I’ve never written any proper postmortem posts during my LD carreer, but my LD#29 game, The Last Beacon was so successfull in my aspect that I have to do that now.

To understand, why I think so, I should start explaining, how the event have happened for me.



The titlepic, which is currently my best artwork ever. I suck at drawing, ‘uknow.


You must know that the days preceding LD#29 were completely empty for me. I just sat at home, played games, and sometimes visited social media, but that’s all – and I did it literally all day! If you wouldn’t know, this way of life drains creativity. Okay, of course I went to shop to buy food and fuel for the great event, but pretty much that’s all.

The fuel for my creativity was the fact that I decided sleeping some hours in the afternoon before the theme announcement. I woke up at 10 P.M. (CEST), did some Redditing, then I went back relaxing for an another, so I was kind of relaxed when I was waiting for the theme.


So, theme is “Beneath the Surface”. I didn’t know what to expect so my first (kind of reflex-like) reaction was a fist into the air with an “OH YES!!” shoutout. Then I started thinking of ideas, “just for some time before I go to sleep”. heh, well, I didn’t go sleep…

My first thought was “beneath the surface of an alien planet”, and I started thinking on that way. Only about half an hour was necessary to find out every features, excluding a proper title, a story and a protagonist. All of these were decided was later.

And the key to my success just have come…


Alpha Shot 2

This is it: the annoyingly precise draft of ALL tasks to do!

Around 4:00 A.M., I had an idea to build up a literally FULL to-do list to surely cover every aspects, and avoiding mistakes I did last time, when I screwed up 20 hours a completely useless map loading system – not even talking a lot of other, terribly implemented features.

It looked like this:

  1. I did some basic definition (where does it take place of, what’s the genre, will there be a controllable character, etc.)
  2. then I explained it (what’s the aim, how the player can reach it, will there be premade levels or procedural generation, what may happen on levels, etc.)
  3. then I decided about main features (behaviour of environment, basically) and I thought over, what kind of sprites should I draw (including technical sprites such as particles – yet they weren’t implemented at last)
  4. these technical features were great start for a to-do list
  5. and the final step, which is the most important: I made a detailed explanation for every points of the to-do list! I could follow these statement while I did the stuff that must be done and it not only speeded things up, but also made them “bug-proof”! It was especially useful when I created the map loading and saving system.

So yeah, this was the true power of my LD. And the energy drink.


“I hid behind this bottle perfectly…I hope no one takes notice!”

The structure plan was done around 5:00 A.M., so the actual work was started around that time. Then I worked on the game without stopping for 24 hours, and I literally didn’t stop!

At the end of this period, I had a fully working engine, and a partially working editor. The rest of the second day was spent for smooth features, such as the ridiculously simple intro sequence, the HUD, the rest of the textures, an autodetailer – and of course the creation of maps and story.

End was a little bit in hurry, and had to rebuild both the .exe and the .love file twice after Comp ended (first time for a bug that makes lvl2 unbeatable, and second time for the buggy appearance of the ending text), but fortunately, I could finish the game.

It was at 32+ hours of pure working, if not more. I can not estimate it precisely, due to time spent on YT, Facebook, Reddit, Twitter, and so on, but I’m pretty sure it was at least 30 hours of pure work. This is my personal record, and I’m not sure I’m gonna be able to exceed it.


  • Excellent structure plan. Explained above; it was the best thing I did during the whole Compo. Incredibly, how does it increase efficiency.
  • The peak of my LÖVE knowledge. Eight months of programming in Lua with this amazing set has proved that I’m actually able to make a great game.
  • Arcade feeling. The placement of HUD was a great choice, and that small and direct interaction with the skytexture also works well.
  • Quick basic implementations. After 3 hours of work, there was a working player affected by gravity. After 4 hours of work, both the grapplink hook and the laser beam were added. In those circumstances, they were absolutely free from bugs.
  • Reusable functions. “The Last Beacon” seems super easy either to port e.g. to C++ /w Allegro, or extend into a more complex game.
  • Design aspects of the levels. The tileset I did was proven to be great.
  • My first idea worked perfectly. Usually they aren’t as good, but now, it was great.
  • And I could do the whole planning phase without sleeping.


  • Ratio issues. After Hackfield, I wanted to use 16×16 tiles instead of the usual 32×32 system, but my 8×8 tile pixelstyle seemed way too small; I couldn’t even notice the player. Though the playfield would have been too small with 32×32 tiles, so I decided to stay in the middle: 24×24. Which looks perfectly well, but screwed up all the HUD and placement ratios I did, since 32 is a power of 2, while 24 isn’t, so I couldn’t stay at the 512×512 size of playfield. It was 504×504 (a.k.a 21tilesx21tiles at the end, and it made some crazy situations at smooths placements. Also, after all, the sie of window is 792×592 instead of proper 800×600…it’s not a big problem though, I think.
  • Lack of smooth movement. I wanted to implement it, but when I realized that it’s impossible (because player is moved by tiles, not by coordinates), it was so deeply rooted into the code, that I simply wasn’t able to recover without serious relapses. The disadvantage of quick implementation.
  • The other features I skipped. Force field, that block laser, but lets player go on; portals’ ability to teleport player & grappling hook; keys for doors; explosions…I could not implement them. I had to simplify as I could, and these more complex ideas were not good ideas to add, especially after 10 hours of work. I especially agonized a lot with explosions, because I simply wasn’t able to add explosion effects.
  • The rest of the bugs. I hate that I had to rebuild project files twice, because of last-time-recognised bugs, and I also hate that laser can’t burst through breakable blocks properly.
  • The lack of music and sound effects. I didn’t have time for doing them. Also, Abudant Music didn’t do music that was fitting to the style of the game.
  • Not challanging levels. Even though levels look beautiful in my view, I suck at making puzzles, and I even didn’t have much time, so in the aspect of gameplay, “The Last Beacon” is possibly a boring game.
  • Not enough levels. Number of levels is also too small – I planned 21 at the beginning, but I was also bored and running out of time, so 3 hours before Compo ending, I had 15 levels. I made the right decision though.
  • Hurry at the end. It made a lot of unpleasant “featues”, including a ridiculously small story and a kind of short ending.


Even with every unpleasant events and bugs, LD#29 an outstanding success for me, and this thirs time was the most entertaining for me among all of them. I still can not believe I actually created this…

I hope I’m gonna have a good ranking at the end. Also, I’d like to extend this game in the future, like Alan Zucconi did that with 0RBITALIS.

I hope I can do that. But until that, good luck for every competitors, and thanks for reading!

– Katamori

Tags: ,

One Response to ““The Last Beacon” – Postmortem, a.k.a the success of planning”

  1. SenorDiablo says:

    Energy Drink! My best friend!!! <3

Leave a Reply

You must be logged in to post a comment.

[cache: storing page]