This week I’ve been continuing adding persistent town loot back into the game which lead into a deep dive into our spawning system.
After setting up a whole bunch of extra loot spawn points in all the towns (extra points compared to the set that the loot frenzy events run off) and configuring our spawn director to spawn loot into them I noticed that the spawns weren’t really behaving as expected and often the explicitly set loot points weren’t being used and the system was falling back to random positioning instead.
This led to me going deep into our spawning system to find bugs and get to the bottom of what was going on. I ended up finding a couple of problems and regressions since legacy, mainly a race condition with our loot position marker registration and our spawn director initialization, also another bug where SpawnerCellUpdatesPerFrame in ServerConfig was being ignored due to cell updates not being tracked properly.
With the spawn director fixed up I’m now moving onto setting up then iterating over and tweaking the spawn configurations for these towns trying to find the sweet spot in loot density and respawn times.
We’ll be releasing the next experimental patch on Thursday the 13th of December which will include the return of persistent town loot, rebalancing the vehicle system (removal of insurance and vehicle destruction, return of destructible wheels and making building a vehicle more achievable), return of the large mining rocks in the final biomes and craftable dynamite along with a bunch of bug fixes and other small changes.
Been another interesting week of collecting footage for the trailer with a focus on PVP combat. Last week we found the major dificulty was trying to achieve realistic gameplay footage yet keeping the visuals fluid enough to cut nicely in with rest of the cinematics.
We captured various scenarios including early stage combat, raiding and long range gunplay. As we progress through the trailer we keep finding better techniques during the capture that i want to implement across the board. Was lots of fun, however now I have a heap of footage to trawl through, I better get stuck in.
Hey Hurtworld Fam,
This week has involved a lot of trailer footage shooting and more progress on the Mozzy damage model. Things are coming along nicely but won’t be ready for the rolling out in the patch on the 13th. My hopes are that the upcoming patch will put the meta in a state we are happy to officially release V2 to the main branch based on the feedback we’ve been collecting from the commuinity. If everything goes well this will likely be Early Jan. We still need to do a serious optimisation pass and stabilise a few core bugs but nothing I think we can’t handle by then. With Mils working through LODs for everything and Splatt consolidating map resources we should have V2 running faster than V1 by release.
V2 will be rolling out without a “No EAC” version, so if you are still one of the few people who can’t connect to EAC servers please notify us on the forums so we can get it sorted before release.
Hey ya’ll. So we have some easy wins performance wise that we hope to achieve pretty quickly in regards to poly counts in the game. So far we have made a fair amount of LOD’s (Level of Detail) meshes in Hurtworld, but we haven’t always had time to create LOD’s for everything. Sometimes this is due to not knowing if a mesh will stay in the game long term and sometimes we just need it in game asap and don’t have time to make them.
Good news is, we now are using Simplygon to create our LOD’s and it’s super fast. In the past we have hand built these and it is time consuming. They look great using this method but Simplygon can do this in 10-20 seconds. The results are almost as good too. This was also very daunting when it came to skinned mesh attachments. Skinned mesh attachments are anything that use a skeleton to deform the mesh ie; clothes, body parts etc. Simplygon also instantly creates the skinned mesh LOD’s with some kind of evil beautiful code magic and the new meshes instantly deform just like the original. SO. GOOD.
The past week I have been testing the reliability of the resulting meshes that have been coming out of Simplygon. I’ve been fiddling and tweaking the settings to get a nice custom preset or two which so far creates great LOD’s for all situations.
Below: The top row shows completely Simplygon generated LOD’s and the bottom row shows the same but with hand tweaking of the lowest LOD mesh. In the end you can barely tell the difference and the Simplygon mesh came out less quality but lower, so we’ll automate everything and go back and fix anything that stands out in the game as we play over time.
I have 580 items to check through, not all will need new LOD’s created but a bunch will so I will be busy on that grind for the next week I’d say.
For all you people out there running slower machines, this should really benefit you. Also all you car hoarders who like to keep 50 cars in your garage for bragging rights this will help with that also.
Also I made Coal 😛
The world item that drops on the ground for coal and the Icon looked freaking horrible so I quickly remade this.
Hello again, this week I’ve run through the rest of this list I have of manual touch ups the map always needed, managed to get through it all and get a good build of the map out. Feels good being able to just hit stamp and know that everything will just work, and even if a couple things here and there screw up, it’s still a million times better than it was before. Fixed up a bunch of biome manager stuff like no builds being slightly off and radiation zones creeping too far on the playable zones. Also blended the biomes a lot smoother so it’s not such an immediate switch from one to the other. Next step is sorting through a bunch of doubled up splats and building LoDs for the town events as they are currently hitting performance quite hard. No pictures this week sorry.