Postmortem: Poor Ball

Posted by (twitter: @fifif0x)
August 23rd, 2011 12:22 pm

Poor Ball - Click the image to rate the game !

Poor Ball – Fififox – Competition Entry

Now that the stress of coding/packaging/testing/submitting is gone, the time is right to sum up what I’ve done, the good choices and the bad choices.

What I’ve done right (based on my previous and first LD experience, LD19):

  • I finished my game (I haven’t been overly ambitious);
  • I only made graphics and sounds after the game was working;
  • I made something (relatively) nice and with good sound effects;
  • I made a title screen and wrote a short story to introduce the game mechanics;
  • I spent time on IRC, but only time which I didn’t miss as the deadline approached;
  • I didn’t feel stressed until the very end !

What I’ve done wrong:

  • I thought that good ideas would come by browsing dictionnaries and wikipedia and probably lost time by not really concentrating;
  • I got used to my game’s difficulty and didn’t see that removing the two easiest levels was a bad idea.
  • I spent a long time trying to repair a library instead of just replacing it by another one directly;
  • I tried to cross-compile my game and the libraries I was using instead of booting that old slow Windows box.
  • I should have written more progress reports instead of being on IRC for time.

What I’ve done during those days:

First day:

  • 9 am (theme announcement at 4am for me): Woke up and looked for some ideas using theme list / dictionnary / wikipedia;
  • 10:00 — Created a base system in C++ with SFML2;
  • 10:20 — Thought, not looking at the computer;
  • 11:00 — Tried to write down an idea;
  • 11:40 — Threw that idea away and thought again (also played prototypes seen on IRC);
  • 13:00 — Back to code, looked in detail at the SFML API;
  • 14:00 — Suddenly had an idea: posted the idea on LD blog and started coding like mad;
  • 14:45 — Had the basic physics working, so screenshot and lunch on LD blog;
  • 15:00 — Kept working on physics;
  • 16:00 — Took a break;
  • 17:00 — Back to code, added controls;
  • 17:30 — Created a minimal user interface, worked on graphics;
  • 20:00 — Had the physics working, posted progress report and screenshot on LD blog which is getting slow, tweeted, talked on IRC;
  • 20:40 — Created some bad background music (removed from the final game, was very bad 😛 );
  • 22:00 — Tried to get music working, SFML uses OpenAL, which is apparently badly broken on my computer, tried to find a solution to that (but I didn’t);
  • 02:00 — Stopped searching, went to bed.

Second day:

  • 08:00 — Woke up, 20 hours remaining, eating some croissants (yeah, I’m french 😀 );
  • 08:30 — Took a crash course to use FMOD for sound in my game, and set up background music;
  • 09:00 — Heard about LD server failure;
  • 09:15 — Tweaked physics and changed gameplay idea, implemented some GUI;
  • 11:40 — Added some graphics;
  • 11:45 — Fixed a race condition-related segmentation fault, added levels, lives and death to the game;
  • 13:30 — Got tired of adding game code to the physics thread and edited some graphics;
  • 13:35 — Talked on IRC and looked at some other sound libraries;
  • 14:00 — Got back to gameplay, and eventually broke the whole physics/gameplay thing, causing dead locks every now and then, but somehow I got it working nearly seemlessly;
  • 15:40 — Talked on IRC, helped testing some games;
  • 16:20 — Wrote a progress report on the really slow but working LD blog, while still talking on IRC and failing at finding FMOD documentation on compressed sound;
  • 18:00 — Added code for sound effects and created some, which made me remove the background music;
  • 19:45 — Refactored a part of the physics code to remove an old feature which looked more like a bug from the outside;
  • 20:10 — Created other sound effects, reworked the sound load/play system;
  • 21:00 — Talked on IRC, posted an update on the LD blog, and tweeted;
  • 22:00 — Rewrote the GUI and added a graphical title screen
  • 00:00 — Sumbitted my entry with description, screenshots, sources and Linux x86_64 executable.
  • 01:00 — Wrote a post about my entry on the LD blog;
  • 01:10 — Talked on IRC and started packaging for Windows and Linux i686;
  • 03:00 — Read about FMOD’s API on Windows which doesn’t support C++ (only C) when using MinGW, converted the source to use FMOD through it’s C API, compiled, packaged and sent;
  • 03:30 — Reuploaded different versions with more and more libraries for people to play the game just by unpacking and running: thanks to people on IRC who tested my game until the very end, at 04:00 for me !
  • 05:30 — Encoded my timelapse and posted it on Youtube and on the LD blog.


Leave a Reply

You must be logged in to post a comment.

[cache: storing page]