Delaney, Timothy C (Timothy) wrote:
From: Dennis Allison [mailto:allison@sumeru.stanford.EDU]
How about re-engineering the interpreter to make it more MP friendly? (This is probably a bigger task than a Masters thesis.) The current interpreter serializes on the global interpreter lock (GIL) and blocks everything.
To me this would probably be the most interesting thing to tackle - especially since it has been tried before with partial success but overall failure. At the very least that gives a body of work which you can refer to both as a starting point for your work, and to show how your approach differs from and improves on existing work.
It would also be of tremendous value to Python IMO if it could be done without negatively impacting performance on single-processor machines.
Whether it is too large for a Masters thesis I don't know. Does a Masters thesis require *success* in the stated goal? I've been thinking about doing my own Masters in the not-too-distant future if I can find the time ...
Success as in what you set out to do was actually beneficial? No, just as long as something is learned. Successful as actually finishing the darn thing? Yes. Basically a masters thesis needs to require some research, such as looking at other implementations, and some original thought if possible. The problem with a masters thesis, though, is that I have a fixed timeframe (want this done in about a year's time for doctoral school applications) and I don't get to spend a large portion of my time on it (I still have to take normal classes during this time, although I can fenagle my schedule to minimize my work load). I will still consider this, though. -Brett