Tuesday, December 20, 2005

Bruce Eckel's post, and Ruby/Python dialogue

Firstly, I'd like to say that I would like to see more positive dialogue between Python and Ruby users and developers (This isn't just because I would like to see Ruby steal some smart Python people via a process of assimilation). The Ruby community is usually well known for its open-mindedness and general niceness, I don't understand why some members have a dislike of Python (I personally prefer Ruby, but Python would be my 2nd favorite language). I don't think superficial Python critics ( e.g. "significant whitespace, wtf? must be a bad language") are representative of the Ruby community as a whole (I hope not), and similarly I don't think superficial Ruby critics ("influenced by *Perl*, wtf? must be a bad language") are representative of the Python community. In this spirit I should probably take Bruce Eckel's recent hyper-enthusiasts post as a welcome sign that more and more prominent programmers are taking notice of Ruby (and Rails, of course), and ignore the labels and generalisations he applies to Ruby and Ruby users.

However I do take some issue with Bruce Eckel's post (My point is similar to the one David H. Hansson makes in his post, but he makes it in a more outspoken style). Eckel's main criticism of Bruce Tate's book is that he dismisses languages without taking the effort to learn them. In Bruce Eckel's post, and especially in his previous post some years ago, which I clearly remember (the rest of the post was quite similar to the quotes that remain in on ruby-talk IIRC), Eckel seems to criticise and dismiss Ruby without showing that he has taken the time to learn it. First time round, this was the paragraph that I remember most strongly:

For some reason, the creator of the language saw Python and decided to do a clone, and people who had never used Python thought it was a good idea. Harsh, maybe, but that's my impression: if you've used Python at all, you wouldn't give Ruby a second glance.

This time it seems to be along the lines of "there are influeces from Perl, ergo it must be a bad language". So the problem to me is that he comes across as the pot calling the kettle black.
The fact that he criticises different parts of Ruby this time around, without acknowledging that he criticised Ruby before, and that some of these previous criticisms were somewhat rash and unfounded, is a secondary issue, which I think we would all be willing to overlook if he wrote a second post, critiquing Ruby from the perspective of an example script he's written, or some evidence that he's taken the time to learn Ruby and to try to think in the language, something he should be quite good at doing :-) Of course, if he suddenly got excited about Ruby and decided to write "Thinking in Ruby", that would be fantastic. Stranger things have happened ;-)

Monday, December 19, 2005

Will the "long tail" start wagging the dog?

The title of this post questions whether long tail style information dissemination techniques, such as Google searching, blogs, and wikipedia, may start influencing opinions (of at least the net savvy), more than mainstream media, whose influence seems to have been on a steady rise since the invention of the television. Chris Anderson, editor of Wired Magazine, has some interesting things to say about these new technologies and approaches to information dissemination, which rely on the fault tolerance of our brains toward information sources (i.e. we look at multiple sources in the course of making an opinion, if one is a bit wrong or inaccurate, that is OK), in The Probabilistic Age.The main issue in my opinion is getting an accurate view of a topic or issue. The problem with blogs and google searches and wikipedia is that some of the information may be biased or inaccurate. The advantage of these new forms of information dissemination is that we can sample multiple times from this sea of information to get a broad view of the facts, and our fault tolerant brains can weed out most inaccuracies and bias. The problem with mainstream media is that it usually conforms to one or two simplistic viewpoints. It is less likely to be factually inaccurate, but it is still usually biased. The problem is that with the mainstream media (I'm mainly thinking newspaper, I can't really talk about television as I don't have one) the biases all reinforce one another, for example if we always read the one newspaper, we may read five articles on one topic which all agree with one another, therefore we think that opinion is the truth. We get less diversity in opinion, due to editorial pressures, and less diversity amongst the people giving us these opinions, compared to say blogs or the results of a google search. The advantage of the diversity search results and blogs is that we have more opinions to choose from, and the biases tend to cancel each other out (we get an averaging effect).

As an example, imagine we are interested in a topic such as "What are the facts regarding the Schappelle Corby case? (this will only be familiar to people from Australia, sorry). By reading blogs and searching the web, we are exposed to say five articles which express variations on three main viewpoints
  • "She is innocent, the drugs were planted on her, this is so terrible!!!!!"
  • "I think she is innocent, Indonesia sometimes seems corrupt, however we don't have much information"
  • "She had 4.1kg of marijuana in her boogie board bag, in a specially designed plastic bag that was cut to fit the outer bag. She allowed officials to search her bag, but got upset and refused when they asked to look inside the compartment holding the drugs. Her brother was in prison at the time she was arrested, and has previous drug convictions. Make up your own mind."

Of these five articles, one is excellently written, two are well written, and two articles are poorly written. We are still much better off, however, than if we are exposed to five articles which are all reasonably well writted, but with some of the immediacy and frankness edited out, and all conforming to minor variations of only one of the three viewpoints (say maybe "She is innocent, the drugs were planted on her, this is so terrible!!!!!", because it is believed that this is what their audience wants to hear). If this mainstream media viewpoint is not accurate or only contains part of the information required for us to make up our mind, then by relying on mainstream media we can get a distorted view of reality.

Of course, what I have discussed here is only part of what Chris discusses in The Probabilistic Age, and The Long Tail, which discusses some similar issues. A further article worth looking at is Web 2.0, by the inimitable Paul Graham, especially the section on Democracy which talks about the damping effects of editors (getting rid of the extremes of good and bad writing), and the Common Thread section which insinuates that this new style of information dissemination (Google searching, blogs, Wikipedia) is nothing more than using the web in the way it was meant to be used.

Sunday, December 18, 2005

Thesis finished

The final version of my thesis "Retaliation and Reconciliation Conventions in the Repeated Prisoner's Dilemma" has now been passed and submitted to the University library for posterity. If you want a copy of it (of course you do), here it is.