email sol follow sol rss feed of the blog wishlist Sol::Games

Tekram-S

Ludum Dare's 4th 48h game contest

Downloads

Tekram-S

Win32 version (also contains sources and timelog)

Postmortem

This was the second time I entered an LD48h contest. The only preparation I did before the beginning was to try to get my microphone working.

While I was slightly disapointed at the selected theme, I decided to write a simple puzzleish game for the LD48 contest. Unfortunately I fell ill right at the beginning of the contest, but was still able to, more or less, finish my entry.

The name for the entry was primarily chosen to be different. I was rather confident that nobody else picks the same name. (And no, I have nothing against certain finnish grocery chains).

What Went Right

  1. Scope

    The scope of my project was small enough so that I could finish it easily in time. If the theme had been something else, the scope might have been wider, and given my health at the time of the contest, I might not have been able to finish it.

    If you consider the game's scope, it requires very little content, very little scripting, very little tuning and it has no overly complex subsystems (such as physics or 3d collisions).

  2. Scheduling

    The schedule was forced by my health, but I had meant to do it that way in any case (albeit not exactly so aggressively).

    The contest started at 4am my time, but I woke up at 8am (thanks to our dog, actually - I most likely would have slept till 10am without her). I meant to take a shower, eat away from my computer, take the dog out, and have a good night sleep. Due to the illness, I had to call it a day around 7pm.

    It was a good thing is that I got my first full-featured version done at that time, so even if my illness would have gotten worse, I would still have been able to compete. (At 7pm the contest had been running for 15 hours). The next day I still didn't feel too good but was able to tune the difficulty a bit and create the tutorial.

  3. No killer bugs

    Thanks to the simple scope, clear vision and strict programming dicipline, I only had one bug during the whole contest that I had to hunt, and that was found easily.

    I created a simple 'screen' architecture which let me add menus, intro and endgame screens easily, and starting a new game, something that I had problems with in my earlier entry, came for free with this framework. Funnily enough, my last entry post mortem praised that I did NOT do much OOP, while this time doing classes for most important things was a good thing.

  4. SDL

    Like with my last LD48 entry, SDL was all I needed. This time I didn't even plot a single pixel - filled rectangles and image blits were enough. Mouse and keyboard input was a breeze.

  5. Difficulty

    Since I finished the feature-complete version early enough, I had time to test the difficulty of the game. Since I managed to play up to the "sick" difficulty level, and other people managed to solve the "childish" level on their first try, I'd say that the game has a good range of difficulty levels. I also made sure that even the "insane" difficulty level is solvable by limiting the mutation strength, and adding keyboard shortcuts for the tool keys.

What Went Wrong

  1. Scope

    The small scope of the game was also a bad thing. If I had selected a more complicated game, it might have been a better game. However, I most likely would not have had the time to finish it in that case, again due to..

  2. Health

    What a rotten time to get ill! At first I thought it was a food poisoning, as my girlfriend had the same symptoms.

    Unfortunately the fever rose during the day, and when I finally took my temperature it was at 37.5'c (99.5'f). Sort of ironic, considering the theme.

  3. Graphics

    Yes, yes, I know, the comic sans font is horrible. Had I been healthy, I would have redrawn all the graphics and made it look even somewhat like a serious game. I had the idea of rendering the "tekramytic-k" pill box etc, but simply was too ill to do it.

    I also made a bad design choise in tinting the "target disease" towards the red; it felt like a good idea ("evil disease") but this was confusing for some people at first, as you could not *exactly* match the pattern.

  4. Audio

    My microphone died. End of story.

  5. Difficulty

    While the range of difficulties is good, their variety isn't. The only thing that changes from one difficulty level to the next is the size of the grid. A better idea, in retrospec, would have been to stop the grid growth at about 4x4 and make the game more difficult in other ways, such as more aggressive mutations and more than 2 color choises for the tiles; both would have been easy to implement.

Project Stats

1 PC (p4/2.6GHz)

1 bag of chips (350g)

5 bottles of coke (7.5l)

9 classes

11 images

17 source files

59kB of source code

1800 lines

total of about 16 hours spent working on the entry (including shower, eating, taking the dog out, but not sleeping)

Comments are appreciated.

Site design & Copyright © 2017 Jari Komppa
Possibly modified around: May 01 2010