Thursday, June 16, 2005

Hacker article on TLC

Nice article

The Learning Channel has a good article (or collection of articles) on hackers and it's history. I'm adding it to my list of links.

-Just something to read in an afternoon

Hey, look. Blog offers titles now.

So I guess I don't need this title any more?

Windows and its vulnerabilities. You don't really have to wonder why people still use windows, even after hearing of all its security exploits. Myself, I'm moving away from windows. I've got it on my PC at home, and on my laptop as well. However, I'm not going too far from windows. My problem is, windows is a gaming platform. Not too many games are supported on Linux at the moment. How am I going to get my Half-Life / Half-Life 2 fix on a *nix operating system?

I'm removing windows from my home PC, Kubuntu is taking its place. My laptop, however, is still carrying Windows, but only for the games. I've got Cygwin on there and it's configured very well.

-I've also heard good things of minGW

Tuesday, June 14, 2005

Research follow-up

My professor, I think, has finally realised how little enthusiasm I have for the mathematical sciences. Or, forbid! she has access to my blog.

She's proposed that I use my program to help out the two entomologists we've been working with. I have had questions all along in regards to how I should handle certain events in my simulation. Gary, the field guy entomologist dude, has suggested I try my question in my program, varying it so I can get an answer myself. The answer will be when we compare my data and his and draw a logical conclusion from that.

Sounds good to me, and I get to avoid the mathematical analysis. I do get to write an advertisement for my simulation. And I've gotten permission to post is as open-source.

-Building a better resume
Summer Loving

I don't know what to make of my summer research assistant position. My professor, she's a nice lady, but I don't think she knows what she wants with my research. I'll give her credit, she's juggling three students at the same time, each with a project unrelated to the other. Justin with his simulation of a swimming nematode. Brock with his recolonization by squirells in clear-cut forests. And myself, with my dispersal of codling moths in the presence of a network of pheromone traps. I found that the longer the title, the more impressed people are with your work. But that's getting off topic. Rebecca, my professor/supervisor, has her own research going on as well. And the school is forcing her to take her vacation time at the moment, which she is taking up until the end of this month.

However, her research runs parallel to mine. She's covering codling moth dispersal using a mathematical model. I'm doing a cellular based model on the computer. This is where it gets hair-brained. We started on my project last summer to coincide with hers. Toward the end of last summer our work was beginning to overlap. At that point, all I had was a simulated orchard with a central release point. I'd release a mass amount of moths in 4 trials of 2000 moths each. The moths would fly around their orchard and either die, become trapped or fly out of bounds; all of which would remove them from the program. OK, that's cool, but then she wanted more features. And then more features. And then she wanted me to throw together a presentation. Then she wanted me to run the experiments to gather data, from a program that I felt still needed to be polished.

The data was run, imagine 1000 trials of 500 moths each in a small simulated orchard. Now a program needed to be written to injest the 1000 large files containing data from the simulation. Enter MatLab, another language to learn. Now for the whole school year all I did was type what Rebecca dictated. I felt I could better be used in polishing the program, fixing my bloated code. I'd only had two semesters of programming experience, and that only in Java, before she had me writing this simulation for her. So 8 months of meeting with my professor for 3~6 hours a week, writing a program she was more than capable of writing herself. She knows Matlab inside and out. She could have written up the analysis program in a week, where it took the two of us almost 8 months! And I couldn't have done it that quickly on my own, I had my own classes to tend to.

So now that it's summer time again; she informed me that I was to have a paper written by the end of the summer with the intentions of having it published early next year. So the whole first month of my summer position I was modifying my program, fixing the code, tweaking it, generally making it speedier. When the program takes 30 minutes per trial of 500 moths (doing a trap efficiency experiment) you don't think that's too bad. 30 minutes 1000 times, well, that's getting a bit much. But when the program takes 8 hours for one simulation of 2000 moths, and you have to run 4 of those experiments to have enough data to say you've emulated what scientists do in the field... yeah, that's a bit extreme. But fixing my program, with the knowledge I gleaned this year in classes, I have the program going as quick as 2 hours per quarter. It takes only 8 hours to run the whole experiment!

As I said in the previous paragraph, I'd spent the first month making my program more efficient. But she wanted something I could show for. So I added sterile female moths to my program, to accompany the sterile males I'd had for the year. Though, we're not even close to ready to use these female moths in our simulation. I feel we still need to polish up the male aspect of the simulation. Get the no-wind model analyzed again now that I've made some modifications. Then we can move on to an all-male analysis using wind, since I just recently (this last week) fixed a minor bug that prevented my moths from moving in certain directions (they should be able to move in a full 360 degrees along the x-y plane). Then after we've got that data, and have analyzed as best we can, maybe we could then how the data changes when females are introduced.

But wait! Sterile males and sterile females aren't enough. Now we need to add wild males and wild females. Because, you know, we can. And when that's done, we need to have these moths mating with each other... you can see where this is going. This simulation is getting bigger and more complicated. I've got my boss looking over my shoulder (once a week, Wednesdays) telling me, "Looks good, now do this". And all the time, no research is being made. I'm just programming and programming. But don't get me wrong. That's what I'm here for, I'm here to program this awesome simulation that takes in the important parameters so the Scientist can use my program, tweak the numbers, analyse the data and write the paper.

However, my role in this project has grown. I went from just having programming duties to having to analyse the data. Then creating presentations. And with the presentations came the mathematical formulas I had to make sense of (I'm not a math student by far). I just want to go back and polish my program. I want to make it user-friendly. I want to write the document to allow the world to use my simulation (open-source if I had my way) to study the movements of more than just these Codling moths. The program could be modified to work with, say, salmon swimming up stream. Maybe even cows wandering across a field, eating grass and rolling in mud along the way. I'm here to program, that's how she sold this job to me.

And when you consider how much this program is worth, think of this. I've been employed for the last year and more to work on this. 4 months of 35 hours weeks, that's 560 hours right there. And 8 months during the school year of, say, 5 hours a week. That's another 160 hours. Plus the last two months I've been doing my summer position, another 280 (I hope I'm getting my math right). So that's a total of, wow I didn't this number, 1000 hours. And let's say I'm being paid an average $9.50 canadian for this (none of which I have managed to save) and that's $9500.00 that has been put into this simulation. So, thus far, this simulation is worth $9500, and if it were commercial, you'd want a product that you can sell to make a profit. If you wanted a profit, you'd be polishing the program and making it the best thing you could. But sadly, as you've read this far, that is not how this is going.

So now I'm going to read the stack of scientific papers I have sitting in front of me. Hopefully this will give me some insight as to how and what I should write about. That's the topic of meeting with my professor today, we're going to decide what I am to write about. All I can think of is writing an advertisement for my simulation. How it may benefit the scientific community. How to use it.

Providing I'm not contractually obligated to keep it under wraps, this is going

-to be open-source

Monday, June 13, 2005

Guess who's back

Gorillaz!!! They're 'performing' a virtual concert today for 99x listeners (everything alternative, radio Atlanta, Georgia). Click and enjoy.

-I've got sunshine in a bag