PC v1.x Developer Diary

[beerfly]

@Laz Man You should do a Jar Factory POI. Don`t just smile at that, we are waiting

 

[FramFramson]

@Laz Man You should do a Jar Factory POI. Don`t just smile at that, we are waiting

lmao, this is perfect.  Do it do it.

I understand that multithreading is not worth waiting for? I understand that this is a huge job, but without it, the game is unlikely to ever work well. Of course, the game with multithreading will remain difficult for computers, because this is not some kind of miracle. But still, without this, we are unlikely to ever see normal performance in cities or with a large number of zombies.

By the way, could you tell us a little more about the work and the difficulties that need to be overcome to implement multithreading? Even if it never will be. It's just that in the view of many players, this simply requires changing a couple of lines of code that "lazy developers" can't get to in any way. Although this is not the case, because in that case you would have done it a long time ago. The same applies to bandits — few people realize that this is a very large amount of work, especially for a small team. Thank you for putting up with us)

IIRC, significant multithreading improvements are unlikely simply because Unity itself has problems with multithread utilization. It IS a fairly old engine after all.

 

[SylenThunder]

IIRC, significant multithreading improvements are unlikely simply because Unity itself has problems with multithread utilization. It IS a fairly old engine after all.

Except that it isn't actually that old. The version of Unity the game client will be running on will very likely be less than a year old. Add to that the fact that Unity has been working to improve multithreading operations in the last couple of years, and you could see quite a bit of optimization in that regard. TFP isn't pulling a WildCard stunt and remaining on an obsolete version of the game engine because "it's too hard". Wildcard didn't even have to update versions from UE4 to UE5. It was just a minor update to UE4 that ocurred while it was in EA. If you look, every major release of 7 Days is on a new Unity client version. 

 

[Roland]

Except that it isn't actually that old. The version of Unity the game client will be running on will very likely be less than a year old. Add to that the fact that Unity has been working to improve multithreading operations in the last couple of years, and you could see quite a bit of optimization in that regard. TFP isn't pulling a WildCard stunt and remaining on an obsolete version of the game engine because "it's too hard". Wildcard didn't even have to update versions from UE4 to UE5. It was just a minor update to UE4 that ocurred while it was in EA. If you look, every major release of 7 Days is on a new Unity client version. 

of course they aren’t likely to update to any new Unity version which would push them into the new pricing plan.  Unless….maybe THATS the reason for the price increase….

 

[User]

of course they aren’t likely to update to any new Unity version which would push them into the new pricing plan.  Unless….maybe THATS the reason for the price increase….

Oh, did unity end up keeping that new pricing plan that they got a lot of flak for a while back? I don't remember.

Edit: Does anyone know if that cowboy hat has been seen in a22? The one that was shown in the character animation preview a while back. I almost forgot about it, but its a pretty dope hat so I hope that it is.

 

[Roland]

Oh, did unity end up keeping that new pricing plan that they got a lot of flak for a while back? I don't remember.

Edit: Does anyone know if that cowboy hat has been seen in a22? The one that was shown in the character animation preview a while back. I almost forgot about it, but its a pretty dope hat so I hope that it is.

That cowboy hat is one of the new paid DLC costumes. 
 

KIDDING!!!

 

[HB_H4wk]

Did someone say Zombie Deer?

 

[Roland]

Did someone say Zombie Deer?

Yes but you must never look at a deer unless there is a vulture lined up perfectly between you. 

 

[Arez]

I know that through Telltale there was Michonne from the Walking Dead as a playable character on the console version.

Are there any plans to include 3rd party IPs as DLCs going forward?  

I can suggest a couple.

The George Romero Pack. Various POIs, zombies and survivors from the George Romero movies.

The Return of the Living Dead Pack. The medical supplies building with the mortuary and graveyard as a high tier POI. As well as zombies and survivors.

Yes, I know these can be modded in, and some already have been in the past. But there is one group of people that would not be able to partake in such mods....console players.

I know there are enough people that would go absolutely bananas if this stuff was added. 

Think about it. I'm not too sure how things work, but even if TFP takes an extremely small cut of the DLC's profits, I would be SHOCKED if these DLCs wouldn't get a substantial amount of people to buy the game.

 

[FramFramson]

Oh, did unity end up keeping that new pricing plan that they got a lot of flak for a while back? I don't remember.

Edit: Does anyone know if that cowboy hat has been seen in a22? The one that was shown in the character animation preview a while back. I almost forgot about it, but its a pretty dope hat so I hope that it is.

No, Unity completely backed down and also fired the CEO (which puts him 0 for 2 on his attempts to do things like this at various companies).

Yes but you must never look at a deer unless there is a vulture lined up perfectly between you. 

Underappreciated post, lmao.

 

[Ricowan]

Here's the thing with multithreading:
Threads are independent, they are NOT aware of what each other is doing.  If you spin off a thread to work on one thing, but main thread changes that thing, then the first thread is now invalid.  So you have to add in specialized thread managers, or just live with the fact that you will occasionally waste processor time by throwing away incomplete thread work.

Once a new thread has completed it's work, that work has to be re-integrated into the main thread loop somehow.  In many cases that will add more performance hit than the new thread saved.

Multithreading is great for splitting up a single task that works on data that doesn't change.  This is not a common scenario in a fast paced multiplayer game where the terrain can be changed by players and NPCs at any time.

 

[Riamus]

Here's the thing with multithreading:
Threads are independent, they are NOT aware of what each other is doing.  If you spin off a thread to work on one thing, but main thread changes that thing, then the first thread is now invalid.  So you have to add in specialized thread managers, or just live with the fact that you will occasionally waste processor time by throwing away incomplete thread work.

Once a new thread has completed it's work, that work has to be re-integrated into the main thread loop somehow.  In many cases that will add more performance hit than the new thread saved.

Multithreading is great for splitting up a single task that works on data that doesn't change.  This is not a common scenario in a fast paced multiplayer game where the terrain can be changed by players and NPCs at any time.

That's a good simplified explanation of multithreading.  I like it.  People often think that multithreading is some easy thing to do that works in every situation and automatically speeds things up.  That just isn't the case.  And even games that have multithreading usually only have it for very specific things and not for the game as a whole.  In a game, you often have many different things all working in an interconnected way and trying to pull a piece out and run it separately and then intertwine it back in and still get some kind of performance increase rather than a loss isn't always possible.  It works best when you have some part that is entirely, or at least almost entirely, separate from everything else that can be done in its own thread without affecting other things.  There are places where this is an option but there are many more places where it isn't a good option.

A very simple and basic example that might help people to visualize the issue...

Let's say AI is on one thread and building/destruction of blocks are on a different thread.  The AI sees that it can walk a zombie forward without anything in the way.  Someone builds a wall at the same time that would prevent the zombie from going forward.  The AI thread doesn't know about that wall because the two processes are in different threads and so still tells the zombie to go forward.

Now, there are ways to integrate threads and provide some interaction/updates to them, but the more things you do to tie the threads together, the less benefit you can get from them being separate threads.  Also, to be clear, that example is NOT how they are doing things and isn't what I'd suggest as a good way to do things.  It is just a very simple and basic example about why something would be a challenge.  So you don't need to tell me that it's not how someone would do multithreading because I know it isn't. 

 

[User]

Here's the thing with multithreading:
Threads are independent, they are NOT aware of what each other is doing.  If you spin off a thread to work on one thing, but main thread changes that thing, then the first thread is now invalid.  So you have to add in specialized thread managers, or just live with the fact that you will occasionally waste processor time by throwing away incomplete thread work.

Once a new thread has completed it's work, that work has to be re-integrated into the main thread loop somehow.  In many cases that will add more performance hit than the new thread saved.

Multithreading is great for splitting up a single task that works on data that doesn't change.  This is not a common scenario in a fast paced multiplayer game where the terrain can be changed by players and NPCs at any time.

Huh. The more you know. Lots of smart people on this forum

 

[FranticDan]

Just a random question, does anyone on the dev team know for the next major update if the head hitbox for the tourist, cop and mama have been improved?
Currently, you have to hit these zombies above the nose for it to count as a headshot. 

EDIT:
Additionally, in the current version and for quite some time now, for a few frames while a zombie is picking themselves up off the ground, their ENTIRE hitbox disappears. Has this been fixed in the next major update?

 

[FramFramson]

So here's a question which probably won't get an answer right now, but I'll ask anyway. 

In the prefab editor, there are a number of very interesting, yet incomplete skyscrapers. A bunch of them, including an absolutely glorious 25-story tall monster under "aaa_skyscrapers" and a rather civic-minded/government-looking one under "skyscraper_05".

I know we're getting at least a couple more large buildings in, but might we hope for the completed versions of any of these specific ones in A22/1.0?

And is there a plan for that 25-story behemoth? Because hooooolllly sheeeeeeeet.

 

[Laz Man]

So here's a question which probably won't get an answer right now, but I'll ask anyway. 

In the prefab editor, there are a number of very interesting, yet incomplete skyscrapers. A bunch of them, including an absolutely glorious 25-story tall monster under "aaa_skyscrapers" and a rather civic-minded/government-looking one under "skyscraper_05".

I know we're getting at least a couple more large buildings in, but might we hope for the completed versions of any of these specific ones in A22/1.0?

And is there a plan for that 25-story behemoth? Because hooooolllly sheeeeeeeet.

Some of those are proof of concepts and/or facades for possible future content.  They are not all guaranteed to go into production.  Dishong is still currently the tallest skyscraper in A22 / 1.0.

We are still focused on downtown remnant POIs currently, but it is still possible for extra unplanned content to squeeze in if we have time so there is still hope.

Out of curiosity, what new POI would people like to see in A22 / 1.0 if you could only pick one?  Sorry, no airports or yachts...

 

[faatal]

All my interest in A22 just evaporated.  Not feeling any interest right now in the next release.

That is too bad. I can't wait to play it myself. All this work being wrapped up, so I finally get to enjoy playing it for real.

 

[faatal]

I suppose. It's just I'm used to game devs playing fast and loose with terms like alpha, beta, and live, that I don't put much faith in game versioning. Of course in other sectors there are clearer requirements for versioning software, especially enterprise software, but there's too much salesmanship and too many semi-amateurs in the games sector, and no outside body or customers with fixed expectations. Some players would have been happy to see A16 called 1.0, others like yourself feel it's not done yet, and neither opinion is wrong per se.

I just focus on the content and functionality. Does the game work as advertised? How likely are promised future updates? What's the studio's track record, etc. Versioning doesn't reliably correspond with any of that from game studio to game studio, at least not in my experience.

Yeah, alpha/beta/gold/whatever, early access and version numbers tend to be arbitrary these days. There is no industry standard. Games release in all manner of states and have a large variety of post launch development.

I actually like how No Man's Sky did their versioning with code names and 1.x, 2.x, ect. They may of had a rough start, but I certainly got my money's worth from that game with all the free content they have released over the years. I feel 7dtd dev is similar to theirs, except they never labeled it early access and simply chose to update the game several times a year for many years. I actually get turned off when I'm browsing new games and see "Early Access" as I never know what state it is in and tend not to buy them, so I'm glad we are removing that from 7dtd.

faatal also posted on this forum over a year ago some of the AI work specifically for bandits he was working on such as retreating, taking cover, and flanking. 
 

In addition, perceptive people will see how the outfits and 3rd person view animations for characters are also groundwork for adding bandits. 
 

We’ve also seen bandit models posted on social media. 
 

it’s pretty silly to claim that they haven’t been working on bandits or that they are only going to start work after 1.0. 
 

Exactly. 

My memory might not be the greatest (ask my wife), but I certainly remember working on bandits. We definitely laughed when we saw them spawn out in the wild and fight a zombie or die and get eaten. Jumped when one was suddenly beating you with a club or shooting you. In the end, they needed a lot more work, so we put that on hold for higher priority tasks of which there are many. I look forward to resuming the work and then inevitably adding the settings to make them easier or turn them off after we get sick of being murdered by them.

 

[faatal]

What do you mean about "event"?

Like Radiated zone to avoid? Massive zombie groups in cities? Attack/raid of zombified wildlife/bandits?

How will it be displayed? What are TFP's thoughts here?

It's pretty interesting to have more info cuz it looks like a major improvement for the late game

I don't have all the details, but my initial desire for an event is a mini blood moon, that can happen any day, with some warning to get back to your base. They could also be bandits or a mega zombie. Vulture swarm. Bandits lobbing mortars at your base that you have to go out and stop. A tornado or earthquake would be cool, but you can't fight that . Land shark burrowing up into you base. Sand worms. Trader's wife taken hostage, so you have to get her back or pay ransom. Bandit duel where you each get one shot. Bandit drag race. Timed race to get somewhere. Bandits kidnap you and wake up in random location. I'm sure I can make some more up.

 

[beerfly]

All of these. I'll probably double my gameplay time if events like these are added.