[Devlog] May 2025 - Week 1
Welcome to the week 25 devlog for the development of Cosmic Construct. In these devlogs I go over development goals set at the end of the post from the previous week. However, this week did not have anything set, so it will be more freeform than these usually are. This will go for next weeks as well, before returning to normal.
If you would like to follow development of the game, there are a few ways you can do that.
With that out of the way, I normally have a planned work section. Since there is no planned work, we are just going go over what was done in not particular order.
---
Pawn Interactions
Code for Pawn Interactions is underway. Pawn Interactions are going to mostly replace the system of pawn report logs, and switch over to an "animal crossing like" conversational system to get feedback from pawns. I don't entirely know what this will look like after it is finished, but there will be chat cut-ins when interacting with a pawn. It should bring more life and character to the game, and provide a means to get a higher resolution art of each of the pawns. Which will make the tailoring more visually interesting.
Pawn AI Improvements
Since all of the new crops were added, they were not recognized by the pawns doing the general gathering job. General Gathering still remains as less useful than intended, due to how the jobs are broken down. I neglected to add a general farming job, which is unfortunate. I'll be reorganizing the job lists so that they make more logical sense in the early game. Instead of having the player select between specific harvestables objects like berries, wheat, or linseed.
The search function was very, very poor computationally. I was using true diagonal distance, to find the nearest grid square. Then even worse, checking all 10,201 squares diagonal distances each time a pawn changed a state. This was fine when I wasn't checking over a list of possible objects to find. The second I needed to check the entire list of harvestable objects, it became O(n^3) complexity and ran the game down to 40 FPS.
I rewrote all of the grid search functions to check manhattan distance in a spiral. This is good because pawns, when harvesting a crop, are likely surrounded by many other similar crops. So it will terminate on an adjacent square almost immediately. Also, nowhere in that check is a square root ever used. This might not be a joke, but it could in some cases be more than 100,000 times more computational efficient than what I was doing before. That is how bad the search algo was.
I did everything I could to stress the game out, and could still do double that and run above 60 FPS. The screenshot above is the worst framerate dip I could find.
Game Balance
About half of the buildings still don't have build costs because none of the game testing has gotten to that point of gameplay. However, this is going to change in the upcoming weeks as I start building the mid game. I figured I would go ahead and do the math necessary for all of those buildings, and see what level of commitment a player needs to craft those items. I'm trying to keep from timegating the player from having access to the resources they need to make progress. Which is why the Market was added as a structure. The idea is that once a colony is established, the player can just buy most of the common resources instead of having the harvest them.
New BGM Tracks
4 New BGM tracks were added to the game. Cosmic Construct has two music playlists, common and rare. All 4 of the new songs are in the common rotation, which makes them just as likely to show up. In total 16 minutes of BGM audio, which brings the grand total to 1 hour and 40 minutes. I still don't think this is enough, but I probably will not add any more BGM audio until after the game is released.
Recording For Trailers
I have committed to two gameplay trailers for a couple of events that are going to happen in June. Once those trailers are finished, I'll be going on a small marketing run presenting the game to some interested people. The final pre-release trailer will be for Steam Next Fest. I don't know which next fest Cosmic Construct will be participating in. However, I have hard numerical requirements and goals that need to be meet before I consider sending the game out streamers.
Game Testing
Much of my time this week has been spent doing game testing and tweaking what I see. Removing annoyances, and shifting direction way from mechanics that cause those annoyances. This is part of the reason I am lessening the Logging mechanic, in favor of a more interactive means of communicating with pawns.
---
As like last week, I will be doing work that I do not have planned.
I will be preparing a new steam demo, and do not know yet what is required to have it ready to play.
If you would like to follow development of the game, there are a few ways you can do that.
- My bluesky account where I post frequent updates.
- I also have a Cosmic Construct community discord server.
- Don't forget to Wishlist Cosmic Construct on Steam.
- Follow the steam page if you want more of these weekly devlogs
With that out of the way, I normally have a planned work section. Since there is no planned work, we are just going go over what was done in not particular order.
---
Pawn Interactions
Code for Pawn Interactions is underway. Pawn Interactions are going to mostly replace the system of pawn report logs, and switch over to an "animal crossing like" conversational system to get feedback from pawns. I don't entirely know what this will look like after it is finished, but there will be chat cut-ins when interacting with a pawn. It should bring more life and character to the game, and provide a means to get a higher resolution art of each of the pawns. Which will make the tailoring more visually interesting.
Pawn AI Improvements
Since all of the new crops were added, they were not recognized by the pawns doing the general gathering job. General Gathering still remains as less useful than intended, due to how the jobs are broken down. I neglected to add a general farming job, which is unfortunate. I'll be reorganizing the job lists so that they make more logical sense in the early game. Instead of having the player select between specific harvestables objects like berries, wheat, or linseed.
The search function was very, very poor computationally. I was using true diagonal distance, to find the nearest grid square. Then even worse, checking all 10,201 squares diagonal distances each time a pawn changed a state. This was fine when I wasn't checking over a list of possible objects to find. The second I needed to check the entire list of harvestable objects, it became O(n^3) complexity and ran the game down to 40 FPS.
I rewrote all of the grid search functions to check manhattan distance in a spiral. This is good because pawns, when harvesting a crop, are likely surrounded by many other similar crops. So it will terminate on an adjacent square almost immediately. Also, nowhere in that check is a square root ever used. This might not be a joke, but it could in some cases be more than 100,000 times more computational efficient than what I was doing before. That is how bad the search algo was.
I did everything I could to stress the game out, and could still do double that and run above 60 FPS. The screenshot above is the worst framerate dip I could find.
Game Balance
About half of the buildings still don't have build costs because none of the game testing has gotten to that point of gameplay. However, this is going to change in the upcoming weeks as I start building the mid game. I figured I would go ahead and do the math necessary for all of those buildings, and see what level of commitment a player needs to craft those items. I'm trying to keep from timegating the player from having access to the resources they need to make progress. Which is why the Market was added as a structure. The idea is that once a colony is established, the player can just buy most of the common resources instead of having the harvest them.
New BGM Tracks
4 New BGM tracks were added to the game. Cosmic Construct has two music playlists, common and rare. All 4 of the new songs are in the common rotation, which makes them just as likely to show up. In total 16 minutes of BGM audio, which brings the grand total to 1 hour and 40 minutes. I still don't think this is enough, but I probably will not add any more BGM audio until after the game is released.
Recording For Trailers
I have committed to two gameplay trailers for a couple of events that are going to happen in June. Once those trailers are finished, I'll be going on a small marketing run presenting the game to some interested people. The final pre-release trailer will be for Steam Next Fest. I don't know which next fest Cosmic Construct will be participating in. However, I have hard numerical requirements and goals that need to be meet before I consider sending the game out streamers.
Game Testing
Much of my time this week has been spent doing game testing and tweaking what I see. Removing annoyances, and shifting direction way from mechanics that cause those annoyances. This is part of the reason I am lessening the Logging mechanic, in favor of a more interactive means of communicating with pawns.
---
As like last week, I will be doing work that I do not have planned.
I will be preparing a new steam demo, and do not know yet what is required to have it ready to play.