Hi everyone !
The 30th edition of LudumDare will start soon, it’s a good moment to think about what happend in the previous edition. Here is the link of my compo : http://www.ludumdare.com/compo/ludum-dare-29/?action=preview&uid=484
LD29
.
I consider that it was a success. My goal was to create a game with a start and an end and Unminor meets the exigence. For my previous participations, I used to program in Python with Pygame. It was painfull to develop with it, I really love Python and use it almost everyday at work but there is so many missing common features you have to code yourself with Pygame … I think it’s too low level for Ludums. Oh, and the deployment part is horrible. I still don’t know how to generate a windows executable that just work.
For LD29, I tried Haxe with HaxeFlixel framework. It was a pure pleasure to use. There is all builtin functions needed to quickly develop a game (loading a map, handle velocity for objects, …). My favorite part was the sunday night : in five minutes I had working builds for windows, linux and flash. I will definitively use it again for LD30.
The main issue I remember was to load a map from tiled. I misused the functions to load a map so it took me a too long time to have a functionnal stuff.My map class is more than 100 lines long, and the biggest majority of those will probably be the same in every projects.
So this week end I will write a Map class to speed up the development time for handeling a map from Tiled. It was very helpful to have a powerful external map editor.
In the feedbacks, there was multiple complains about the lack of wasd controls. In fact, me neither, I don’t like directional keys. But I’m french and here, we use zqsd. Usually, games allow users to remap their keys, but who want to map their keys for a 5 minutes flash game ? Some games ask the user if they use zqsd or wasd controls but I did’nt had time for that. After thinking about it, there is no conflict having wasd and zqsd working in the same time (and directional keys for people who wants). Also, the support of a gamepad like a 360 controller could be a cool feature. So for the LD30, I will map wasd, zqsd and directional keys. This week end, I will develop an input handler to speed up that part of the development. It will be great if I can support console controllers and virtual gamepad (for mobile) with that.
By looking in my source code, I see that I should create a generic Menu class to spend less time on it for LD30. I remember I spend maybe one hour on it, and it’s just a waste of time to display a title, create an animation and 3 buttons. So, one thing more to do this week end.
There is two sounds in the game : beep and bip. I used bfrx. This is an amazing tool, but I don’t know how to use it except by randomly change all parameters until I hear something correct. I should practice … For LD30 I don’t know if I will use Bfxr or Jfxr, those tools seems similar, I don’t know if there are big differences.
My game don’t have any music. It’s a bad thing. I tested many games of ludumdare and … A game without any sound is just awfull. Any game that plays 3 notes in a loop enhance greatly the game experience. So if I can I will use Lmms for composing a little music. Unfortunately, I have no music knowledge. I think that one day, I will buy Rocksmith to learn to play guitar so I will be able to learn music composition. Until that moment, I will try to do a partition with good looking patterns and pray that it generate a good melody.
For graphics, I used Gimp. I was happy with what I’ve done even if it was ugly. I will use Krita for LD30 and I should customize the work panel to fits my needs.
In term of organization, first I managed to find a project I was sure I was able to achieve in a week end. It’s what was wrong for me during the last editions. Secondly, before the tasks I started, I choose how much time I will spend. If this time is over, I stop working on it and try to do something functional with what have been done even if I didn’t finish. This is because during the LD28, I spent one complete afternoon trying to do with rotations of the canons on my boats (yes, I sucks in trigonometry). This was a big waste of time, I should have found a workaround even if it lower the quality of the game (a bad game is better than no game for a compo, no ?). The big problem in this situation is that you think something like “oh, no, I spend two hours on it, it would be stupid to abandon after so many time, maybe I will found the solution quickly now !”. So now, I try to avoid that with my timeout for every task I do. Oh, and another good thing about this method : when you work on polishing, it’s hard to know if you have to stop or continue and work on another thing. With the timeout method, no need to ask, you try to be better until the clock says to stop (except of course if you are proud of what you did and don’t see how to do better).
During LD29, I was working at school with some friends, maybe we spent too many times for our meals, but if it can maintain the moral up, it’s a good thing.
Maybe my postmortem was too late, but as we say here, it’s better later than never.
Oh, and as I said before, I’m french so excuse my language ( http://danstonchat.com/1337.html ).
Tags: haXe, postmortem, preparation
