Tim-Oliver bothered me today with a question about A* pathfinding. It has been some time since I wrote our A* implementation in DVW, so I had to dig into this topic again. I used this tutorial and wrote some code. Note that this is just a straightforward implementation of the pseudo code presented in this article. I wouldn’t write productive code that way. Anyway, this might come handy for someone, so I’ll present it here. Next time, I will replace the open list with a binary heap. Click here to see the source:

Many people are posting interesting queries for Google code search. Here’s another one which might make Java engineers shake their heads.

Tonight, I added the batch processing. On the left image, you can see a set of six input images on the left, and three output images on the right (all images at 50% zoom, another feature I added). The packing is not optimal in a mathematical sense, but it’s effective anyway, as the shot on the right shows. These (100% zoom) images were randomly picked from a set of output images which were created when I reorganized all of the plants we have in DVW to test my tool.

Tonight, I enhanced the tool (codenamed TexTorture, hehe) with a little rectangle packing algorithm. Taking an unsorted list of sprites (i.e. rectangles), the outcome is already quite lovely. Next, I will try how presorting of the rectangles by width, heigth or area will affect the result. Finally, I want to be able to merge n input textures into m output textures, where m<<n, hopefully.

I just wrote a quick and dirty Java/Swing tool. It loads and autodetects DVW sprites from a texture file and allows to reorganize them. This is useful for stupid programmers like me who don’t want to mess around with mspaint.exe or The GIMP but feel the urge to move sprites around for whatever reason. I am planning some improvements like a textual export of the rectangle coordinates and a cool algorithm to align the sprites automatically including batch processing, but for now, manual moving of the sprites is exactly what I wanted. The autodetection is done using a simple floodfill approach which proved to be sufficient for my test texture. I might find other textures where this approach isn’t enough, therefore I will provide means to apply a dilatation operation on an image and to merge detected rectangles. While I wrote this tool, I recognized a little problem in the AWT drag and drop framework, java.awt.dnd.DropTargetListener in particular. I used DropTargetListener#dragOver(DropTargetDragEvent) to listen for dragging but found that there was no (direct) way to find …

Many of us are enthusiastic about getting into the game industry. And the more enthusiastic we are, the more we might wonder why people are happy to leave the business again. At the end of 2004, a blog entry widely known by the authors nickname, ea-spouse, rocked the boat. I remember that many professionals joined in complaining about the working conditions. Recently, I read two similar articles again: Jurie Horneman of ex Rockstar Vienna wrote that he was glad when the studio was shut down. He refered to another article entitled The joyful life of a lapsed game developer whose author was glad to have left the industry for good. How can it be that so many want to get in, and at least some are happy to get out again? Let’s face it. The trouble is that the game industry still is an industry. Like in every other business, be it software or whatever, there are customers, deadlines, and consequently there are crunch times. And I’d expect my employees to commit themselves to the …