Hordes.io retweetledi
Hordes.io
347 posts

Hordes.io
@hordesio
Survive faction battles and build your clan. Browser-based MMORPG.
Katılım Eylül 2016
111 Takip Edilen1.1K Takipçiler
Hordes.io retweetledi
Hordes.io retweetledi
Hordes.io retweetledi
Hordes.io retweetledi
Hordes.io retweetledi
Hordes.io retweetledi
Hordes.io retweetledi
Hordes.io retweetledi
Hordes.io retweetledi

Micro Devlog Time
PTR Test will run on Nov 4th, sign up if you wanna join :)
Mainly spent time on the spell engine, most of these bugs related to obscure rules for parsing order of spell interactions (e.g. buff queued entity while alive, dies in same frame, buffs added after death flag).
In order to do decently comprehensive testing on massive combat engine changes, i'll generally just generate 500 bots that all fight each other + monsters in the same spot, with action priority lists that include every spell and buff in the game. Ultimately though the best testing suite would be a fully implemented APL system that acts deterministically.
While testing I also found some decently easy to implement perf improvements, which resulted in some massive speed ups on the server. Expecting to roughly 1.5x the amount of connected players we can handle.
Also added a login queue.. 2 days went into that unfortunately, that stuff was overly complex for side cases such as players reconnecting during the queue or how to handle players that log out for a short time (dont wanna force an actively playing player to re-queue, ever).
Character creation limit lifted, its now a moving window that limits you from creating too many characters in 10 minutes, 1 day, and 1 month.
Remaining task for the next PTR test are mostly non-critical. I'll do some visual i terations on guardstone so there will be more pretty vids and pics.
Have a good weekend!


English
Hordes.io retweetledi
Hordes.io retweetledi
Hordes.io retweetledi
Hordes.io retweetledi
Hordes.io retweetledi

Open World Editing Trick - Modulate heightmap based on texture values, rather than painting textures based on heightmap.
1. Paint Macro Heightmap features like Mountains
2. Do a texturing pass
3. Modulate heightmap based on texture
(Advantages of writing custom editors)
Reason i think this is better:
Its easy to reason about large scale features like mountans and little hills at low vertex counts, but when it comes to detailing like this, doing it "blindly" without any texture values is difficult to intuit. when you only see the mesh, you'd never modify the heightmap like that because it just seems arbitrary. The height map is a lower detail density thing that supports your much higher detail texture information by giving it a good silhouette.
IMO its these kind of details that end up creating a lot of depth and make things more believable. I only discovered this approach for myself today - incredibly satisfying to apply to everything.
English
Hordes.io retweetledi
Hordes.io retweetledi
Hordes.io retweetledi
Hordes.io retweetledi
Hordes.io retweetledi

