Fog Creek Software
Discussion Board

Master's thesis adds value to work experience?

I am starting my masters thesis in cs part time while working full time as a developer.I have completed all required courses and have my thesis left.I have 3 years of work experience mostly in Java and C++, in a wide variety of domains.
Currently I am doing J2EE/Web stuff. Previously I have worked in computer graphics and data mining in 2 other companies.
I want to do something practical in my thesis, which would add value on top of my work experience.
My choices are -

1. theoretical work involving new algorithms/graph theory
2. Work on wireless network simulators
3. Visual tools for software engineering

I am pretty much equally inclined on each of these topics, but which would add more value in my work, rather then in an academic sense?

Sunday, May 30, 2004

1 is fuzzy - you'll get some interest, but it sounds like "typical MS stuff" - in other words, not really much in the way of practical use. UNLESS you plan to work in areas where it has distinct practical applications (engineering, financials, etc)

2 sounds interesting, and may be a hot thing in the next decade as more and more wireless comes to market - I'd recommend studies dealing with scalability and signal density.

3 has potential - if you do a credible analysis of what software engineers really need/want in software UI's (see the emacs thread - it's a nice objective analysis), then you could end up working for MS, IBM, Sun, etc as a human factors engineer. But it could also end up to be a lot of wasted time. :-)

My $.02, anyway.


Sunday, May 30, 2004

You mean you can only choose from three different topics? Weird.

However, don't take the theoretical one, unless you're really really interested - it will be hard to come up with something original otherwise. And originality is the key when it comes to a master's thesis, as is my experience. Do something new and you'll be halfway already.

Good luck.

Monday, May 31, 2004

Do whatever amuses you as a hobby. None have significant income potential.

Monday, May 31, 2004

Were I you, I would work on #3.  Check out Visio, Understand for C++, and UML Studio.

While the perfect visual tool is currently a holy-grail of the software industry, I believe the Rational Rose / RUP / UML school of thought is off on a tangent. 

This is still a fertile field of research, I believe.  We very much still need inexpensive, effective visual tools for organizing software.  This has the largest potential bang for the buck -- if only we could get it right.

Current approaches are too large, expensive ($5,000 per seat) and unwieldy.

Monday, May 31, 2004

>>I believe the Rational Rose / RUP / UML school of thought is off on a tangent

The instructor for my Software Engineering class works for Rational (now part of IBM). I learned that they don't eat their own dog food (at least not his group).

One other interesting thing. His #1 SW Engr advice was to use state diagrams.  He writes one for every piece of code that he creates.  I thought that was overkill, but I did it for a small app I was working on at the time. I was surprised to find that it really made coding parts of the app much simpler.  Such a simple tool - so easily overlooked.

yet another anon
Tuesday, June 01, 2004

I'm with anon - I've had my MS in CS for about 15 years now and nobody has ever asked me what course work I've taken.  It doesn't matter.  All that matters is that you have the degree from a university that they've heard of.

BTW, I did mine in image processing and AI.  What a waste.  I should have been learning TCP/IP networking, OOP or Linux.  My thesis was trying to match faces found in photos.  Not a lot of that going on in industry, AFAIK.  I did write all the image processing software by hand which took weeks; I could have do most of it in Paint Shop.

Tuesday, June 01, 2004

I would suggest something with model-driven-archecture.  Look up the work of Harry Robinson.

If your school has a project or capstone option, you may want to consider those.

Matt H.
Tuesday, June 01, 2004

*  Recent Topics

*  Fog Creek Home