Dear friends,

A fresh magazine is ready and can be found here:

Aum 112

Have a great day.
George

Thanks for this new aun ^^

Thanks a lot for the new AUM, the kind words and the occasion to talk about my game!

EDIT: Just two small things about the new AI code:
- Right now the enemy oftentimes shoots against the wall in front of him because the c_trace only uses enemy's origin. USE_BOX could be a solution but IMO oftentimes leads to mediocre results and other problems when you try to realize visibility checks. I think two c_traces, one from the head and one from the hand (in shoot animation position, at least approximated) could work out just fine and make the enemy much more dangerous.
- In the move_bullets() function you set bullet_speed.x = 50*time_step; prior to the while loop where you move the entity. This can work but only will do so consistently when you have a steady framerate. Otherwise the outcome can and will vary (consider for instance a moment of action or when the player looks a certain way and the framerate drops to 30, a bullet created then will move at double the speed when the framerate goes back to 60).

EDIT2: Nice to see my multiplayer basis being helpful! However, there are two small things about the current pan implementation, that is it needs to be sent in the my.pos_update case as well (and my.pan +=... is missing a *time_step). Currently, the code does not really work when one client rotates his character before another client joins the server (or ent.pan gets a little desynchronized during simulation).

what a surprise, thanks!

Cool, nice to read about Superkus game and to see another FFA-Contest. Having some cool ideas already...

Thank you for the kind words, guys!

@Superku: thank you for the suggestions. I am aware of the enemy origin issue; I talk about it in the Aum article almost as if it's a feature but I will try your suggestion and see how it works.

From what I remember, the bullet_speed.x value should be pretty consistent because the frame rate is limited to 60. It's true that I didn't test the code on a very slow computer; that could cause problems.

Concerning the multiplayer code: first of all, thank you again for providing such a clean, well coded piece. I didn't try to rotate a client before starting another (I have always started two clients one after the other) but I can see why that would not work properly. I wanted to send the pan angle at the beginning, but then I thought that this value isn't critical and the code was working fine with my setup. I'll take care of it.

awesome, thanks george!

Thanks a lot for the new AUM, nice nice !!!

Thank you, guys!