четверг, 19 июля 2018 г.

Event system works great. Debugging

After moving from update for each tower towards event system (producer->consumer) I made it fast again.

Using too much Update can hurt performance even if objects count is low.

With this architecture decision I made 111fps from 15fps. So good to view this.

Now I has interesting bugs to overcome - when lower modules was destroyed - stack should collapse on one row to down.

After subscribing to Unity Plus I found analytics page. Will use it in game to see how it feels and update it to become more player friendly.

вторник, 17 июля 2018 г.

Porting complete - profiling

Porting to latest Unity 2018.2 is complete. Now I profile my code. Garbage collector spikes is not terrible (but I can override them with object pools).

Subscribed to Unity plus and now use editor with black interface. Wishes to store/build on cloud, but Unity says that it should have 3 seats licenses to have team (to use cloud build/cloud storage)

My code for grid row movement uses too much time (40ms - 15fps). So I rebuild it to subscription model and now test it.

Physics is overkill for my abstract game. And it not feel right - top down view is preferred to not use physics (and stacking physics objects give me some errors in calculation and falling blocks)

вторник, 10 июля 2018 г.

Plan to port to Unity 2018.2

Implementation of async module movement is slow. And profiler in Unity 4 is not accessible. So I must move to the latest version of unity.

Now I understand why many game developers said: "Your project should not see same season twice"

The latest attempt to port to unity 5 is failed because of many assets I have used. Now I think to use only code and prefabs with small assets that can not be removed easily (or upgraded to latest unity version)

Good news - I can use Fungus for visual novel part of my game. And it fully production ready and can provide me with all the tools I need.

Now I understand - if you make a game that spans more than a year you should write your own engine to remain in control of changes and can fix bugs for yourself.

Unity 4 keep exiting when I develop my game (maybe 2-3 times per day). I get used to restart it and wait some time. Think to move project and unity to SSD as well to speed startup time.

I should finish this project with Unity (to minimize loses) and publish it.
And I will use Unity again after trying something else (ala Defold/Godot or own engine written in C# using MonoGame)

воскресенье, 17 июня 2018 г.

Updated site for studio

Google site moves to new version with easier design and content management.


Now easier to collect my games in one place.
Use itch.io only for good games and demos. But I wish to share all my creations. And google sites is good for it

суббота, 16 июня 2018 г.

Moving from physics and towards ZenJect

Setup classes and components is tedious

When I have the small amount of towers - all fine. But now each tower (from 30 models) has 4-7 components. And all should be configured on creation.
This is very tedious and error-prone process. After investigation of my tower registry, I found that some towers do not have valid components on them.

Found ZenJect and will try to setup it to inject all controllers/sensors automatically on launch.
Maybe it will lessen the pain of support this large model line.

Physics 2d is difficult beast

After fixing too many errors I found, that physics has many drawbacks. I need the only movement and checking when to stop. And it can be done by code (and fast).

All bugs with repelling objects and falling apart grid will be gone when I remove all physics interaction. And make them only trigger.

Maybe this is overkill to move them by myself. But I find it easier than fix every time new bug in physics interaction.

суббота, 9 июня 2018 г.

Combination logic. Tweaking gameplay

Now it can be played for free. I decided to give alpha in player hands so they can help me improve it with feedback.

Grab it here
After some awhile I fixed combination logic and boosted energy generation so player can have much more towers fast. Will tweak it to make game more engoying.

I enjoy when bigger towers waporize enemies with "who-o-osh"-es. So fun to see big towers do it fast.

This can be interesting though, that I will tweak generation of energy more. Game now faster.

When generator has small output power - game feels more like survival. When you have powerful generator - it feels more as arcade. I have thoughts to provide ability to upgrade generator or make it permanent (per mission).

пятница, 11 мая 2018 г.

Debugging cannons logic

After awhile started developing BoomTris again. Nothing can stop you if wish to make it alive.

Now use hacknplan to track progress. Trello is great tool, but game development have too many hats to wear and it should be grouped. And statistics is great also.

Debugging cannons and other towers to work fine (ironing bugs).

Life can interfere and put you back with disappointment, lack of direction and motivation. Health can float down and after healing - only one thing that keep me moving forward - it stubborness. I should make it.

Three years of development should be made to finish. Maybe this year (I hope).

Developing small games is helpful to bring joy to game development when you stuck with current project.

And next one should be smaller. I have enormous pile of experience and work with clever people in game dev because of this project.

So good adventure.