Alpha 18 Dev Diary!!

[SrslyGTFO]

Why ask for 19 when 18 isn't done?

 

Well, you did mention redoing the weather for A19 while in an A18 thread.

 

That makes it sound like A18's mechanics are pretty well laid out, and you're already planning A19. No matter, I'll be thrilled for even 17.3 at this point. I must say, A18 is really sounding like a stellar build. I can't imagine what you have planned for A19 -- I figured Gold was coming after A18 with all the progress being made. Have a great weekend!

[Roland]

If I may, i think the disconnect between the company and the community is the assumption we wanted mods in the form we received them.

 

There was no assumption that the community wanted or didn't want the mods in the form you received them. They didn't have a fan generated checklist of wishes they were trying to fulfill in their planning meeting but somehow misunderstood. They put things into place that they wanted and hoped that the community would like it too. I really don't get the whole "We never asked for this" mentality among some in the community as if the developers shouldn't design anything before getting the fan stamp of approval.

 

People were asking for mods but things like silencers/scopes/magazines/nails etc. NOT mods in PLACE of quality stats. It seems like one part of the community feedback was received but the other part ignored. Mods are nice, as an addition TO quality crafting stats etc. NOT in place of them.

 

I think it was more like TFP had plans to do mods the way they did them and some aspects happened to align with what people were asking for and other aspects did not. But it wasn't a situation of TFP designing mods to fulfill the wishes of the community but then ignoring a part of it or misunderstanding what the players wanted. I think you are assuming that community wishes have a greater impact on development choices than they do. Community wishes and reactions have a strong impact on how TFP adjusts their designs and tunes them but it has almost no impact on the core design decisions in the first place. The choices they are making are their choices for their game the way they want it to be but they are willing to dial it back a bit in response to community feedback as we have seen each and every Alpha release.

 

Think of how many more combinations we could have had if we had both quality and mods. You reduce the silly mods like paint adding damage to everything, add back the quality stats and have mods give small bonuses or functionality. Hell the first time MM showed off the mods EVERYONE was wondering what the heck was up with Quality 1-6. And most expressed concern over it.

 

I can't dispute this. I believe you are correct when you take 600 steps of quality that are defined by various gun parts that also have 600 steps of quality and that need to be assembled and then add in mods-- the combinations are many for sure. My only point is that for whatever reason they chose not to go that route it wasn't because they assumed anything about the community wanting or not wanting them to do it before moving forward with their plans.

[Crater Creator]

Added Block VehicleHitScale for vehicle collisions, removed heavy hit based on mesh tag and setup blocks (road debris and small cactus x999, signs x8, shopping cart x4).

 

It would be great if there were a property like movement_factor or BuffsWhenWalkedOn that applied to vehicles, so that e.g. driving on grass incurred a penalty versus driving on asphalt, and driving through other 'soft' things like a cotton plant or hay bale could slow down the vehicle without necessarily applying damage.

 

Fixed SDTD-8948 Wandering horde has too many duplicates side by side (GetRandomFromGroup takes a last spawned id and tries 3 times for a different id. Applies to all spawners except biome).

 

Fantastic! I'll keep putting in a good word for procedurally generated zombie appearances, though.

[JaxTeller718]

Yes and some refinement as I see things that could be better. Here are my interesting changes for the last 2 weeks:

 

Fixed SDTD-8497 Vehicles are a base wrecking machine (vehicles do half damage to land claimed blocks and make the sound, use driver's block damage scale).

Fixed SDTD-8480 request for having bikes (2 wheelers) not do damage to blocks when they fall over after dismounting (increased minimum velocity required to damage and reduced damage amount).

Fixed vehicle doing more damage on collision separation, block check could fail and block hits tended to undo the motion and repeat the damage.

Added vehicle hopForce forward offset, so front tire lifts more on 2 wheelers and motorcycle does wheelies.

Adjusted mass of 2 wheelers.

Increased vehicle self damage.

Optimized HUD target/crosshair update and reduced garbage gen.

Fixed SDTD-8839 smoking minibike when placed/spawned.

Added Block VehicleHitScale for vehicle collisions, removed heavy hit based on mesh tag and setup blocks (road debris and small cactus x999, signs x8, shopping cart x4).

Fixed SDTD-9349 mushrooms can be damaged when run over by minibike (set crops to VehicleHitScale .1).

Fixed SDTD-8948 Wandering horde has too many duplicates side by side (GetRandomFromGroup takes a last spawned id and tries 3 times for a different id. Applies to all spawners except biome).

Optimized EntityAlive updateCurrentBlockPosAndValue.

Fixed UI ItemStack dirty flags (always updating) and optimized UI array access.

 

Fixed fleeing animals will get stuck in corners or run into walls (Bigger random angle. Shorter paths. A large AIPathCostScale now makes paths that don't go through blocks. Run away short paths switch to random escape direction).

Refactored AI run away logic and added timeout.

Fixed CalcPositionInDirection not height testing correct position.

Improved AI run away values.

Adjusted animal move/panic/turn speeds.

Merged deer mesh and prefab and centered near front feet.

Increased timid animal vision angle.

Added path start position motion estimation.

Increased path point reached distance when side stepping.

Added AI slows down near end of path and won't turn when close.

Improved AI blocked distance, side stepping and direct movement.

Changed AI returning home to use aggro speed (at 80%).

Fixed entity RotateToGround jittering on flat terrain, rewrote math and added springiness.

Fixed SDTD-9255 gyrocopter floats in air if the pilot gets disconnected/logs out (vehicles will fall once their chunk loads).

 

Read every single line of this and each one made me smile wider and wider. This is a great list of ai changes, especially in the areas of animals. Looking forward to it.

 

- - - Updated - - -

 

There was no assumption that the community wanted or didn't want the mods in the form you received them. They didn't have a fan generated checklist of wishes they were trying to fulfill in their planning meeting but somehow misunderstood. They put things into place that they wanted and hoped that the community would like it too. I really don't get the whole "We never asked for this" mentality among some in the community as if the developers shouldn't design anything before getting the fan stamp of approval.

 

 

 

I think it was more like TFP had plans to do mods the way they did them and some aspects happened to align with what people were asking for and other aspects did not. But it wasn't a situation of TFP designing mods to fulfill the wishes of the community but then ignoring a part of it or misunderstanding what the players wanted. I think you are assuming that community wishes have a greater impact on development choices than they do. Community wishes and reactions have a strong impact on how TFP adjusts their designs and tunes them but it has almost no impact on the core design decisions in the first place. The choices they are making are their choices for their game the way they want it to be but they are willing to dial it back a bit in response to community feedback as we have seen each and every Alpha release.

 

 

 

I can't dispute this. I believe you are correct when you take 600 steps of quality that are defined by various gun parts that also have 600 steps of quality and that need to be assembled and then add in mods-- the combinations are many for sure. My only point is that for whatever reason they chose not to go that route it wasn't because they assumed anything about the community wanting or not wanting them to do it before moving forward with their plans.

 

Not to worry, I was not trying to drum up some community uproar over it. Its way more of an "imagine what could have been" scenario where we had hundreds of different combinations. I DO like the news of mod slots from crafts being limited and looted weapons having more slots.

 

My biggest gripe over 17 overall was the reduction of a need to go out and interact. Dungeons are awesome and still are I feel. I just want a reason to do them besides a chest with a pistol and 15 bullets. And this looks like it will hit that area nicely which I am grateful for.

 

18 is looking very promising. With the books etc all coming I suspect a lot of others feel the same way.

[Doctor3D]

I've learned not to get too excited by what promises or ideas are made...I'd rather not be let down.

 

 

Sorry...just the truth...

 

A18 doesn't look any more 'promising' than A17 did. Hard to trust that kind of statement or thought.

 

All I want to hear about is bandits and Ai - I can just mod the rest...About the only thing that's gonna make this game last beyond my Alpha trial each time is a clear, fun challenge and I need some bandits shooting at me to do it at least. I'll mod them how I need to as far as accuracy and aggression but I need some proper ones. I don't care about anything else frankly. I need the game to be hard again.

 

TLDR.. I don't care about anything else for A18 except bandits..that's what I expect for it. Small little tweaks to various systems..who cares..we do that ourselves..I want to be shot at..I want to feel threatened.

[Gazz]

Any chance we'll see a slightly more robust bleeding mechanic? I saw a suggestion over in Reddit regarding three tiers of bleeding debuffs. It would add white and orange tier bleeding debuffs, with different levels of health loss, of course.

 

With proper armor, perks or whatever, you could mitigate the bleed debuff somewhat. It should be extremely rare or impossible to get a critical (red) bleed debuff in full modded steel armor against a vulture. Maybe just a mild bleed.

Yes, that's how it currently works.

Bleeding can stack up to 3 times. Armor mitigates the chance of getting a bleed debuff and therefore it stacking higher.

[beerfly]

Whatsup, Doc ?

 

Bandits are on their way.

If some "stuff" is not properly done/developed before them, forget the bandits being proper, too. Thats it. Cheers.

[The Gronk]

Sounds over engineered. How about it has an increased chance to rain every day that it doesn't? Lets keep it simple so weather doesn't get cut completely.

 

It's a little more complicated than a global precipitation variable but not by much and the frame-rate impact would be very low as clouds don't tend to move that quickly.

 

I'd say the global precipitation variable is too simplified. Just because you're on the final stretch is no reason to cut corners. There seems to be a tendency recently to grind down details instead of polishing them.

[FileMachete]

Question about the mod 'Rad Remover'.

 

Looking at the item_modifiers.xml it looks like the "weapon" tag is what determines if the Rad Remover can be installed.

But the machete has that property tag and I couldn't install a RR.

The fireaxe doesn't have the tag and it won't accept a RR, as expected from the xmls.

Both the meleeClubWood & meleeClubIron have the tag and a check of the iron club showed it was installable.

 

Thinking ahead to the more focused toon builds, wondering if all melee options will allow the Rad Remover, or something similar? Or will it be restricted to certain weapons?

 

Just curious if say there may be things like 'defensive tank' & 'offensive tank'; where a 'defensive' build might not have something like the Rad Remover, but offer better damage tanking, while an 'offensive' build might offer the RR for certain melee weapons, so more damage output, at the trade off of less toughness?

In mmorpg terms a sword & board vs. a large weapon build.

[AtomicUs5000]

Just because you're on the final stretch is no reason to cut corners. There seems to be a tendency recently to grind down details instead of polishing them.

 

I feel like they started with a very unique cookie, and now that it's cooking, they are trimming the edges to make the same cookie they would have had using the same old cutter.

[madmole]

That makes me a bit sad. Ever play Don't Starve (Together)? Seasons can add quite an interesting dynamic to a game. It wouldn't be harsh on noobs either. Spring and Fall feature mild days and slightly chilly nights, summer features hot days and in the winter, it snows. Hell if done right you could eliminate the snow biome and just have forest, wasteland, and desert biomes. Desert would be hot all the time. Wasteland? Anything you want, extreme thunderstorms? Acid Rain? Who knows. The forest would be the only biome to undergo seasonal changes.

 

I personally find the latest changes to temperature/climate/weather a bit boring. I preferred Alpha 16's weather. However, I'm one of those that enjoys the current alpha.

 

Sure if done right it could be cool, but Don't Starve's whole game is designed around food and seasons. We have 50000 more features than them that need polished up to go gold so adding new features isn't a priority for us. \

[madmole]

Stabby McStabbins sounds like the kind of character you'd hear mentioned in the game's flavor text.

 

I'd be amused if there was one of those "found note" quests with the late Stabby McStabbins telling you where you can find his secret scroll of stabby skills.

 

One of the guys on the Ark server I played on had hilarious dino names.

[madmole]

This was suggested before, but here it is again... one simpler idea would be for precipitation to raise the water table, and it takes time to come back down. Then weather would have both a short-term and longer-term effect, without having to implement seasons. Flooding and droughts would impact survival more than clothing/temperature alone, and it would specifically introduce a dynamic hazard to underground living.

 

We're not flooding thousands of voxels with this tech. Knocking down a small building is hazardous to the framerate I can only imagine entire water tables moving and huge floods with voxels putting the minimum pc requirements to 2150 computers.

[Pegasus]

Then you go another 10 feet and its raining again. My point is say you have 3 or maybe even 4 intersecting biomes its kind of immersion breaking to change spectrums and its suddenly not raining by leaving that biome. I'd rather the rain make it colder and there is no way out of it. The whole server should be talking about the ♥♥♥♥ty weather not one guy in one biome. From a survival standpoint if rain makes it colder then its better game play if there are no real known ways to escape that.

 

Before you do that ... can you please fix windows and doors and hatches so that rain and snow don't come in through your ceiling or windows? I mean, what's the point of having a fully upgraded vault hatch, if it leaks like sieve? And even though the windows SAY they keep out the rain ... they don't. It snows inside your house. It snows inside your bunker duge 50 meters below ground with vault hatches on top. It's ridiculous. Just as it's ridiculous to be 50 meters below ground, and have the light change because the sun is up or down. So before you make it so I get cold when I get wet (great idea, miss it) ... please make it so I can get not wet by being inside my home.

[madmole]

If I may, i think the disconnect between the company and the community is the assumption we wanted mods in the form we received them.

 

People were asking for mods but things like silencers/scopes/magazines/nails etc. NOT mods in PLACE of quality stats. It seems like one part of the community feedback was received but the other part ignored. Mods are nice, as an addition TO quality crafting stats etc. NOT in place of them.

 

Think of how many more combinations we could have had if we had both quality and mods. You reduce the silly mods like paint adding damage to everything, add back the quality stats and have mods give small bonuses or functionality. Hell the first time MM showed off the mods EVERYONE was wondering what the heck was up with Quality 1-6. And most expressed concern over it.

 

I think most people would agree the following are true statements:

1.) Finding gun parts and not whole guns was stupid. It was merely a gamey mechanic to slow players down from getting great firearms too soon.

2.) Teaching new players how to assemble, repair guns or parts was confusing.

3.) Mods that don't do damage are largely uninteresting.

4.) The current system at its core is fine, its way better than repairing individual parts (clunky AF)

5.) It just needs some sort of random ranges and it could be interesting

 

Given the context of those mostly true statements you can see how we fell on the design we did. Fallout 4 has all sorts of gun mods and the only ones worth a crap are the ones that increase damage and the scopes. I just don't see how to make enough interesting mods without getting too gamey "adds extra damage to super mutants", again, its even damage and its totally meh. All you go for in Fallout is the incendiary legendary weapons and its game over. At the end of the day damage is the main thing players are after and we didn't want 5 damage mods and 100 meh mods.

[Pegasus]

Probably not that hard but lets leave some work for the modders to do. Thats kind of a hardcore personal tastes mod, that I would install myself, but average guys need simple farming.

 

Oh, this reminds me ... any chance of getting grow lights? They have to be hooked to power and such, but you could grow your farms underground....

[madmole]

My biggest gripe over 17 overall was the reduction of a need to go out and interact. Dungeons are awesome and still are I feel. I just want a reason to do them besides a chest with a pistol and 15 bullets. And this looks like it will hit that area nicely which I am grateful for.

 

18 is looking very promising. With the books etc all coming I suspect a lot of others feel the same way.

Yes I agree, which is why all the top tier weapons will be find, buy or schematic craft only, and I added 100+ books with collector perks (more reason to find them all).

 

I'm not 100% satisfied with guns, but I don't think we have time to change them for A18, but hopefully one small overhaul with random ranges will spice things up and the introduction of legendary items.

[madmole]

I've learned not to get too excited by what promises or ideas are made...I'd rather not be let down.

 

 

Sorry...just the truth...

 

A18 doesn't look any more 'promising' than A17 did. Hard to trust that kind of statement or thought.

 

All I want to hear about is bandits and Ai - I can just mod the rest...About the only thing that's gonna make this game last beyond my Alpha trial each time is a clear, fun challenge and I need some bandits shooting at me to do it at least. I'll mod them how I need to as far as accuracy and aggression but I need some proper ones. I don't care about anything else frankly. I need the game to be hard again.

 

TLDR.. I don't care about anything else for A18 except bandits..that's what I expect for it. Small little tweaks to various systems..who cares..we do that ourselves..I want to be shot at..I want to feel threatened.

 

Good thing you won't be disappointed, bandits are not coming until after A18.

[madmole]

It's a little more complicated than a global precipitation variable but not by much and the frame-rate impact would be very low as clouds don't tend to move that quickly.

 

I'd say the global precipitation variable is too simplified. Just because you're on the final stretch is no reason to cut corners. There seems to be a tendency recently to grind down details instead of polishing them.

 

LOL, no we're being sensible professional designers. Over design is a noob designer trap. More complex means more bugs, harder to implement, and harder for players to understand. The magic under the hood is of no concern, just know that it might rain and thats good enough.

 

Sometimes grinding down details is the right call. Too many moving parts makes the game harder to balance, and players are wondering why they are freezing but there's 10000 variables at play. It was overdesigned in pre 17 and broken in 17. We'll land on something more simple that works and doesn't break immersion.

[Lonestarcanuck]

Yes I agree, which is why all the top tier weapons will be find, buy or schematic craft only, and I added 100+ books with collector perks (more reason to find them all).

 

I'm not 100% satisfied with guns, but I don't think we have time to change them for A18, but hopefully one small overhaul with random ranges will spice things up and the introduction of legendary items.

 

Collector Perks? I like it, I always enjoyed doing Archeology quest in WOW to get the special Perks. so a reason to find sets of books/schematics in order to earn a rank/perk/status would be cool.

 

I still think there is room for level specific items like clothing that shows your a level 100 with X kills = zombie skull on a stick mod for a club or something.

[Laz Man]

Would be funny if the game gave call outs too;

 

"Roland was crushed by "The Peoples Hammer"

"Roland was cut to ribbons by "McStabby"

"Roland had his melon popped by "Rusty Trombone"

 

+1 but instead...

 

"Roland was killed by Gazz with The Lubricator"

[The Gronk]

<snipped for space>

 

I'd say you were choosing quick over quality but that's a judgement call.

 

The complexity and attention to detail is what first drew many of us to the game in the first place.

 

Why bother having smooth terrain? It's been a constant source of long-standing bugs, is complicated, and has caused many issues with frame rate. You do it because you know in the end it's innovation that sets the bar higher for everyone else to follow.

 

It's a sad day to see someone constrained by boundaries they could easily brush aside.

[Lonestarcanuck]

+1 but instead...

 

"Roland was killed by Gazz with The Lubricator"

 

there was a mod called "Death Announce Mod" or something like that in Rust a few years back that did just that. some something like "Quicken hit Roland in the head/hand/arm/foot/♥♥♥♥/ass/leg/ankle/foot/etc. with a spear from 20 meters causing his death - not good" as you can see with the many body parts you could get some fun combos with tools+body part.

[Tin]

LOL, no we're being sensible professional designers. Over design is a noob designer trap. More complex means more bugs, harder to implement, and harder for players to understand. The magic under the hood is of no concern, just know that it might rain and thats good enough.

 

Sometimes grinding down details is the right call. Too many moving parts makes the game harder to balance, and players are wondering why they are freezing but there's 10000 variables at play. It was overdesigned in pre 17 and broken in 17. We'll land on something more simple that works and doesn't break immersion.

 

heh, I just hope you don't forget to also add in a 'check' to the system that forces it to stop raining after 'X' amount of time, along with having a 'check' that makes it rain after 'X' amount of time.

 

One of the downfalls of the old system was, rain (along with the constant fog that came with it) got stuck every so often and just stayed, till you exited the game and restarted it (or you loaded the game onto a server, and ran it that way).

 

The old weather system wasn't terrible, it just needed to have a global effect system as well, instead of being fully biome driven.. well imo.

If it rained in the forest, it needed to rain/snow in the other biomes at varying levels.

 

Another area that needed adjusting (for me at least):

Was/Is how you are using the different spectrum filters for every biome, which was the cause of the flickering when you had dithering between them.

The dithering was -> immersive <- because then you didn't feel like you just crossed over an invisible barrier and got transported into a new area. Instead the dithering allowed you to ease into the new biome.

 

Solution:

Use the different spectrum filters like you use the blood moon spectrum (which gets applied globally), and use the different spectrum(s) every couple/few days. Use them globally instead of just per biome, and keep the dithering between biomes.

[sillls]

Then you go another 10 feet and its raining again. My point is say you have 3 or maybe even 4 intersecting biomes its kind of immersion breaking to change spectrums and its suddenly not raining by leaving that biome. I'd rather the rain make it colder and there is no way out of it. The whole server should be talking about the ♥♥♥♥ty weather not one guy in one biome. From a survival standpoint if rain makes it colder then its better game play if there are no real known ways to escape that.

 

So are you planning to make it so the whole map is having the same weather then?