to salvage the bits from old readdisk to make a new one that was perfectly useful to me. It was easy and even fun, even if did not produce an application I would attach to my resumé. We programmers do not have shame from releasing our open-source turds to the world -- particularly when we see that the next guy's turd is sometimes even less well formed and smells perhaps even worse. Anyway, I see no particular reason to expend great effort for what will be quite temporary. A few versions from now, Linux Kernel will again support affs, and the readdisk program will be a lame redundant application again. This is as should be.

So, what we are doing is not very professional, at first, but it is something we all can do. This is how you make communities. But! I do believe that open projects should support beneficial forms of self-organizing such as meritocratic election of editors with greater powers or ability to decide fate over entries in case of Wikipedia. However, as this raises the barrier of entry, it should only be done later in the life of a project, preferably as soon as some significant minority of the early smart guys recognizes the necessity for system's better control. Naturally this all depends on your values, for instance, whether you prefer quantity over quality, consistency and cohesion over breadth, and so forth.

Strcuture and control seems evil to those who do not understand the need for it. But successful projects, in the end, become structured to deal with their growing complexity. That is a fact to contend with. And evolution has only one way up, and that is towards greater complexity.

Finally, I'd like to clear up what I believe McHenry intended with the final passage:

The user who visits Wikipedia to learn about some subject, to
confirm some matter of fact, is rather in the position of a
visitor to a public restroom. It may be obviously dirty, so that
he knows to exercise great care, or it may seem fairly clean, so
that he may be lulled into a false sense of security. What he
certainly does not know is who has used the facilities before him.

A badly chosen analogy, as mr. Krowne responded with:

McHenry is essentially asking us to suspend all higher brain function
at this point: in the above metaphor he simply pretends that the
reviewability features he just based his entire article on do not

As I see it, the reviewability is not the point. The point is accuracy, and you can't guarantee accuracy from being able to review the history of changes. However, there may be some clues. If the entry does not meet professional standards in its style, you know it probably isn't very accurate. But suppose someone not knowing anything of the issue comes and cleans up all grammatical mistakes and so on, hiding the trash under the carpet. Now you see a perfectly well-written entry whose content is nevertheless bogus. You will not be able to determine the accuracy of the entry by viewing its history. So I believe that some form of authorityship needs to be endorsed; you need to be able to elect experts that can be trusted to know the matter they are writing about.

Remember that the average Internet user is just a moron and unless you raise some barriers, you will tend to this mediocrity.