Currently Decay hits are quite random and seem to be able to hit one object multiple times in quick succession while leaving others unscathed for long periods of time. I'm going to speculate that what this looks like at the code level and hope loftjorb can confirm or correct. The map is divided into larger areas (maybe as large as a whole super-grid square) and each game 'tick' a random X/Y coordinate is chose in each area and a decay hit is computed there. This can be though of as 'dartboard' randomness. If this is indeed the way its done then I propose changing to a kind of 'deck of cards' type randomness which would apply hits is a more consistent but still semi-random way.
In lay terms a list of numbers from zero the the size of the area being worked on is shuffled and then decay hits are applied to each location as they are 'drawn', once a cycle is completed the deck is reshuffled and a new one begins. So long as their is full randomness within each cycle and cycles do not repeat in any any predicable way the player never be able to predict decay outside of knowing that hits can not come any more often then back to back and that over a certain time frame (a week?) everything will take at-least 1 hit. In practical algorithm terms a number of tricks can be applied to avoid literally shuffling numbers around every cycle, a large number of arrays would be created and shuffled at start-up with a different one chosen randomly each cycle as well as a different start location within each (existentially cutting the deck), also its a simple matter to do multiple layers say a top list of 1000 controlling sub lists of 1000 doing the work of a list of 1 million entries, its just a memory/speed trade off issue. If done properly it should not be noticeably slower (and might even be faster) then the current system.
While I'm sure more pressing design maters are going to be in the updates for the near future I think this change would help to resolve some long standing issues with decay and make it a more robust system to build on for later improvements like weather, climate, natural tree life-cycles and many other aspects of the ecology, particularly with regard to orchards and tree-farms which are at present quite unpredictable.
-
serpentyngallery
- Posts: 244
- Joined: Thu Apr 01, 2010 12:32 pm
Re: Algorithmic improvement of Decay
Ooo, gotta dig this one up from the grave, it sounds like a neat idea. +1 and as for the guy who put up the weather search... wouldn't have found it without you, thanks.
Ezekiel wrote: He called me what I believe to be an "ass fetus" in Russian and then killed me.
-
theTrav
- Posts: 3464
- Joined: Fri May 29, 2009 11:25 pm
Re: Algorithmic improvement of Decay
I reckon loftar probably knows about that option. Read some of his writings. He is a uni student but not the average dummy out of high school
-
Granger
- Posts: 9255
- Joined: Mon Mar 22, 2010 2:00 pm
Re: Algorithmic improvement of Decay
Decay is good as-is. Having all tiles getting hit on every cycle would suck: you would have to repair everything, replow the whole field, ...
What could change is decay resistance of some type of stuff, but J&L are tweaking that from time to time.
What could change is decay resistance of some type of stuff, but J&L are tweaking that from time to time.
⁎ Mon Mar 22, 2010 ✝ Thu Jan 23, 2020
-
serpentyngallery
- Posts: 244
- Joined: Thu Apr 01, 2010 12:32 pm
Re: Algorithmic improvement of Decay
True, didn't think about that... I guess I shouldn't have brought this one back to life, but it was something I hadn't thought of, and sounded neat... Meh, back to drilling holes in logs.theTrav wrote:I reckon loftar probably knows about that option. Read some of his writings. He is a uni student but not the average dummy out of high school
Ezekiel wrote: He called me what I believe to be an "ass fetus" in Russian and then killed me.