Slashdot Mirror

http://ptth.net/slashdot/gimpdemo.jpg

Evidence for my assertion in my other post.

Here you go:

http://ptth.net/slashdot/2d_rubik's_cube.JPG

Couldn't find squish, my little web game, even given the hint that it was on ptth. Google finds 4 pages of relevant hits.
http://ptth.net/squish/

http://accoona.com/search.jsp?qt=squish+ptth&col=w c&charset=utf-8&la=en

http://www.google.com/search?q=squish+ptth&start=0 &ie=utf-8&oe=utf-8&client=firefox-a&rls=org.mozill a:en-US:official

http://ptth.net/squish/ WTF is that? Sorry but I'm not going to run any unsigned java code. Can anyone vounch for this?

I made a couple of small freeware games, followed by a couple of successful quake mods. The quake mods were successful enough to have been played by the game developers who later interviewed me for a position at Blizzard North, which I got.

Then later after finding out how painful it is to work in the gaming biz, I got out. But if you want in, make some games.

Now I just make fun little games in my spare time, like squish:

http://ptth.net/squish/

If you check it out, journal me how to improve it please!

Since I often wonder just how much performance i'm giving up using java instead of c, I try an experiment like this about once per java version.

The current result
c (ms visual c, optimize speed): 4391 ms
java (5.0, -server): 10031 ms

So there you have it, java is only a little over twice as slow as c, even at something that ought to be c's greatest advantage.

If anyone knows how to optimize the java outcome better without obfuscating the code or changing the contract of the functions, comments would be welcome.

I also compared replacing the strlen calls with a hard-coded length of string value (tail = 19;), to make sure that my strlen implementation didn't suck, but the outcome was pretty much the same, c twice as fast as java. (c: 2359 java: 5171)

Here's the code (link to avoid violating compression filter):

C:
http://ptth.net/slashdot/stringreverse_c.txt

Java:
http://ptth.net/slashdot/stringreverse_java.txt

Since I often wonder just how much performance i'm giving up using java instead of c, I try an experiment like this about once per java version.

The current result
c (ms visual c, optimize speed): 4391 ms
java (5.0, -server): 10031 ms

So there you have it, java is only a little over twice as slow as c, even at something that ought to be c's greatest advantage.

If anyone knows how to optimize the java outcome better without obfuscating the code or changing the contract of the functions, comments would be welcome.

I also compared replacing the strlen calls with a hard-coded length of string value (tail = 19;), to make sure that my strlen implementation didn't suck, but the outcome was pretty much the same, c twice as fast as java. (c: 2359 java: 5171)

Here's the code (link to avoid violating compression filter):

C:
http://ptth.net/slashdot/stringreverse_c.txt

Java:
http://ptth.net/slashdot/stringreverse_java.txt