Tuesday, April 05, 2005

Satisfying ways of doing nothing

I, the owner of a computer inhabited by M$ software of doubtful authenticity had recently realized that M$ is a monopolist. Since it is a monopolist, it can't be all that good - as there is a loss to the society whenever there is a monopolist around. Elementary economics tells us so. Irked by this deduction, I decided to spend all my idle time (which could mean my entire life) on figuring out the intricacies of alternate software distributions.

There are many Linii around. There's the ubiquitous Mandrake - and the tribal Ubuntu or the impossible Debian - not to mention Fedora, Gentoo, Vector Linux, Suse, Novell....... Installing each of these versions is a rather challenging affair, and a nice way to spend half an hour.

The problem with most Free/Open source software is, of course, the geeky documentation and the rather steep learning curve. M$ Word is intuitive and easy; LATEX is impossible for a person who does not have a background in C/C++/Java/Vi. Oofice, if it ever opens, is hardly a replacement - and MPlayer needs to be compiled from source. The Samba server is benevolent: it reaches a state where you share data with others easily - but getting data from others is almost impossible. All these problems are not problems, actually. They were designed that way. Onky the perseverent geek shall survive.

It isn't that extra little bit of functionality that people use Free Software for. It's for the fun in learning it. Writing a technical document (My Master's Thesis) in Word is what you would call a bore. (Usually, anything you have to do with a mouse is a bore - such as equation editor - perhaps with the exception of a gem of a game called 'Copter'). Add to this ennui of writing equations using equation editor, the ennui of putting references and figures in order.

That's why I turned to LATEX. Though I am almost positive that it shall take me less time to compile the entire thesis in M$ Word, the intellectual thrill of composing a document in TEX bears no M$ paralells. Word is like building a mansion and living in it. TEX is like building a hut (poor ol' vi!) and then finding out (after compiling and looking at the pdf) that the hut is actually a palace. It is this thrill that has me hooked on to TEX. Not common sense. Just a drug-like addiction.

We do have this amazing program called Fluent (which assumes an almost biblical status to most Master's Students here at IIT Madras) which solves the bitch (the N-S equations) with all sorts of turbulence models thrown in. But people still like to write their own code - to get the same results. This often takes them more than a year - and a lot of agony. But they march on - criticism notwithstanding. After a lot of trials and tribulations they try to compate their results with the same Fluent. They fine tune their code till they agree with Fluent. The thrill of getting your own method to work is amazing. But doesn't it border on masochism?

Coming back to the point: There are a lot of times when you can do a lot in Linux: you can work very hard; spend a very satisfying morning working on trying to get something to work; learn about the beautiful intricacies of OS design but your actual problem is something else. Linux is an amazingly satisfying way of doing nothing. Most students do nothing anyway; so they love Linux. I do too.

No comments: