Hacker News new | ask | show | jobs
Claude Shannon -why prefer the bit to the dit, nit and Hartley? (i-programmer.info)
17 points by mikejuk 6113 days ago
4 comments

Father of Information Theory is a little-known character ? I think not. His master thesis is the "most significant masters thesis in the 20th century".
Shannon is only a little-known character if one has been stranded on a desert island for the last 70 years, without any contact with the rest of the world...
Sadly that applies to a lot of the general population, I would say it even apply to several people having a degree or working in fields related to computer science.
Maybe this is due to the fact that Shannon's work is quite theoretical. Sure, his work had an immense impact, but it's "behind the curtains", so to say. The point I am trying to make is that Coding is an enabling technology, i.e., it serves to enable reliable communication, but it's a means towards an end, not an end in itself. A lot of people think communication is about two antennas and something floating around in the air is transmitted between them, and somehow the receiver is able to guess the message that the transmitter sent. Well, the transmitter can do that because of coding schemes, but that's too technical for most people.

Another field which suffers from the same problem is Controls. It's also an enabling technology in the sense that it allows man-made mechanisms and devices to work as desired. Most people think airplanes fly because they have wings, when in fact they fly because they have controls.

I wished the article went more into why decimal pales in comparison to binary. Is it because hardware components are easier to build and more reliable with two states instead of 10?
The question of whether hardware components are easier to build and more reliable with two states instead of ten is more subtle than it appears. In the early days the appropriate princples for building hardware were in flux requiring co-desigh of software (number base) and hardward. I'm particularly remembering the decatron: http://en.wikipedia.org/wiki/Dekatron. The comparison is between doing binary with triodes and doing decimal with triodes, dekatrons and anything else you might invent. Is binary really better or have you merely failed to invent a suitable tube?
Shannon's "A Symbolic Analysis of Relay and Switching Circuits" is hardly a 'paper'. It's a thesis towards Shannon's MSc.
And the thesis can be found here: http://hdl.handle.net/1721.1/11173
Thanks for letting me know more bits about Prof. Shannon's works, they are just ... beautiful