Monthly Treeverse Update #21
Monthly Treeverse Update #21
The twenty-first edition of monthly Treeverse updates. Read below for some game development insights on what we’ve worked on in the last month.
An update on Treeverse. A look at our development and art team progress. We will be posting monthly for you to follow along the development of Treeverse, so make sure you subscribe!
Exciting day today before the monthly update. We released a 15-minute uncut gameplay dungeon walkthrough. Check it out below:
And… vtubers! We teased them a while ago. Tomorrow they go live :).
Apologies from me, Loop, for the care package delay. I have personally signed every single note in them — and this month it is 100% coming so stay tuned. We will most likely do a pre-order (join our Discord to get the notification early!) and then a FCFS wave available to all.
Art Update
By Scott
For the month of August, we are expanding our dungeon-themed monsters and environment.
We have also streamlined the production process to maximize our output. From here on out, monsters will have a core body, and then modifications on top to reflect different elements and typing. You can see this with the beetle line, which based on different typing, will sport a slightly different look. We hope to make it evident so the players can tell what typing monster is at first glance. At the same time, we want to be even more efficient with the animations and rigging we already have. Using this method, we can achieve this goal, with the additional benefit of monster variations being more lively, avoiding them feeling cheap, which would have been the case if we just simply changed the colour of the monsters. (However, we do still plan on using simple colour changes when it’s appropriate, such as with the wolf fur material).
We are also working to try to expand our asset library. This month we are working on the skeleton dungeons, The original skeleton dungeon from pre-alpha was the cave environment, which we felt could be used for a more general purpose of an animal-inhabited environment. So we have decided to create an extra pack specifically for our skeleton-themed monster line.
Our goal is to try to refine our production pipeline to maximize the impact of the art assets that we produce as a small art team. With the new asset pack approach and monster typing, we are moving towards that goal step by step.
Engineering And Design
by Aizea
Last month I mentioned that our focus was too scattered and that we were implementing a ton of features. Whilst these features are important, they really should have been worked on after the core experience had been set. We’ve gone through a couple of important systems like the quest and interaction system which took a significant amount of time, but have now finally aligned ourselves with really focusing on the core user experience
The core experience is about cooperative dungeon raids, fighting challenging monsters in waves and battling across multiple arenas. A monster hunter-like experience where after killing the boss you have a chance to get some of the crucial resources required to upgrade your gear and progress to the next set of monsters.
Capsule Heroes has a phenomenal combat feel that has already been validated internally and externally in playtests. We’re taking some of the mechanics that make it feel so good and implementing them into the combat of Treeverse as well, which will save us a lot of iteration time.
We spent a lot of time improving our combat tool and improving our AI this month. We have significantly better target detection and selection now as well as some basic horde control, to help manage the chaos inside of dungeons.
The dungeons can get quite challenging, so in the spirit of improving cooperative play, teammates can revive each other during the dungeons.
We added spawners inside of dungeons to control how the monsters appear and when they appear after clearing each wave. The system used is quite extensible, allowing the enabling of various other things when certain actions or waves are cleared.
Previously, the combat capabilities of our abilities were quite primitive. Once it had started casting an ability, it was locked into that mode, unable to change aim or move during it. We’ve added a few simple nodes which have vastly improved how dynamic the gameplay feels.
With regards to effects, we’ve added on hit-effects, and the ability to stun enemies if we hit them with a Knockback and they smack into a wall. We have 3 main monster types. Mobs, Elites and Boss. They have various types of crowd control resistances and we have indicators for those as well. We also experimented with adding hitfreezes, and a scalable timer for abilities, letting us fine-tune the speed and feel without having to commit to different lengths of animations immediately.
Implemented Damage over time effects
Implemented perfect dodge, allowing you to avoid all damage if timed correctly using IFrames.
We are also in the midst of implementing different types of triggers for abilities based on input, allowing us to create charged abilities and combo them in interesting ways.
We implemented traps and other interactable objects many months ago, however, these became out of date and required re-implementation. They are working properly now with the improved code base.
We’ve added a telegraphing system, making combat much more readable and enjoyable. Telegraphing is basically just indicators on the ground for the damaged areas of a monster.
We onboarded a new engineer into our team, and his expertise in ECS has helped us improve the codebase and the way we write our code. We will perform more refactors as the months go by, making the code base more maintainable as the project grows in size.
Many refactors
We initially went with a huge switch statement to handle the logic for incoming packet systems. It was fine for a while, but maintenance costs became too high so we split each network command into its own system.
Ability casting and systems were refactored.
AI systems were refactored.
Target detection and selection refactored.
Many bug fixes
Refactoring code can often cause unexpected issues too, such as refactoring network commands into multiple systems revealing an issue that only happens when packets are not reacted to immediately.
We had a bug with colliders on dead entities that we fixed.
We moved all of our client scenes into sub-scenes and had to chase a lighting issue in the game for a while.
Aiming and magnitude of abilities had some issues since we used the length of the aim input to determine the range of ability.
One of our consultants added an Orca system in the game, which is a local avoidance mechanism allowing monsters to chase targets whilst trying to minimise crashing into each other.
We implemented a simple stats system in the game. Initially, we flirted with the idea of a complicated stats-based approach, however considering we are mobile-first focusing on shorter play times, we are going for a much more visual and effect-based progression system like Pokemon instead of a purely old-school stat-based type of progression.
Going Forward
For those of you who have the chance to play capsule heroes, it really is a phenomenal game. Every group we have tested with externally have thoroughly enjoyed it, the cheering and applause and laughter during game play is incredibly inspiring. We are looking forward to all of the feedback you bring to us.
For those of you eager to play Treeverse, I was really hoping to have some of you play the game by now, but it simply isn’t ready. Thank you for your continued patience.
So far, we really have been focusing on building the engine of the game rather than gameplay specific things. Hence the focus on the tooling for our AI and combat. Normally I would say that it’s taking longer than expected, but I've learnt from saying this enough times, that the longer timelines are now just expected. It takes a long time to be able to do these things well, and it just isn’t possible to rush it unfortunately. Our reluctance in giving specific dates comes from the fact that estimating software is just incredibly difficult.
Instead we hope to be able to invite players when the team is ready for it. In the meantime we hope to provide more video footage of our progress, and when the team and code base is ready for our community we will invite you to play. Promising specific dates and not being able to meet them will just burn out the team and result in more delays than if we didn’t.
Disclaimer: Please note that anything written in this document should not be taken as financial advice. Endless Clouds is dedicated to producing a fun, enjoyable game that integrates NFTs. NFTs are a new and highly experimental technology and should not be bought for investment purposes.