C Hees . info

Finally a result!

…
Generation 42: 1.0
DEBUG: inserting step in the trace: CALL_METHOD, test () ON REF 0
DEBUG: inserting step in the trace: CALL_METHOD, plus3 () ON REF 0
DEBUG: inserting step in the trace: CALL_METHOD, plus3 () ON REF 0
DEBUG: inserting step in the trace: CALL_METHOD, plus3 () ON REF 0
Generation 43: 1.0
DEBUG: inserting step in the trace: CALL_METHOD, minus1 () ON REF 0
DEBUG: inserting step in the trace: CALL_METHOD, plus2 () ON REF 0
DEBUG: removing step from the trace: CALL_METHOD, plus2 () ON REF 0
test 33 -9 24
Generation 44: 0.0
Covering trace found!
The fitness of that trace is: 0.0
The length of that trace is: 54

Method: Examples.GATest.test_0
Target Path Count        : 2
Covered Path Count before: 1
Covered Path Count after : 2

Just so you know… this is a big breakthrough for me ;)
It’s really cool cause I can improve the coverage of T2 using my Genetic Algorithm :)
Now to make it work in more cases =)

Sogyo, the company where I’m doing my thesis project, had a video made:

At 2:13 you can almost see where I was working on genetic algorithms. But hmm, probably no more genetic algorithms for me since it turns out they might not be a very good solution to the problem. So byebye GA, hello dynamic symbolic execution? or thin path searching? or something else?

I’ve been searching for companies and ideas for my thesis and I ran into cloud computing again. It caught my attention before so it would be cool if I could do something with it for my graduation project. While I was reading up on the subject I couldn’t resist to check out the new browser by Google which is supposed to be good for applications running completely online (as in.. it’s sortof related to cloud computing :P).

I wanted to know if it was just a Mozilla ripoff or that it is actually something new. If I’m to believe the Google Chrome comic/book they really used a lot of new interesting technologies. One of them being another rendering engine (WebKit which is also used in Safari and which came from KHTML which is used in Konqueror which I really loved). I wonder if websites in Chrome look the same as in Safari. Either way, Chrome seems to be really fast. It starts up fast, shows pages quickly and executes Javascript quickly (see some benchmarks at scriptNode). What I also love is that one slow site doesn’t hang your whole browser. For example if you run this raytracer in one tab, your other tabs will get very slow in Firefox while in Chrome you don’t even notice the difference. I think this should also work for slow or hanging plugins like Flash which would be great.

So far Chrome seems to be nice, except for one bug I noticed:

When you hover over the title bar while it’s not maximized, the tooltips for the close, maximize and minimize buttons show up in the wrong places. It’s of course a very minor bug but it reminds me that it’s just a first release and probably still contains many other bugs. I guess I’ll see in the coming days if I actually like Chrome or not.

Today I went to the university library cause for some reason I thought that I could get some information there. What was I thinking?

This enormous building with 90 km of books has a computer science section. The computer science section is highlighted in red:

Yes, that’s a whopping 2 bookcases. And most of the books are about things like Algol60 (a programming language dating from 1960), Fortran and cutting edge computer graphics… from 1991. Completely useless. There were hardly any books on recent topics, let alone books that were really up to date. I guess it’s hard for a library to keep up with new techniques, but this is ridiculous. At least I know I won’t ever have to go back there again.