Pinned Tweet
I am so proud of this community to give us the strength and motivation to post this. The team is locked in and I hope you are ready for what is to come. 🥰hytale.com/news/2025/11/h…
23
40
852
216,052
Slikey retweeted
We dont really have a concurrent playercount target as it heavily depends on the way you use the server and such metrics often end up overly academic with too many variables.
The main problem is the nature of realtime simulations. If you have many players in one spot with very few NPCs the count is different compared to many players into completely different parts of the world with thousands of NPCs and many chunks loaded.
Its different again if you completely seperate the worlds through instances. Generally we want to have 100-250 player servers with vanilla exploration / adventure to be possible if you have the hardware.
Right now the server still depends too much of single-core performance and we are not leveraging the concurrency capabilities of the ECS enough. We lack a scheduler in the ECS which is capable to parallelise work where it's possible.
Giving you a concrete example of a simulation example:
(1) Imagine you have a solar system: Every body affects the other bodies, so you have to split the logic into a gravity calculation, force application and velocity application step. It's possible to parallelize.
(2) Imagine you have a collision system of multiple bodies colliding. Who bounces where? There needs to be a "communication" between the bodies - it's much harder to parallelise.
This is the core of the problem - we need that scheduler that plans the work and we need every system to be looked at and optimized / rewritten to enable splitting work.
On top of all of this comes the fact we don't just have 32-core CPUs but also 4-core CPUs and some people even just 2-core where the overhead of parallelisation isn't even worth it and further tanks performance on already weak systems.
I think we will make progress but it will be snail's pace.
The main problem is the nature of realtime simulations. If you have many players in one spot with very few NPCs the count is different compared to many players into completely different parts of the world with thousands of NPCs and many chunks loaded.
Its different again if you completely seperate the worlds through instances. Generally we want to have 100-250 player servers with vanilla exploration / adventure to be possible if you have the hardware.
Right now the server still depends too much of single-core performance and we are not leveraging the concurrency capabilities of the ECS enough. We lack a scheduler in the ECS which is capable to parallelise work where it's possible.
Giving you a concrete example of a simulation example:
(1) Imagine you have a solar system: Every body affects the other bodies, so you have to split the logic into a gravity calculation, force application and velocity application step. It's possible to parallelize.
(2) Imagine you have a collision system of multiple bodies colliding. Who bounces where? There needs to be a "communication" between the bodies - it's much harder to parallelise.
This is the core of the problem - we need that scheduler that plans the work and we need every system to be looked at and optimized / rewritten to enable splitting work.
On top of all of this comes the fact we don't just have 32-core CPUs but also 4-core CPUs and some people even just 2-core where the overhead of parallelisation isn't even worth it and further tanks performance on already weak systems.
I think we will make progress but it will be snail's pace.
3
6
80
1,746
What really strikes me on reddit everytime I go there is the amount of cynical comments.
I discovered that life is so much nicer when you remove cynical people from your life as they are not interest in solutions - all they do is see problems without an positive outcome and drag everyone around them down with them. It's really this fundamentally hopeless existence that I can't and don't want to understand.
It is a major reason I am more active on X rather than Reddit and I think it will only continue. I was a lot more active on reddit in the beginning of Hytale but it really fizzles out. On X its a lot easier to start with and I just mute cynical people.
I guess the modern word for them is "doomer".
I discovered that life is so much nicer when you remove cynical people from your life as they are not interest in solutions - all they do is see problems without an positive outcome and drag everyone around them down with them. It's really this fundamentally hopeless existence that I can't and don't want to understand.
It is a major reason I am more active on X rather than Reddit and I think it will only continue. I was a lot more active on reddit in the beginning of Hytale but it really fizzles out. On X its a lot easier to start with and I just mute cynical people.
I guess the modern word for them is "doomer".
21
17
393
17,960
And to be clear, I don't just mean the hytale subreddit - it is everywhere on reddit. It's not about the occasional sarcasm or irony - it is about the people who see something nice and they immediately spin it into a negative direction.
Imagine you are a child and proud of something you learned and show your parent, just for the parent to say "you are still making mistakes and other kids can do it better".
These people are unhappy with themselves and project it on everyone around them. Cut them out.
Imagine you are a child and proud of something you learned and show your parent, just for the parent to say "you are still making mistakes and other kids can do it better".
These people are unhappy with themselves and project it on everyone around them. Cut them out.
5
0
90
1,821
My take on this: SBMM is often poorly done in multiplayer team games. For players it often feels like they are thrown into team that matches another team, while the individual performance and playstyle is secondary.
Generally X% of matches are lost no matter what you do. Y% of matches are won no matter what you do. And finally Z% of matches actually depend on your performance.
SBMM in team games with randoms feels like the X and Y are way too large. The expectation should be that you always have to try and this is actually the case when you reach the very top of the ladder.
The issue is really in the middle where you also have strong differences in teammate performance - I have seen players and streamers go full tilt and enter a massive decay streak losing match after match because they play well below their performance.
The other thing is that in top play you generally have people who mastered the game and they can swap roles or tactics to play the other team. The middle ranks often have people stuck with one-tricks and single tactics which gives them a certain win-rate but it keeps them there, so that it's very easy to get frustrated when this player rather loses than play what is needed.
Generally SBMM is supposed to fix something but just shifts the problem and you end up with players rather wanting none again and occasionally run into the pro with no real-life to get absolutely railed but at least not lose rank points or ELO for losing.
Generally X% of matches are lost no matter what you do. Y% of matches are won no matter what you do. And finally Z% of matches actually depend on your performance.
SBMM in team games with randoms feels like the X and Y are way too large. The expectation should be that you always have to try and this is actually the case when you reach the very top of the ladder.
The issue is really in the middle where you also have strong differences in teammate performance - I have seen players and streamers go full tilt and enter a massive decay streak losing match after match because they play well below their performance.
The other thing is that in top play you generally have people who mastered the game and they can swap roles or tactics to play the other team. The middle ranks often have people stuck with one-tricks and single tactics which gives them a certain win-rate but it keeps them there, so that it's very easy to get frustrated when this player rather loses than play what is needed.
Generally SBMM is supposed to fix something but just shifts the problem and you end up with players rather wanting none again and occasionally run into the pro with no real-life to get absolutely railed but at least not lose rank points or ELO for losing.
Can someone knowledgeable explain to me why people complain all the time about SBMM (Skill based Match Making) or put it as a terrible, terrible thing for players?
To someone that doesn't play shooters competitively it sounds like a good thing. Pros play pros, noobs play noobs.
3
3
69
3,943
Slikey retweeted
First version of the downhill
The road is procedurally generated and is infinite (that's why I was creating the infinite dropper, it won't be released as a mod). The skateboard is still in a very rough version.
#Hytale #HytaleModding
The road is procedurally generated and is infinite (that's why I was creating the infinite dropper, it won't be released as a mod). The skateboard is still in a very rough version.
#Hytale #HytaleModding
20
54
650
24,018
Important: It is a preview as we are still working on it but it is a crucial step to get feedback early and pave the way to bring the node editor into the game and also pave the way for visual scripting :)
I am really proud of our tools team to bring this new Noesis based editor into the engine <3
I am really proud of our tools team to bring this new Noesis based editor into the engine <3
Hytale's New Asset Editor has been released in Early Preview with pre-release v0.8.0-pre.4.
Enable via Settings -> General -> Development -> Use New Asset Editor.
13
16
228
5,355
You can now. Enjoy 👌 If there is something missing or not modular enough, you can hit me up and I can try to get it into U8 before release.
This is super impressive work guys. It can’t have been easy to make this fully compatible with existing worlds not having it change the seeds at all. I can’t wait to dive into playing with making new biomes for wg v1 and having new ores, structures, and more.
9
4
221
8,484
Slikey retweeted
We grew up on Tekkit and Feed The Beast, so we started building "Rewired" for @Hytale.
It's still early, but seeing all these machines working together brings back memories.
Check out the factory at play.aselion.com. Maybe start your own?
#Hytale #HytaleModding #Aselion
It's still early, but seeing all these machines working together brings back memories.
Check out the factory at play.aselion.com. Maybe start your own?
#Hytale #HytaleModding #Aselion
20
78
1,123
29,948
I'd solve this using rested XP and an XP fatigue but it highly depends on what audience the devs want to attract.
Generally it comes to the choice if you want players min-maxing progression or do you believe the game is fun enough by itself and progression is mostly based on season length?
Either way I'd encourage them to bring rested XP to the progression elements individually as I don't really enjoy assault but with rested XP I'd probably use that budget to level assault whenever my boost is full.
In the end the game itself is fun and it seems the devs reason for the early reset is aligned with them wanting season duration to lead the progression rather than unemployment 😅
Generally it comes to the choice if you want players min-maxing progression or do you believe the game is fun enough by itself and progression is mostly based on season length?
Either way I'd encourage them to bring rested XP to the progression elements individually as I don't really enjoy assault but with rested XP I'd probably use that budget to level assault whenever my boost is full.
In the end the game itself is fun and it seems the devs reason for the early reset is aligned with them wanting season duration to lead the progression rather than unemployment 😅
ANNOUNCEMENT 📢
Season 2 was meant to be a fresh start, and for a lot of you it hasn't felt like one. We sped up progression on deliberately and tried to balance the game better, as set out in the patch notes, but it was the wrong call.
Some players are reaching end game gear like Tier 4 armour and MBTs within days, so the wipe never really landed.
Some things are clearly out of balance and we'll fix them. Others have the community split, so we want to see how those play out first.
We won't be doing a knee jerk reset this weekend, as that would be unfair on everyone already enjoying the climb. Instead, we're looking at significantly bringing forward Season 3 and some of its content, so you get a proper fresh start. More details in our dev vlog next week.
This is what Early Access is for. Thanks for playing and for the feedback.
1
0
67
4,974
Slikey retweeted
This is time! We're again looking for animators capable of working in #Blockbench.
Please share your most impressive blockbench animations in this thread, or please RT it if you want to help someone get a job on @Hytale 💫
Please share your most impressive blockbench animations in this thread, or please RT it if you want to help someone get a job on @Hytale 💫
52
103
989
50,152
Slikey retweeted
Fixes were shipped tonight with 0.7.1 that help with players who were having issues connecting via LAN to each other.
If you are still encountering issues please submit a bug report in game and shoot us the report ID in the replies as it helps our team greatly pin down issues.
If you are still encountering issues please submit a bug report in game and shoot us the report ID in the replies as it helps our team greatly pin down issues.
8
7
153
4,692
Slikey retweeted
We are trying to publish Hytale Early Access Chapter I, but GitHub is having issues, so there's no ETA yet; you should get the update in a few minutes. We will try to re-run the deployment now.
Microsoft is just trying to slow us down (jk)
Microsoft is just trying to slow us down (jk)
107
118
3,429
85,249
Update 8 (after chapter 1) teaser for worldgen V1:
We are breaking our rule of "no changes for V1". For the last weeks we rebuilt how V1 loads its data. Until now, V1 lived in its own special corner of the server: its own file format, its own loader, file references, and no way to touch it the way you touch everything else in the game.
That'll be over. Every zone, biome, cave system, block mask, surface cover, tree and prefab pattern, and noise function will be a regular asset in the same system as items, blocks, NPCs, and prefabs. About 8,800 of them.
I. Same tools, same rules. Worldgen assets are plain JSON with a schema, so it works with the same editing, validation, and autocomplete you already use for other assets.
II. Readable names and folders. We have reorganised all V1 assets to fit the asset type ordering style.
III. Less duplication. Copy-pasted definitions were merged. The surface cover library alone went from 2,383 files to 1,171 without changing a single block in the world.
IV. Node editor authoring. V1 graphs can be edited visually, with live noise previews instead of guessing numbers. The NoisePlayer is dead.
V. Your worlds do not change. We held ourselves to one hard rule the whole way: the generated world is byte for byte identical to before. Same seed, same world. Every change was checked against thousands of generated chunks.
VI. Native modding in the overworld. You can add and modify content in v1 easily. Examples will be provided to the modding community. Existing V1 mods will no longer function.
VII. Oh.. It's also roughly 15% faster during runtime and 70% faster during boot.
The old loader and the old asset format are deleted. Worldgen V1 will still be phased out eventually and fully replaced by V2 - however it takes more time and we needed to take some pressure off the team and decided to modernise V1 ONE LAST TIME(tm).
Additionally as part of Update 8 we will make V1 compatible with cubic chunks. Details follow.
We are breaking our rule of "no changes for V1". For the last weeks we rebuilt how V1 loads its data. Until now, V1 lived in its own special corner of the server: its own file format, its own loader, file references, and no way to touch it the way you touch everything else in the game.
That'll be over. Every zone, biome, cave system, block mask, surface cover, tree and prefab pattern, and noise function will be a regular asset in the same system as items, blocks, NPCs, and prefabs. About 8,800 of them.
I. Same tools, same rules. Worldgen assets are plain JSON with a schema, so it works with the same editing, validation, and autocomplete you already use for other assets.
II. Readable names and folders. We have reorganised all V1 assets to fit the asset type ordering style.
III. Less duplication. Copy-pasted definitions were merged. The surface cover library alone went from 2,383 files to 1,171 without changing a single block in the world.
IV. Node editor authoring. V1 graphs can be edited visually, with live noise previews instead of guessing numbers. The NoisePlayer is dead.
V. Your worlds do not change. We held ourselves to one hard rule the whole way: the generated world is byte for byte identical to before. Same seed, same world. Every change was checked against thousands of generated chunks.
VI. Native modding in the overworld. You can add and modify content in v1 easily. Examples will be provided to the modding community. Existing V1 mods will no longer function.
VII. Oh.. It's also roughly 15% faster during runtime and 70% faster during boot.
The old loader and the old asset format are deleted. Worldgen V1 will still be phased out eventually and fully replaced by V2 - however it takes more time and we needed to take some pressure off the team and decided to modernise V1 ONE LAST TIME(tm).
Additionally as part of Update 8 we will make V1 compatible with cubic chunks. Details follow.
26
41
675
29,033