?

Log in

No account? Create an account

Previous Entry | Next Entry

Sep. 10th, 2007

Today is roughly the 25th anniversary of the start of my fourth year at MIT. This has special significance because I changed majors back to CS from EE, after realizing it wasn't what I wanted to do. I was nervous about the decision, for reasons I've discussed earlier. I'd registered for 6.033, the computer systems class (the one with lots of papers and written assignments), and also signed up for a UROP in the lab the 6.033 professors belonged to.

It is fitting that I should remember 6.033, which explored (somewhat) the "essence" of software engineering, since I'm asking myself similar questions today. I wish that I had considered more of the implications of some of the readings, such as the "Tar Pit" chapter in Brooks' Mythical Man-Month. At that time, I was more concerned with trying to understand the concepts covered in the book, in order to get a good grade, rather than asking if the book might be trying to tell me something about what might lie ahead. (I'd had summer jobs, but hadn't worked on a major project with lots of unexpected "gotchas".)

I wish I'd known more about the history of computer science and software engineering from the standpoint of how notions of "what is important to learn and understand" came into being. Brooks' text covers this somewhat. There is much more on this topic at Michael Mahoney's History of Computing Articles page. I will probably be commenting on several articles in the near future.

The Finding a History for Software Engineering article addresses some of the questions I'd raised 25 years ago. At the time, I met some of the professors I've mentioned before (David Reed, Dave Clark) and also some people who would go on to become top network protocol designers (Noel Chiappa). I didn't know much about networks at the time, but one thing I noticed was that the 6.033 staff's focus was primarily operating systems based, while there was another, more mathematics-based focus in books such as Tanenbaum's Computer Networks, drawing from probability and information theory. So I wondered about the difference, and also what it was that made Reed, etc. "good". What sort of things did I need to do to become a principal contributor to computer networking? Did I need to focus more on operating systems or information theory? To what extent did one lead to understanding of the other? I was starting to come to understand the issue of not having enough time to cover everything, but not knowing which were the most important things to cover.

Comments

( 1 comment — Leave a comment )
nsingman
Sep. 11th, 2007 04:59 pm (UTC)
Thanks for pointing out Michael Mahoney's page. I just started skimming a couple of articles, and it's a rich source.
( 1 comment — Leave a comment )

Latest Month

July 2018
S M T W T F S
1234567
891011121314
15161718192021
22232425262728
293031    

Page Summary

Powered by LiveJournal.com
Designed by Tiffany Chow