Alpha 20 Dev Diary

[Rabbitslovecactus]

18 minutes ago, Khalagar said:

 

Dearest Faatal,

Please spend between 70, and 450+ developer hours, valued by the company at 60 to 80+ dollars an hour, working on an obscure and very complicated feature for me. This feature has no resale value or marketing value, and will not entice a potential buyer into actually making the final purchase. In fact, my feature will appeal to no one besides customers who already paid $7 for your product 5 years ago and have put 2,000 hours into it. 

I think this investment of the most highly contested  resource at your company (developer time) will be very well spent on my obscure and barely marketable feature.

Best wishes,

Every gamer who has never worked at a big company before and has never had to answer the "And who will pay for this? What else could that developer time be spent on?" line of questioning

 

 

 

 

Dear The Fun Pimps,

 

I appreciate that you have spent over 7 years working on implementing, balancing, and programming the zombies, but I regret to inform you that I do not, in fact, like zombies. Please add an option to remove this unwanted feature from my interactions with the game. Furthermore, I also regret to inform you that I do not appreciate your addition of crafting, eating, or looting, or most of all, shooting. Please add an option for those of us who wish to run in an empty field and complete jump puzzles

 

Sincerely,

Random obscure internet forum poster #2342394

Actually there is an option to remove zombies from the game and make it just a survival game.  😁

[Blake_]

58 minutes ago, Jost Amman said:

The best way IMO is to cast a "ray tracing" (was that the name?) from spaces (air blocks) near the bandit, up to the player, and if there's a "free" block nearby that has a collision in the direction of the player that is cover. (did I say this right??) 

 

Example:

(player)          (cover 1)                     (bandit)     (cover 2)

 

In this case, both free blocks are cover, but you can have the bandit choose the nearest one by checking that there's a wall between cover 2 and player.

The HUGE problem I see is that you'd have to tell the TYPE of block cover 2 is made of and keep only the ones that won't allow projectiles through. 

 

Yeah. Basically raycasts, pathing code and tagging selected blocks+ knowing interior from exterior would solve all of that. It's a work akin to making 300 POIs, but for programmers. Maybe level designers could start to implement "cover areas"(made by a programmer of course). That would solve interiors. Exteriors would only need to tag trees, boulders and a few environment assets.

[Jost Amman]

1 hour ago, dcsobral said:

Ah, but can they path to these spaces? I mean, I could provide some really nice covers close to the entrances of my base with just the slight inconvenient that it will take a long time for someone to go from the entrance to them, all on plain sight.

Yeah, it's complicated.

Also, I'm worried if they have the processing power to calculate all that "in time" and not make it a lag fest. 

30 minutes ago, Blake_ said:

Basically raycasts

Ah! So, that's how they're called! 

[Laz Man]

1 hour ago, Gamida said:

This is just a curiosity question about what I noticed in the Sept 15 dev stream.

Who in charge is obsessed with the number 8?

I am pretty sure every building I saw that was numbered started or included the number 8. Example there was 800, 888, and 8000.

 

Is this random or were you guys ordered to use this number? 

 

 Those are random number block helpers. 

[Jost Amman]

Just now, Laz Man said:

Those are random number block helpers. 

Yes, I counted (I think) 8 of them... 

[MechanicalLens]

1 hour ago, Gamida said:

This is just a curiosity question about what I noticed in the Sept 15 dev stream.

Who in charge is obsessed with the number 8?

I am pretty sure every building I saw that was numbered started or included the number 8. Example there was 800, 888, and 8000.

 

Is this random or were you guys ordered to use this number? 

 

 

Surprised we didn't see 8008 anywhere.

[Laz Man]

1 minute ago, Jost Amman said:

Yes, I counted (I think) 8 of them... 

They are not random in the editor...but work correctly in an actual game.

[faatal]

56 minutes ago, Blake_ said:

Yeah. Basically raycasts, pathing code and tagging selected blocks+ knowing interior from exterior would solve all of that. It's a work akin to making 300 POIs, but for programmers. Maybe level designers could start to implement "cover areas"(made by a programmer of course). That would solve interiors. Exteriors would only need to tag trees, boulders and a few environment assets.

I'm not having them tag anything. The world is full of cover, you just have to scan and find it similar to how destroy area works with the added problem of you need to be on a side that covers you. Destroy area find is pretty expensive, so this probably more so. Having them duck down would help too, since that is cheap. Lay in the grass. Now I can't see you.

[Jost Amman]

42 minutes ago, Laz Man said:

They are not random in the editor...but work correctly in an actual game.

Sorry, was just a bad joke about the number 8... 

[Gamida]

4 hours ago, MechanicalLens said:

 

Surprised we didn't see 8008 anywhere.

I am not lol. I never mentioned it because I didn't know if it was in the last part of this video (that I never watched over as I saw it on the night they made it) or if it was in the first one somewhere, but I was almost positive I did see an 8008 on one building. I may have to watch them again now to see as it will drive me crazy.

4 hours ago, Laz Man said:

 Those are random number block helpers. 

ah ok so there is a method to the madness

[stallionsden]

12 hours ago, Matt115 said:

 

 

Honestly i have bad opinion about human enemies in zombies games and i will try explain why spend a lot of resources and time for them is pointless.

Okay : 1. Often human AI in games just do stupid things- i think everyone who play a lot of fps saw some stupid human enemies behaviour. In 7dtd it will be even harder - Because there is bandits VS zombie VS player - so there will be a lot of stupid sitations. One of the best example can be hospital in hl2 or days gone. In this games "human" ai just wait to be killed by zombie AI. @faatal don't blame me but Bend studio had a bigger budget , more employers and they coudn't do this good in indestructible world so it just can't work good in 7dtd. Honestly that's why we don't have human enemy in cod zombie ( Atlas Strike Team is exception but they are "invisible" for zombies so they just fight with player). Honestly i wish it could works good and the closed to achieve this is Dying light but is still far from being good (sometimes bandits ignore zombies and are focusing on player and zombie deal rly small dmg to them).

2. Balance problem - okay - in days gone most have guns. You have guns. Melee weapons are only "last stand" weapon not main for 80% of game.  In dying light most people use melee weapons and you use melee weapons most of  time. So everyone have similiar way of fight. In 7dtd you have more styles of fight - you can focus on few melee weapons, spears, few types of guns.  In diffrent games ofc you have this same sitation but simpler- in dead island  you have few melee characters , 1 throwing character and guns characters. They have diffrent "special" skills but finaly they are almost this same in playing. 7dtd is more complicated so this can be impossible to balance. in Dying light guys with guns can kill you very fast but when are this guys there is almost any zombies ( prison island). So there is a chance to be bandit  too easy or too hard. So it would need a lot of time to balance this - maybe 1 or 2 rly big fixes

 

 

@faatal i know you guys promise them in kickstarter so my advice is : just make it in bordelands style- melee bandits just behave like bordelands psycho, guns bandits just stay somewhere or just walk few steps and shoot.  And make them respawning in bandit camps - let them just sitting there similiar to bordelands 1 and that's all. I know you guys are ambitiouse and i appreciate it but a lot studios tried and failed so better not to waste time. In zombie game zombies works best- they can do stupid thing and people will forgive this because they are zombies. But reviewiers and players critize stupid human AI behaviour 

They were  part of kickstarter goals. And a promise which if not put in the game could lead to legal ramifications on tfp part for not having them in fairly certain with kickstarter rules.

Edited September 18, 2021 by stallionsden (see edit history)

[Burrfly]

Just a random short thank you to The Fun Pimps for interacting with their playerbase. Via this forum, we can ask a question about the game and almost all of the time we get an answer, which is amazing. We can even leave our own input/opinion about certain aspects to the game and knowing that some devs will read that, is just epic. I only know a very few amount of game companies that do so. I wish WildCard did this for ARK, because that game is officialy released while being less finished than 7 Days to Die that is in alpha

[Guest]

@faatal, here you wrote that you might add improved snow if you have time before the release. But why not do it after the release? After all, it is already relatively soon, and you have a lot of plans, and such trifles will immediately go to the tenth plan. And there are a lot of things that you would like to add, but do not have time before the release. So why not do it after?

[littlegreen]

@Roland @faatal

 

Guys, are there any future plans to enable super sampling for the game? Either DLSS or FFX-SR (preferably DLSS lol)

I know it is best to optimize as best as you can so that stuff like that is not needed at all, but I am struggling to display Ultra settings on anything above 1080p in high frame rate, and I happen to be running a 3080, so... super sampling maybe makes sense? 😇

 

EDIT:

OK, just to be clear - I'm not bragging about having a nice GPU, I'm asking a genuine question. Screen resolutions are ever increasing and DLSS is a very nice way to get around that, whilst at the same time keeping system requirements low - thus targeting a wider audience

Edited September 18, 2021 by littlegreen
Additional (see edit history)

[Blake_]

9 hours ago, faatal said:

I'm not having them tag anything. The world is full of cover, you just have to scan and find it similar to how destroy area works with the added problem of you need to be on a side that covers you. Destroy area find is pretty expensive, so this probably more so. Having them duck down would help too, since that is cheap. Lay in the grass. Now I can't see you.

Sweet. Non-scripted behaviour is amazing and it always creates unexpected emergent gameplay when you least expect it. It's obviously expensive, but it has to be to make them believable and smarter than an Arlene.  Thanks for the answer.

 

I've never seen non-scripted jack of all trades AI in a voxel world (cover, surround,alert,path perfectly,fluff behaviour). And I've played quite a few (for example, even a static world like Rust has scripted AI, Rising World has only basic attack, Minecraft is the closest, but units are cheap as dirt and they don't scan properly... etc). So, in a sense you are pioneering on this area like Robert is for RWG.

[meganoth]

8 minutes ago, Blake_ said:

Sweet. Non-scripted behaviour is amazing and it always creates unexpected emergent gameplay when you least expect it. It's obviously expensive, but it has to be to make them believable and smarter than an Arlene.  Thanks for the answer.

 

I've never seen non-scripted jack of all trades AI in a voxel world (cover, surround,alert,path perfectly,fluff behaviour). And I've played quite a few (for example, even a static world like Rust has scripted AI, Rising World has only basic attack, Minecraft is the closest, but units are cheap as dirt and they don't scan properly... etc). So, in a sense you are pioneering on this area like Robert is for RWG.

 

Pretty sure it will produce hundreds of special cases where the AI fails and is exploitable with it, because players can recreate any such condition. If anyone is worried about the 100 bugs left now he should be in panic about the number of bugs once bandit AI is in 😉

 

[Matt115]

1 hour ago, meganoth said:

 

Pretty sure it will produce hundreds of special cases where the AI fails and is exploitable with it, because players can recreate any such condition. If anyone is worried about the 100 bugs left now he should be in panic about the number of bugs once bandit AI is in 😉

 

that's why i think the simple idea is the best idea : just put them into bandit camps - few of them will be fighting melee some of them will be shooting and that's all. nothing complicated just speaking zombies with guns

[Fanatical_Meat]

15 hours ago, faatal said:

I was never impressed by decision trees. Technically we use decision lists, but some AI like the vulture is just a state machine. They all get the job done and are not hard to do, until you get to the gory details of what all of those decisions have to do to actually work.

 

Health < 50% run away. Cool, but to where? Taking cover is going to be one of the hardest parts of bandits, since what is cover in an arbitrary voxel world?

 

People love to gloss over the details, but the details are what take all the time and kill you in the end if you don't do them right.

 

Priorities get in the way of lots of fun. Last Friday I was thinking I could mess with the raft, but nope someone needed me to look at a light block bug. Today I was thinking maybe some raft time....nope new short grass needs to be tested, so artist can get feedback, which sounds simple except for the new mesh generation short grass will use and adding to the biomes, splitting from the tall grass code, shader changes so it blows in the wind since it is too short and lots of testing. All of that eats up time, so maybe it will be done Monday.

 

We have our A20 must fix list going now. 111 currently, 8 of which are mine, but in the beginning it tends to grow as fast as you can fix them, so fun keeps having to wait.

Someday I look forward to a water world. Lots of islands.

They will be functional. They will not be doing brain surgery on each other. Back in the spring I did a full play through of The Outer Worlds. Their enemy AI was not amazing, but competent. If we got to the same level I'd be happy. Their AI does not even find cover, so maybe we can be a bit better.

Regarding AI that is what I expect.

I find it cool that the AI in rust does run for cover however I understand that is not feasible for 7D2D. Rust enemy scientists hide between cars or barrels. The whole event spawns as an entity so I am sure programming it is much easier.

I would be satisfied if bandits kneeled down occasionally when shot at or when shooting back. Not all the time but some times.

I would be happy if they charged you in a semi zig-zag pattern or occasionally one in the group would try to flank you.

As I said earlier I would be happy with bandits hordes attacking the base when some timer goes off or x minutes after a gong or explosion or whatever notice. I would be happy if they were semi smart zombies like some stay away and shoot (this would be a good use of scopes & rifles) and others got close to shoot and some threw dynamite or grenades at the base/walls.

I would be impressed if bandits attacked players when they were out exploring like the player(s) need to get into a POI to fight a group of bandits off.

Edited September 18, 2021 by Fanatical_Meat (see edit history)

[MechanicalLens]

47 minutes ago, Fanatical_Meat said:

I would be impressed if bandits attacked players when they were out exploring like the player(s) need to get into a POI to fight a group of bandits off.

 

Please, I don't need a bandit headshotting me with a sniper rifle from 90 meters away. 😛

[Matt115]

7 minutes ago, MechanicalLens said:

 

Please, I don't need a bandit headshotting me with a sniper rifle from 90 meters away. 😛

yep. same problem, that's why i'm so pessemistic about that 😕 well i have experience in some games when random guy could kill you using smg from 500M but you coudn't hit him with rifle

[Ricowan]

16 hours ago, faatal said:

Someday I look forward to a water world. Lots of islands.

This would be AWESOME!

[Robbing Zombie]

1 hour ago, Ricowan said:

This would be AWESOME!

Also a jungle / swamp option for random gen maps would be nice. That's probably more of a mod thing, but would be a nice add.

[Fanatical_Meat]

2 hours ago, MechanicalLens said:

 

Please, I don't need a bandit headshotting me with a sniper rifle from 90 meters away. 😛

To be clear I mean further away than a puking zombie but still easily seen on the screen. 
20-30 blocks is my guess. Far enough that a shotgun or smg or pistol won’t work but not so far you cannot see them.

Rifle and scope for the players, unseen bandits randomly sniping us would not be fun.

Just make the bandits guns do less damage.

[unholyjoe]

2 hours ago, MechanicalLens said:

 

Please, I don't need a bandit headshotting me with a sniper rifle from 90 meters away. 😛

but you wouldnt mind it if it was you doing the sniping at 90m away at the bandit!

 

equal rights for bandits and zombies

[jeromeN7]

8 hours ago, littlegreen said:

@Roland @faatal

 

Guys, are there any future plans to enable super sampling for the game? Either DLSS or FFX-SR (preferably DLSS lol)

I know it is best to optimize as best as you can so that stuff like that is not needed at all, but I am struggling to display Ultra settings on anything above 1080p in high frame rate, and I happen to be running a 3080, so... super sampling maybe makes sense? 😇

Yes please, for now I wish they'd at least get exclusive fullscreen working because without that I can't even use any down sampling resolutions without first setting my desktop resolution the desired DSR resolution. Even with TAA, a little bit of down sampling helps reduce blurring.