So much has changed since the initial "Twin Demon Slayers" game (made for a game jam), that was the origin of all this.

First of all, I will stop using the original game title from the jam. The current game evolved into something much more than its original. One thing remained contstant though - its combat system does resemble the one XCOM.

Also, I've decided to put the game in a steampunk-dystopian settings - hence the title of this blog post.

Anyway, here's the progress made in May (well, in the second part of May - as previous blog post was written on May 14th):

1. Overwatch mechanism

2. New graphics

3. UI/QoL changes

4. Performance improvements

1. Overwatch mechanism

The idea behind overwatch is simple - a character can spend their action to going into an overwatch state - this will allow them to perform a ranged attack during enemy turn - if the enemy moves within the shooting range.

Such mechanics allows for more tactical depth during combat - but let's not get into details here, as XCom has already exercised this concept.

Anyway - do see it in action below: both player characters enter overwatch (and an icon indicating this is being displayed). Then they take their reaction shots when the enemy starts closing in on them.

2. New graphics

So I knew I told myself to wait with the graphics until all core game features are ready. However I came across opportunity to work with a great pixel artist (whose name is Kurt Prieto) and I didn't want to waste such chance. The result of our cooperation are this first concepts of the rogue/ranger character type - check them out:

Rogues, baby!

Now, the important thing here was to make sure that the game environment (which is tile-based) works correctly with new character model. Previously (when using my 'programmer art') this was easy - a character was of the same height as other tiles. This did not look realistic, but at least did not introduce any extra issues.

The new models are taller, and I had two possible ways to make sure that they blend in the tile-based word:

#1 I will make sure that they are still of the same height as the other tiles, or

#2 The character models will be higher than the tiles.

Big-tile approach on the left, small-tile on the right. For me, the 2nd one is a definite winner.

Approach #1 was looking too rough for me. The price of making sure that the characters and map tiles always take up the same cell space came with a huge price - the allowed character density was not good enough. Also, the walls become too bulky; basically the game scene contained too much empty space.

So I've chosen approach #2. This of course introduced other issues - especially with visibility behind parts of the walls, that I will address in a while.

3. UI/QoL changes

So a good and smooth user experience is something I always value the most in computer games. I will make sure that the same could be said about the game I am working on.

First major improvement is the cover display at the movement target - basically the user is able to see the cover  that each obstacle could provide. Also, if the character would be flanked at the given location, the covers are displayed on yellow.

Yeah, I know that enemies are using the same sprite as the player character.

You can already see an issue with this in the gif above: the user is not able to see the cover information on the tiles which are hidden behind walls. Same for the display of characters themselves. Maybe such walls should become transparent?

Another improvement is targets preview. When the player holds the alt key, they are able to see all potential ranged targets that will become available at the given location - along with flanking information.

Neat features, aren't they? And sure, XCom had them too. And as they were extremely helpful there, I am sure that they will be useful in my game as well.

4. Performance improvements

Yes, I know that early optimization is a bad practice - but surely not in a situation when the pathfinding code starts to tank the framerate when it has to deal with 50x50 tile map. And sadly that was happening before the improvements.

The solution was simple - to make sure that the A* pathfinding algorithm only works on a limited part of the map; basically it now does not perform the calculations outside of the movement range of a character.

What's next?

I am constantly working on making the tactical combat playable. My plans for June include:

• area of effect attacks, including grenades,
• armour mechanics (reducing damage, armor piercing mechanism),
• changes to ammo mechanics (will try out a different approach than the one in XCom!).

Stay tuned and let me know what you think so far!