Beizs

Engine development takes longer, normally. Much longer. I have seen a lot of people complain about how slow DayZ is being developed, in their eyes, and I think that, honestly, the opposite is true. Most original games take two or three years to build, even using a completed engine and just adding on to it with mechanics, asset creation and similar tasks. This means that the game development and release alone takes two years or so for most games. DayZ is not most games, firstly. The servers are smaller than most other MMO's, but DayZ is absolutely comparable, development wise, to other MMO(rpg)'s. MMO's tend to take even longer to develop - more like four or five years at a minimum. But the DayZ dev team isn't even just developing an MMO in the normal way... They're pretty much rebuilding an entire engine while developing the game on this incomplete engine. Look how far the game has come over the last year and a half. You might claim that it's the same, like many, but to name just a few things: Network bubble, increasing performance. CLE, completely changing the looting system. Animals Hunting Farming Fishing Crafting Large scale map changes Zombie AI Animal AI Tonnes of different items Countless bug fixes Incredible Anti-Cheat improvements Vehicles Persistence And that is a just few of the more visible things. Under the hood, there's the physics engine, the multi threading, 64 bit support and new renderer to name a few things that they've been working on almost this whole time. A lot of this is coming fairly close to completion/initial testing release, too. When you actually compare the development of DayZ to anything else, taking in to account the fact that they are building an engine at the same time, for an MMO, it is downright ridiculous to claim that development is slow. Look in to it if you don't believe me. They're actually going pretty damned fast, especially considering that nothing quite like DayZ has ever been done before.

I no longer have the time to go through and research everything I say in replies and ensure that I'm getting everything 100% correct (or as close to that as I can) now, so I'm gonna go ahead and step away from the conversation. Still feel exactly the same way as in the original post, but I don't wanna sit here going purely off of memory and typing up rushed replies and potentially doing more damage than good. If others still want to talk about the general topic etc here, feel free (and try not to get the thread locked if you actually do want a discussion), but there's more than I can be bothered to reply to - or reliably reply to with a decent degree of accuracy even if I wanted to. Did I say that it's been in development for three years anywhere? I'm actually very aware that that's not the case and have been since long before I started this thread. Apologies if I did say that anywhere. One of my key points is that we got this game pretty much at the beginning of its development, which is far from the case with any project people try to compare it to. Nowhere did i state that there's a new engine coming. They are replacing the renderer and physics engines - those are the entirely new parts. The engine, in the way you're talking about it (the entire thing, including the renderer and physics engines), as a whole, is essentially an improved (vastly, vastly improved) version of the engine we started with. However, as you clearly didn't bother to actually read what people were saying when talking about new engines, and your post was dripping with passive aggressiveness, I'm sure you'll pay little to no attention to my reply. Also, updating an engine to a large extent can absolutely make it a new engine. Look at... well... just about any mainstream engine - Unreal, Unity, CryEngine etc are all now iterated upon and improved versions of their original forms. At a certain point, it's classed as a 'new' engine. Eh, I still stand by my original post and think that it's something that was well worth saying (and think I got most of the stuff right there). I think the conversation was also pretty constructive in this threads original lifetime. However, I admit that since it was necro'd, it's broken down a bit and is now pretty full of speculation (which I'm absolutely guilty of too) and borderline flaming.

At that time, the renderer was going to be using DX9 and DX10. However, since then, things have changed. The most obvious change is that it's now going to be DX11 and DX12, which will, without a doubt, have a large impact on performance. They don't want to overpromise, which is fine. But I've also seen plenty of tweets and forum posts from devs describing 'up to four times improved client-side performance' and the likes in internal builds. The original goal of the new renderer wasn't to improve performance dramatically. But that's become one of the main focuses since.

The renderer is still tied to the simulation engine, again. But it's not in the same way as it was early on. I may be recalling it wrong, in all fairness, as it was a year and a half ago and I didn't play DayZ for about eight months or so in that time. I'll try to find some actual information on it, but basically, early on, the frame rate you got was generally the simulation frame rate. That was the bottleneck. They separated them enough so that your rendering frame rate could be higher than the simulation frame rate (refresh rate/update rate would be a better way of describing it). At that point, the renderer became the bottleneck. The two are still, in the code, very intertwined. But the renderer can achieve (by quite a margin) a higher frame rate than the simulation engine, and this has been the case for a long time now. Again though, I'll have to go through and try to find the actual updates I'm talking about. I could, again, be recalling it incorrectly. It's pretty easy to test it though. Change the scene complexity in your config file. This is, essentially, the detail draw distance in the game. It has nothing to do with simulation - it's entirely renderer based. Changing this has a large impact on your performance. At very low numbers, I can pass 60fps even in cities.

Ark has a much smaller map, has just as many performance issues, is actually a very simple game at its core and uses Unreal 4 engine - a very easy to work with, completed engine. RUST reuses the majority of its assets and code from the old version. It wasn't literally completely restarted. At all. Both games have just as many issues and are smaller projects. You also have to bare in mind that RUST and ARK only need mob assets, environment assets and modular building assets. The way the game is created is far simpler and requires a lot less work. Neither game has had to do much of anything when it actually comes to working on the engine they use. Both use engines that, generally, perform very well and are easy to work with. Yet, somehow, these games have just as many performance issues as DayZ and just as many bugs. I own all three games and have hundreds of hours in all of them. I enjoy all three games. But DayZ is a larger project, is using a much more complex system and difficult to work with (and unfinished) engine. Rust has, really, been in development for the same amount of time as DayZ. ARK hasn't, but again, it's a far simpler game. DayZ's performance issues aren't caused by a tight coupling of the sim engine and rendering engine, though. They separated the two for the most part very early in the development process. There is still some ties, sure, but the bottleneck is, by quite a large margin, the renderer at the moment. The issue is almost entirely on the renderer and the outdated, crappy technology it uses now. If it was the 'close tie' between the sim engine and the renderer, you'd still have terrible performance everywhere on the map, especially when there was a large load on the simulation engine. I get the same performance on the coast, FPS wise, when there's a horde of zombies as when there's not. My performance is only poor in large cities, where there's a lot of transparency, complex geometry and a heavy load on the renderer. Again, most of the simulation (virtually all of it) is handled by the server. The issue is client side. It's the renderer. I no longer have lower FPS depending on the server pop or zombie population. There's other performance issues, sure, but they're server side and related to the simulation engine. But the actual frame rate you get on your computer is down to the renderer.

I'm not really start raving mad about the game. I really enjoy the game, but I absolutely recognize its flaws. However, unlike those who absolutely shit on the game, I'm able to remember that I'm playing an unfinished product and the vast majority of those flaws will be fixed in the finished product. The difference between me and somebody who shits on the game is that I have patience and understanding of how games are actually developed. That's really it. I regularly complain about issues the game has. But then I let it go and hope that it'll be taken care of over the course of what remains of alpha and the beta. It's really that simple. I'm just comparing this game to others and pointing out that it's not being developed slowly - and most of those criticizing the game are complaining as if this is the finished product. It doesn't need links. You can google how long it took for just about any game to be developed. It doesn't take much work. I've also gone into more detail plenty of times in this thread alone. You shouldn't need everything handed to you.

The renderer is being literally completely replaced. That's the point. The reason the CPU load is so high (compared to the GPU) is entirely because of the renderer. It's because this game is rendered using DX7 and DX9, which aren't very good at utilizing GPU's. Compound that with the fact that the renderer its self isn't very good when it comes to distributing load between multiple cores and the GPU its self, and what you get is... well... performance like DayZ's. DX11 and 12 are also far better at utilizing multiple CPU cores as well as pushing most of the rendering work over to the GPU. The renderer is, by absolute miles, the main thing when it comes to frame rate in this game. Very little logic or AI is handled by the client. Optimization of anything else will have very little impact on client side performance (though it'll affect server side performance, as that's what the servers are handling). Because the client doesn't handle much other than the input (from the end user and the servers), display and output to the servers.

I agree that everything should be criticized... Including peoples criticisms. I see too many people shitting on the DayZ dev team for absolutely stupid reasons. Also, I don't really think my view is particularly biased. I don't really have any vested interest in the game or the dev team, other than the fact that I enjoy the game... And I enjoy the game because it's a good game. That's not bias. I try to be objective, and, looking at other, similarly sized games and looking at the actual facts on the rate of development of this game, as well as just how fun this game is (in my opinion), I just can't agree with the majority of the criticisms thrown at this game and its dev team. They're doing a far better job, honestly, than any other similar game (H1Z1, RoK, ARK, every other early access game has very similar problems). Games of a similar scale developed by huge dev teams funded by massive companies have, historically, taken longer to develop than DayZ. This game was released to the public at the literal beginning of its development cycle. It's only been a year and a half and it's come miles. If they come close to reaching their target for Beta, it'll have been in alpha for roughly two years - which is actually a very short time for a game of this scale, especially with its relatively small development team. That's my point, and it's not based in opinion at all. Literally all you have to do is look at the very clear facts about other games of a similar scale, regarding how long they took to develop from start to finish.

That's the thing though. In regards to general performance, they have been working on it consistently. Netcode is always changing. They're constantly making improvements to it. They've been optimizing the zombie AI for as long as they've been working on the freaking Zombie AI. They've done the same for animal AI, loot spawning systems, cleanup systems and just about everything else. But none of that will have an impact on the FPS, as it's done server side. That's part of the simulation 'frame rate', rather than the graphical frame rate... And they separated the two very, very early in development. You mentioned 'optimizing FPS', which is really only going to be done on the renderer's side. All of the AI etc is handled server side, again. That will not impact FPS. In regards to 'optimizing' fps, the literal only way they would do that (to any noticeable level) would be 1 - optimizing the old renderer which, again, would be completely idiotic and 2 - implementing the new renderer and further optimizing that. Guess what. They're working on that already. But again, it makes more sense to wait until the game is feature complete to do the bulk of the optimization because, again, new things constantly bring about issues. That doesn't mean they're not thinking ahead. That doesn't mean they're not doing stuff already. But it would be really, really stupid and a massive waste of resources to get everything optimized now... Because I can absolutely guarantee that by the time the game was feature complete, it would need optimizing all over again. Also, Rick, I see where you're coming from, but honestly, posting and maintaining a single thread really isn't a whole lot of work - and I genuinely can't count the number of people over the last year and a half I've had debates with on this exact topic and had them turn around at the end and tell me 'yeah, I see where you're coming from. Guess I just need to be patient'. It's really just a vocal minority imo, and believe it or not there's a good chunk of that minority who actually can see sense when presented with facts and comparisons (both comparing the game at 'release' to now and to other, similarly sized projects).

Zombie killing, friendlies, sacrifices and hordes. Here's a little video I threw together. :) https://www.youtube.com/watch?v=Kw5XXi95eAI Enjoy. :D

I went on to punch out a horde of a similar size by myself the next day. :P I did feel pretty terrible about knocking the guy out, though.

Grimey Rick, the point of the thread was to explain some simple concepts of game development that a large part of this community does not understand. As they are participating in an Alpha, they really should. Emu, I have literally no clue what you were trying to say there, or where you got that idea from. In regards to the roadmap, they have always maintained that the roadmap is a general guideline (as it always is in any software development). Everything on that roadmap will be implemented, even if delayed. Basically everything from q2 is in experimental. But they also have a tonne of stuff in internal builds and far along in the pipeline. We just don't see everything right away. Rauchsager, that's not really how it works. FPS will not improve until the new renderer is implemented. Optimising a renderer they're going to tear out would be stupid. They're currently running the game on DX9 and DX7. It'll be on DX11 and DX12 once the new renderer is finished. That alone will improve performance a lot. Net code is constantly being worked on and improved. The thing is, optimising on a large scale before the game is feature complete is just a terrible idea. Optimisation is ALWAYS possible, but adding new things in will regularly screw up optimisations. They are following the traditional method - inplement shit in alpha, then do the majority of bug fixing and optimisations in beta. That's how the vast majority of games (and software in general) are developed and it's for a good reason. Again, every time you add something new, you'll probably introduce new bugs and break optimisations. It makes far more sense to do it all once the game is feature complete (beta).

Ah, I only played on those servers like the day or the day after the infected servers came up. Gotta try it again. Zombies definitely need a little work though. Think it'd help for them to be sped up and adding a little randomness to their pathfinding (not much, just enough) to stop their beelining. I punched out like seventeen completely solo the other day just lining them up.

Oh god. I hope it wasn't that last one. Might have to take a little responsibility there. But they were super fun, so I dunno. I think it's mainly down to performance. While there wasn't any major issues for me when it came to actually fighting hordes, the servers in general had notably poorer performance than non infected servers. Considering the fact that they've said on several occasions that .58 is focusing on desync and server side performance quite heavily (hence the incredible performance on .58 servers besides the infected ones), that definitely seems like it would be the cause.

Haha, I do like a good bit of 1pp... But honestly, I'm really, really a fan of the new camera for cinematic purposes. :( Shh. Our little secret.

Yeah, I actually just solo punched a horde of 17 to death, too, without even bleeding. I think the main issue is the zombies speed. Really easy to make them beeline when you're soloing them. Also need more attack damage, imo. Maybe stop you running for a couple seconds with grabbing attacks, too. But bare in mind that I am obviously a complete pro at this game. So maybe it's just that. :P

Solo British streamer who welcomes viewers to come find me. I generally, on stream, aim to get as much player interacton as I can while avoiding Elektro and, usually, Berezino. Not too terrible at PvP and aiming for a RP type vibe. Normally, I'll try to stream daily or every other day, especially at the moment while I'm on holiday from both work and college. New to streaming, but have done YouTube on and off for several years and currently have a small YouTube channel which I mainly use for edits from streams. Currently, I'm only really streaming DayZ and a little Reign of Kings, though I'm planning on branching out. DayZ will, however, for the forseeable future, always remain my main game. http://www.twitch.tv/iSpielHalle Streaming in an hour or two.

The AI is still pretty basic, yeah. But what I was looking at was the performance I got when in a large group of zombies. I had no issues. That said, there definitely needs to be some more RNG. Like you suggested, movement speed would be a good one. Waits before a swing. Which attack it's going to use. They have a tonne of different attacks now, but for some reason they all seem to use the same one at the same time. They also probably shouldn't be able to damage each other, as was visibly happening in the video. Overall, however, I'm incredibly happy with the advances they've made with them. It's really not too far to go to perfect them from here. Add some more RNG and then get them to horde naturally, even when not chasing a player (when one zombie comes within x meters of another, they try to stay close to one another).

Yeah, it wasn't too bad, to be honest. It's not too clear from the video as obviously a lot of the combat was sped up. Was considering leaving it normal speed, but didn't for the videos sake. The zombies early on about 50% of the time wouldn't react until they were KO'd, but weirdly enough the entire horde was reacting properly, so maybe the server was just having difficulties at the time (it did freeze for a while around 6:00). Seemed to be absolutely fine after those players lost connection, so maybe there was some kind of an issue with their connections, dragging down the server?

I got a nice little clip of my first ever firefight today. I was in Berezino, as the introduction to the video established (the clip will start just before the actual action) and got attacked by some dude. Ran off after emptying my MP5 clip to no avail. Hid in a green house. Checked the door, shot a guy then, finally, remembered to hit record. Here's the very quick bit of actual action. http://www.youtube.com/watch?v=3qA8m3owe0c EDIT: Skip to 2:06 for the action - the time link wasn't working with embedding. Hope you enjoyed that - and if you went back and watched the whole video, much love! It was only five minutes. :D

So, just read the status report and, forgive me if I'm wrong, but it appears that persistence is to be disabled for the next patch. I'm kind of left scratching my head with this bit of news as persistence really is a huge part of this game - and the implementation of a CLE to stable without persistence is just going to give a false impression to everybody who does not keep up with development. Not only does a Central Loot Economy really kind of need persistence to work properly, but the fact that persistence will be disabled is going to piss off a whole lot of people, especially when compounded with the disabling of zombies. I feel like it really is getting to be a perfect case of one step forward, two steps back. So, I just have to ask, what is everybodies thoughts on this? I have been really, really looking forward to .57 and had been loving the .56 experimental branch. However, with the latest update and look at .57, as well as the news that persistence is to be disabled (and the constant crashing I have been experiencing on exp, though I'll assume rather optimistically that that won't be an issue on stable), I'm starting to feel like I might need to just sit this one out. So... Opinions, everyone? I just don't see how the CLE can be implemented properly without persistence. EDIT: And don't get me wrong. I'm not hating on the devs, or crying about the game. I do get why they're doing this an all. I just think that this move is counterproductive to the development of the game (slowing down the real implementation of a massive feature and cutting down on the amount of large scale testing that's done on said feature) just to avoid people complaining. Like when they went back to legacy loot last stable update. If a system doesn't work enough to be released without disabling larger parts of the game, maybe it just shouldn't be released until it does work. It's like removing the first story of a building because the second story wasn't stable enough. The second story might be filled with awesome shit, but there's no freaking doors, now! EDIT: I just want to make sure my stance is clear, though. I am not trying to hate on the devs or complain about the development of the game. I just don't think this is a good choice. If you are planning on posting in a way to bring it down to that level, please reconsider. I'd like to keep this civil and mature. Quick points before you claim 'alpha' or that persistence isn't working.

I don't think you're quite correct there. The running speeds are unrealistic. The zoom really wasn't. I used to do a lot of track. My school had a 400m long field that we used for it. I could certainly see people on the other end of it. Pretty well, actually - I mean, I couldn't tell who it was, but I could see there was a person there, what colour clothes they were wearing, when they moved their arms or legs, etc. I haven't got great vision. My eyesight is actually pretty fucking bad (-4), but I wear contacts (+3.75). As such, my eyesight is not even at normal levels and I can see good detail a lot further than I can in DayZ (it can be very difficult to spot people past 200m in DayZ). In the lake district, I was able to just about make out people on hills at much larger distances. So, why is this? Because a monitor isn't a good substitute for our vision. We have a very high field of view, with a very high level of detail. In games, the FOV is generally very high, to mimic our peripherals. As such, we lose a lot of the detail, because more is crammed on to a small screen. The zoom function simply gives us back the detail that we should be able to see in... And at this point, it is nowhere near enough.

So, your group doesn't have a 'no kids allowed' rule. It has a 'kill all kids you find, even if they help you' rule. Not like you could have just handcuffed him and walked off as he broke out.

I'm talking about the zoom with making it closer to real life. We can see in more detail than what we can see when we have a high FOV. However, without a high FOV, we have no peripheral vision. The zoom mimics how much detail we can actually see, at the temporary cost of peripherals. The higher FOV mimics our peripheral vision, at the cost of how much detail we can see without zoom. It's really that simple. There is no better way to do it. It's far more accurate to real life than being able to zoom by lowering your FOV then zoom more via RMB. They can't remove the zoom, because it's needed for seeing realistic detail when at a high FOV... So they cut out the lower end of the FOV. The RMB zoom is literally just enough to see how much detail we would be able to see in real life. EDIT: Oh shit, might have misread your original post. EDIT 2: Yup... Read the first part in detail, skimmed the rest. Saw 'min FOV', not 'zoom', assumed you were just talking about FOV in general (as they've changed that, too). In that case, apologies - I agree with you. That said, there's plenty of people complaining about zoom in this thread - and I actually was directing my original post to them. Was just too many to bother quoting. :D

The zoom is to mimic how much detail we can really see in. The high minimum FOV is to mimic the fact that we have peripheral vision. Both are roughly where they should be at this point. Sure, it sucks that you can't see as much detail at a difference in DayZ without zooming. But the only other options would be to have no peripheral vision, or to have far too little detail. Zoom isn't actually zoom. It's literally just giving you the amount of detail that you should be able to see. Removing it would be stupid. This is a videogame. They can't mimic how we see perfectly, unless you're going to get very dynamic depth of field that uses eye tracking, a screen that curves to your entire FOV and a perfect resolution, screen size, viewing distance and FOV balance. This is literally the only way they can mimic human vision. It actually works pretty fucking well, too. The issue with having a lower FOV, besides lacking peripheral vision, is that it creates far too much zoom. Too high FOV makes it far too little. They could set the zoom to a static FOV, but that would be incredibly disorienting from high FOV's and would be higher than very low FOV's. They have to limit the FOV on both ends - and use the zoom to make up for it. I wouldn't call it nerfing. It's getting it closer to real life, excluding the need for zoom (and it is an absolute need). That's kind of a step forward, in my books, when the game prides itself on realism authenticity.