PC Alpha 19 Dev Diary

Status
Not open for further replies.
No, those were placeholders, the one that shipped in 18 is the closest to gold version we've ever had. The old system ate CPU like crazy and was completely different rendering system than static maps were, making a lot of headaches for everyone. At a high level all we want to add now is an improved placement system so that no duplicates spawn and all pois are pushed in and not skipped over, and some bug fixes.
Click to expand...
aaaaand hopefully bigger maps?

Mr. Mole, i know it has no priority. But when you guys do the optimizations for 1.0 i hope you have some time spare to look into that :)

 
Just more loot bags. The quality would be determined by what the zombie would normally drop.
Click to expand...
Here we get to a topic i want to work on for a Mod later:

Zombies that only spawn during Bloodmoons ( like Demolishers ) that have special Gear.

My plan would be to take existing Zombies, create same-looking Copies (as im really just manipulating numbers) and push their strength and health and make them spawn at higher Gamestages only in Bloodmoons. To provide long lasting "fun" when youre in Endgame.

These new-Copy Zombies could drop extra loot and now we are back to your mechanic with the extra loot the longer the Horde lasts.

 
We have a proper world building tool already. I don't really get this vibe that RWG is not in a good spot right now.
Click to expand...
Granted, Haidr and many others do have a point. The terrain, the roads, the mountains (more gentle, not just Mt Everest or nothing) - the world in general just felt smoother in A16 RWG. And last but not least we had 2 more biomes, but I am aware these are not coming back.

On the other hand, apart from maybe iconic POIs, I don't see a problem with a building appearing more than once. On the contrary, I am afraid that avoiding duplicates will be limiting to world size and many people, me included, like big worlds, especially for MP.

That said, as most people on here, I have learnt to love Nitrogen. The inclusion of a comparable map editor could greatly benefit vanilla.

 
A18.2 b5 I tried that and it worked. I wasn't safe only when I put 1 block above me. But 2 blocks make it safe. But i tried that only at first few hordes.
Click to expand...
Sounds like you may have found a bug. If it happens in the current build, I hope you'll consider filing a bug report in the appropriate section.

 
C# is faster than java by a wide margin. However, not all algorithms are created equally.
Click to expand...
I never said to abandon C#, just not to have the bulk of Unity for generation and a simple front end app instead.

 
I have no interest in closing specific AI exploits, I don't use them so they don't affect me. Its the Hydra of games, cut one head off two more take its place. Faatal does all the AI, but he's mainly working on graphics for this alpha so I wouldn't expect much AI changes. Vultures can now attack cages and roofs though, so that will be interesting.
Click to expand...

I finished some changes to AI last week specifically for that case. AI now do some height evaluation in their path, so dumber ones tend to not want to go through the effort of climbing the ramp, but some will and it changes over time. Destroy area chance is now also shared/increased with the zombies around the one that fell, so a little different behavior there too.
Click to expand...
I'm very grateful that Pimps like faatal do have an interest in closing or at least transforming specific AI exploits, even when it's not his primary assignment. It'd be a major disappointment, in my humble opinion, if this game were declared finished while known exploits remain.

 
I haven't tried the forest yet, but most the guys play Ark and said the forest is a real short experience with no replay value.

Ark solo sucks. You need a few mods, the right game settings, and you need a dedi to play on, then it can be a blast. It's nearly impossible solo to beat a boss and get the end game gear. This way crops grow/dinos level when your offline and ♥♥♥♥ happens when you are offline, it feels more alive that way to me. You can see what your friends have built, etc. I don't know why but in 7 days I don't like playing with friends, maybe because there is no secrets to me and I don't like sharing the world so much, but in Ark it doesn't matter everything respawns constantly. Its more about taming that perfect high level dino for bragging rights, or teaming up to smash a cave (that is hard AF).
Click to expand...
Why I wouldn't say that the Forest has inherently low replay value, it's no 7 Days. There's a bit of variety, such as building near a cannibal patrol path (or on one, or on a crossroads) for maximum interactions from the Cannibals, but there are a few safe sections of the map. (Personally, I've never seen the point of building a base in a base defense survival game where you will experience no base defense, but to each their own.) I've already dumped nearly 3000 hours into 7 Days to Die since A17; regarding the Forest, I can see myself putting another few hundred hours into it.

Maybe that's the issue; I'm a solo act online and IRL. Haha

 
Yes this is pretty much it with The Forrest. I liked it a lot, but once you're done with following the main quest there really isn't much at all to do. I tried setting up a dedi and playing around with that, but you just get new people doing the same thing and then leaving.

The only thing I think Ark is really missing is an initial purpose other than survival. You start out and you really have no goals other than not starving to death and building shelter. It's missing that constantly impeding danger that 7 Days brings with the horde. As you stated though, it is only a couple of mods and setting changes away from being a boring grind, to a really interesting game.
Click to expand...
Yeah well, the replay value in the Forest comes from not caring about the storyline at all (or Timmy) and building in different locations, trying new things, or maybe you're just a fanatic about the game in general. But again, most people cannot do this, so to each their own.

If 7 Days to Die were in the same boat, then the only world anyone could play on would be Navezgane and instead of blood moon hordes there would just be wandering hordes and some areas that had more zombies than others. Some people could put a thousand hours into it, others only a hundred. *Shrugs*

Speaking of 7 Days though, I am loving all the additions and changes that A19 will be presenting with us so far. :)

 
Last edited by a moderator:
Yeah well, the replay value in the Forest comes from not caring about the storyline at all (or Timmy) and building in different locations, trying new things, or maybe you're just a fanatic about the game in general. But again, most people cannot do this, so to each their own....
Click to expand...
Which is why I have 130 hours instead of 30. That number isn't likely to change much for a long time though.

...If 7 Days to Die were in the same boat, then the only world anyone could play on would be Navezgane and instead of blood moon hordes there would just be wandering hordes and some areas that had more zombies than others. Some people could put a thousand hours into it, others only a hundred. *Shrugs*
Click to expand...
Hell, the only time I play Navezgane is right after a new build release for the purpose of testing. Until there's a real story there, it gets boring way too fast just because I know where everything is.

 
Hell, the only time I play Navezgane is right after a new build release for the purpose of testing. Until there's a real story there, it gets boring way too fast just because I know where everything is.
Click to expand...
Finally someone who is the same way. This is the reason why I cannot play the same seed twice, for example.

But I'll end my point by saying that what is infinitely more valuable to me than story, is purpose.

 
Last edited by a moderator:
I never said to abandon C#, just not to have the bulk of Unity for generation and a simple front end app instead.
Click to expand...
It's not unity that's causing the problem, it's the methodology used to generate the map. The built-in map generator is far more complicated and slower than NitroGen's method regardless of the engine.

 
Last edited by a moderator:
It's not unity that's causing the problem, it's the methodology used to generate the map. The built-in map generator is far more complicated and slower than NitroGen's method regardless of the engine.
Click to expand...
I have no doubt their method is more complex, but I am afraid it is mostly because they need to jump through extra hoops to make it work in and with Unity. Especially the map previewer they use I think causes extra overhead during the generation proces. Repeated world generations can cause all sorts of errors in that section to appear. If the generation was just that and the preview was a separate tool it likely would already increase the speed of generation. But it seems they have their minds made up on this one so I will leave it at this.

 
<snipped for space>
Click to expand...
They made the extra hoops themselves. Some of them by using geotiles, it takes a lot of time and memory to read and write that many texture files.

Hell, half the problem with crashes on repeated map generation is that many of the terrain calculations are sent to the graphics card which, although faster, is not really what a graphics card is meant to do. If you generate a map on linux without first previewing it you get a black screen because vulkan isn't perfect yet and your graphics card is crying.

 
The terrain generation in NitroGen is only about 10% of the code. (Just a few pages literally) The main complexities are the road-generation, smooth POI placement, city logic and handling the UI.

The main performance hog is the road generation (counting the sheep), and writing everything to those huge png files.

RWG uses a different approach to the terrain generation. Its much more expensive in memory (gfx card).

While it CAN make very nice terrain (the desert looks good) - giving the artist full freedom to create heightmap stamps, its also not a very flexible system in my opinion.

Making the biome-map (that also affects the stamped out local terrain) more dynamic will be a big push into nicer worlds.

But Im not sure if this whole approach will every be very fast, even after lots of optimization. There is just too much data (textures) to shuffle around.

The other question is: do they want to stick to the 1-seed world generation, or offer more options in configuring the worlds.

More options will give players a lot more tailored worlds. The drawback is that this seed-sharing would not work then anymore.

At a high level all we want to add now is an improved placement system so that no duplicates spawn and all pois are pushed in and not skipped over ...
Click to expand...
I always found it funny that this was a problem at all ... I still remember those radio-tower towns ;) Its really not hard to solve.

 
Last edited by a moderator:
I live in Arizona amid the cactus. About 90 minutes north is flagstaff. Mountainous terrain that gets snow covered in winter. After a nuclear winter it would be much worse I'd imagine.
Click to expand...
We had height-based biome distribution (e.g. snow-capped mountains) before. I hope we have it again! It was cool how the terrain mattered, beyond just creating areas that are annoying to travel over. It's the same with caves. I still miss them terribly, as they were arguably the most interesting element RWG ever created. I've said it before, but random world generation does for replayability/longevity what no amount of man-hours making POIs can do.

 
We have a proper world building tool already. I don't really get this vibe that RWG is not in a good spot right now.
Click to expand...
Bigger maps...8x8 is very limited if you have 4 or 5 players on the map. The old 16k maps offered a lot more area to spread out and scavenge. Basically now you pick a city and camp. The game lost the search and explore concept it always had. Its more loot a city rinse and repeat. Its not just me saying this either. Ive been playing since A5-6 with a group and everyone of them asks...when are bigger maps coming back?

you have to expand the playfield..vehicles are useless when you can drive across a map in 5 minutes. Make each biome a map or something that we can load up each time you move into a new area.

You hit the preverbial brick wall on these smaller maps.

Make it so number one...otherwise i dont see late game being anything but boring.

Coughs* Airport Poi* Coughs

 
To be fair, Ark isn't really a Voxel game either.
Click to expand...
I did not mean to imply otherwise. 7dtd is a voxel and that's why I can't run it as smooth as Ark with the same crappy settings. Despite this and given the info we already have, I believe that things will improve as much as 10% from this point until gold. There's not THAT much left to be optimized, and NPCs are a necessary evil on performance.

 
Status
Not open for further replies.
Back
Top