this is the third Factorio weekly update. It is scary how fast the week went by. The summer is definitely gone and it has been raining hard for the past two days here in Prague. We spent the weekend by playtesting the 0.7.1. The bugfixes were mostly finished and we were after the balancing. This resulted in two freeplay games in which we both (me and kovarex) managed to build and defend the Rocket Defense in a little bit more than 11 hours. We both used different strategies. Kovarex went after the logistic robots and the beacons, while I tried to keep my factory relatively small and was really focused on researching the rocket defense asap. With the balancing changes we made, the game seems to be well playable again. No more crazy medium biter attacks after 10 minutes. In my game I saw a first medium biter after cca 4 hours and the first big biter after about 9 hours of playing. That seems allright for the regular settings.
After the weekend we focused on getting the 0.7.1 out. In the end we managed to do that on Tuesday night, after a full day of work on fixing small issues coming mostly from the Lua API refactoring. The 0.7.1 was received rather well. It still has some bugs though. Especially regarding the enemy expansion. There was this funny save where you stand in the middle of your factory and in like 20 seconds out of the blue sky there appears an enemy spawner and two worms right next to you:) So yeah, there will be 0.7.2. We will put in only bug fixes and will try hard to make this one the stable release for 0.7.
After the hectic release we spent the next day trying to relax a bit and put plans together for the next iteration. More or less the plan is now clear. There are three priorities:
Albert has been onto the terrain for couple of days now. We were playing with an idea to use Wang tiles but in the end decided to go with different size variations of tiles for the same terrain. This should break the grid-like feel of the current terrain. Albert developed his own, simple yet clever technique for creating tile variations rather fast. Basically he creates the tileable edge of a single tile and then just changes the inside for every variation. Keeping everything as 3D models allows him to do all sorts of tricks (like changing the height gradient really fast). Some preliminary results look promising. Here is an example of dry dirt terrain with random machines on top of it:
As before the link to the post is in the separate topic on the forum. So you can post your comments there