![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
3:58am - I just realized my last work post covered almost 24 hours. Granted, I accidentally fell asleep for eight hours during it, but wow. No wonder I felt so burned out today.
I wasn't going to do any work today since I worked so much yesterday, but after playing a game for a while I just want to do more.
So, I have two things left to take from the RogueSharp library, and they're both Pathfinding. One's A* and the other's Djikstra Maps. But I already determined that I didn't like the way RogueSharp does A* because it doesn't use cell weights or types. It's too generic and only finds paths on a flat floor in a standard dungeon. The Djikstra Maps implementation seems okay but I only gave it a cursory look. I'll start by looking over it again.
4:22am - Well it looks like the Dijkstra Maps implementation is also flat, which I expected, but it's still pretty good for what it is. Anyways I think since I'm going to be using tile types instead of just flat tiles I'll have to roll my own for both.
Also, oops, I've been spelling Dijkstra's name wrong. XD YouTube corrected me when I went to look for a tutorial on it. Mostly there's just this one: https://www.youtube.com/watch?v=2ExLEY32RgM&ab_channel=InternationalRoguelikeDeveloperConference
The rest of the results are for his actual pathfinding algorithm, which is different and based more on the real world. I'm not sure Dijkstra came up with Dijkstra Maps, I think they were just named after him because they used some of his research. But I could be wrong.
Then for classic A* (which I think is based on Dijkstra's actual algorithm) there's Sebastian Lague's excellent (and already 5 years old!?) video here: https://www.youtube.com/watch?v=-L-WgKMFuhE&ab_channel=SebastianLague
Then on a casual search for Roguelike Pathfinding I found a more recent tutorial by the recent star miziziziz here: https://www.youtube.com/watch?v=rbYxbIMOZkE&ab_channel=Miziziziz
4:41am - Just watched some random vids on pathfinding that I found interesting, but I should actually get down to desgining shouldn't I?
Now, if I'm going to add weights to my cells I should probably figure out how I want to do that. I figure the best way is to add it right into the Cell class. There'll be a built-in weight for every cell based on its type, and the Pathfinding system will have to take entities into account as well, including effect entities. So like, if I want to implement mud that covers a tile of a different type, I don't want to change the tile's type, I just want an effect of mud.
This will be a pretty complicated endeavour I believe. I'm going to review A* pathfinding now by watching that Sebastian Lague series and make sure I know what I'm doing.
6:03am - Alright then, I've made sure I know what I'm doing and it seems like it should work out. I also did a bit of research on the Godot Astar implementation and I've decided to go ahead and just make my own. For now though I've been at it for a couple hours and I think it's best I take a break.
8:07am - Played a game for an hour, then spaced out for an hour, and now I'm back. Gonna make some Pathfinding classes.
10:08am - And we have basic pathfinding. I've been working for another two hours so I'd better take another break.
12:10pm - Oh yeah I should note that I'm back to work again.
1:21pm - I think I'm spent, and I'm expecting a phone call in ten minutes, so this is a good time to wrap up for the day. I need to re-do the Heap impleentation I was working with, it doesn't work right for what I'm trying to do.
I wasn't going to do any work today since I worked so much yesterday, but after playing a game for a while I just want to do more.
So, I have two things left to take from the RogueSharp library, and they're both Pathfinding. One's A* and the other's Djikstra Maps. But I already determined that I didn't like the way RogueSharp does A* because it doesn't use cell weights or types. It's too generic and only finds paths on a flat floor in a standard dungeon. The Djikstra Maps implementation seems okay but I only gave it a cursory look. I'll start by looking over it again.
4:22am - Well it looks like the Dijkstra Maps implementation is also flat, which I expected, but it's still pretty good for what it is. Anyways I think since I'm going to be using tile types instead of just flat tiles I'll have to roll my own for both.
Also, oops, I've been spelling Dijkstra's name wrong. XD YouTube corrected me when I went to look for a tutorial on it. Mostly there's just this one: https://www.youtube.com/watch?v=2ExLEY32RgM&ab_channel=InternationalRoguelikeDeveloperConference
The rest of the results are for his actual pathfinding algorithm, which is different and based more on the real world. I'm not sure Dijkstra came up with Dijkstra Maps, I think they were just named after him because they used some of his research. But I could be wrong.
Then for classic A* (which I think is based on Dijkstra's actual algorithm) there's Sebastian Lague's excellent (and already 5 years old!?) video here: https://www.youtube.com/watch?v=-L-WgKMFuhE&ab_channel=SebastianLague
Then on a casual search for Roguelike Pathfinding I found a more recent tutorial by the recent star miziziziz here: https://www.youtube.com/watch?v=rbYxbIMOZkE&ab_channel=Miziziziz
4:41am - Just watched some random vids on pathfinding that I found interesting, but I should actually get down to desgining shouldn't I?
Now, if I'm going to add weights to my cells I should probably figure out how I want to do that. I figure the best way is to add it right into the Cell class. There'll be a built-in weight for every cell based on its type, and the Pathfinding system will have to take entities into account as well, including effect entities. So like, if I want to implement mud that covers a tile of a different type, I don't want to change the tile's type, I just want an effect of mud.
This will be a pretty complicated endeavour I believe. I'm going to review A* pathfinding now by watching that Sebastian Lague series and make sure I know what I'm doing.
6:03am - Alright then, I've made sure I know what I'm doing and it seems like it should work out. I also did a bit of research on the Godot Astar implementation and I've decided to go ahead and just make my own. For now though I've been at it for a couple hours and I think it's best I take a break.
8:07am - Played a game for an hour, then spaced out for an hour, and now I'm back. Gonna make some Pathfinding classes.
10:08am - And we have basic pathfinding. I've been working for another two hours so I'd better take another break.
12:10pm - Oh yeah I should note that I'm back to work again.
1:21pm - I think I'm spent, and I'm expecting a phone call in ten minutes, so this is a good time to wrap up for the day. I need to re-do the Heap impleentation I was working with, it doesn't work right for what I'm trying to do.
