|
Chernabog posted:Are Facebook/Web games still a thing? Still a thing but the big money is mostly in apps.
|
# ? Jan 8, 2018 22:07 |
|
|
# ? Apr 20, 2024 04:07 |
|
That's what I thought. I worked on a couple of Facebook games back when they were all the rage 5-6 years ago but I haven't heard much ever since.
|
# ? Jan 8, 2018 22:16 |
|
Chernabog posted:Are Facebook/Web games still a thing? Much of it has shifted towards smartphones and other mobile devices. Several of the big Facebook games lost to competitors who got a mobile version of their genre out faster, such as Farmville losing out to Hay Day. Many Facebook gaming giants tried to adapt by developing for mobile platforms or acquiring mobile gaming developers, such as Zynga acquiring NaturalMotion. In Japan, browser gaming is still strong. It's not unusual to see games that are concurrently developed and supported for PC, web browser, and mobile.
|
# ? Jan 11, 2018 22:41 |
|
djkillingspree posted:On the other hand, something that doesn't get a lot of fanfare, but is a huge deal, are the massive advances made in the capability of AIs to navigate environments. In order for a game like Assassin's Creed to work, for example, a fairly large number of AIs need to be able to understand how to navigate a fairly complex environment that can't be defined with a simple navmesh. Those kinds of advances don't get super-buzzwordy-reveals, but it's important to remember that in many games 5-10 years ago, enemies couldn't even figure out how to climb ladders, jump over gaps, or take cover behind objects without a human annotating the environment. I'm sure there's still some hand annotation going on in newer games, but it's clear that AIs have gotten drastically better at navigation in open world games over time. This is an extremely hard problem and has a ton of good work being done on it all the time. This is basically it, and there's still a ton more changes that have gone on behind the scenes: 10-15 years ago (I'll cherrypick HL2 as an example bc a lot of people are familiar with how it works), you might have needed to find paths for 20-30 agents at once, from one thread, on one static data structure. HL2 started with a waypoint graph and later on switched to navmesh for some stuff. The entire map is going to be resident in memory the whole time you need to access it. If you wanted to avoid any dynamic objects, that'd be done as another step after the pathfinding. It's not uncommon these days to have to manage paths for up to 150 agents at once, who might be updating from any thread, on up to 16 cores. Because the maps are so much bigger (32km x 32km is the upper limit of the way most engines do things right now, but people are working on solutions for going bigger), your paths could change lod or even be streamed out entirely completely without warning. There are multiple data structures to use, depending on the situation--many open world games use a navmesh for pedestrians and small agents, and a waypoint graph for vehicles and big stuff. Depending on your gameplay needs and level design you might also use a grid for some stuff. And those data structures are more likely to be dynamic, so moving objects get taken into account at the time of pathfinding instead (or in addition to) a second step. And of course those objects could be registering themselves with the graph from any thread. But the experience to the player is still exactly the same, it's dudes walking from point A to point B. eshock fucked around with this message at 16:49 on Jan 12, 2018 |
# ? Jan 12, 2018 16:46 |
|
More "games" than "games industry" question here, but maybe it can set off some good talk points about game development and the industry anyway. Been following this thread with great interest and wanted to cross-post two questions I posited in the "Recommend me a game" thread (got some great feedback already and thought maybe people posting here had some more insight/recommendations): quote:Any recommendations for RPGs/games that does inventory and gameplay items well? Without generic junk items or design around tedious pack-rat management? quote:I have another question, about greatly balanced games:
|
# ? Jan 12, 2018 21:42 |
|
Downs Duck posted:More "games" than "games industry" question here, but maybe it can set off some good talk points about game development and the industry anyway. The inventory progression and puzzle balance in Shenzhen I/O is very good and very interesting.
|
# ? Jan 12, 2018 22:00 |
|
Downs Duck posted:More "games" than "games industry" question here, but maybe it can set off some good talk points about game development and the industry anyway. You should play Torchlight (or probably Torchlight 2 since it's a more polished version with more features) The entire game is based around getting cool drops and off the top of my head I'd say it has a more forgiving inventory than many other RPGs in terms of the amount of slots you're given. Also Borderlands / Borderlands 2.
|
# ? Jan 13, 2018 00:47 |
|
GC_ChrisReeves posted:Question I have for you. I just came out of my first time trying Oculus Medium and I have to wonder. https://www.facebook.com/oculusmedium/videos/910156012499469/ Seems like it might be catching on, but I don't know how much of this is PR exaggeration vs. the dev actually using it day in and day out.
|
# ? Jan 13, 2018 00:48 |
|
leper khan posted:The inventory progression and puzzle balance in Shenzhen I/O is very good and very interesting. I couldn't program my way out of a paper bag so that's one title I haven't considered before, thank you. Whistling rear end in a top hat posted:You should play Torchlight (or probably Torchlight 2 since it's a more polished version with more features) The entire game is based around getting cool drops and off the top of my head I'd say it has a more forgiving inventory than many other RPGs in terms of the amount of slots you're given. Yeah, I'm installing both Torchlights to check out, thank you, it's great if the drops actually have purpose and isn't just generic items. I was under the impression that the Borderlands games had an incessant amount of generic and at many times unbalanced loot, but maybe I am wrong about that?
|
# ? Jan 13, 2018 09:49 |
|
Downs Duck posted:I couldn't program my way out of a paper bag so that's one title I haven't considered before, thank you. I haven't played the first Borderlands, but otherwise I wouldn't call the loot "unbalanced". It's a FPS diablolike, so the random guns/grenades/armor/etc. there spawn just like any gear grinder, and just like any other gear grinder, you'll not care about a lot of them. The loot is definitely balanced, and that balance is on a continuum. It's worth noting that Legendary and Unique weapons are way weirder than Diablo/Path of Exile/the hour of torchlight I've played and some are definitely better than others. Especially given the oddities of the rarest weapons, if you're min-maxing, then yeah, it's probably "unbalanced" at a given rarity level, but I think the game has the balance the developers intended. Also, the Lascaux might be my favorite joke/gimmick/whatever in a video game, so if any of you did that: props.
|
# ? Jan 13, 2018 19:28 |
|
hey girl you up posted:I haven't played the first Borderlands, but otherwise I wouldn't call the loot "unbalanced". It's a FPS diablolike, so the random guns/grenades/armor/etc. there spawn just like any gear grinder, and just like any other gear grinder, you'll not care about a lot of them. The loot is definitely balanced, and that balance is on a continuum. Yeah, that's me getting my wires crossed between my "generic inventory" question 1 and "balance question" 2. Thanks for the input.
|
# ? Jan 13, 2018 20:41 |
|
Hello gamedev goons - I'm an indie solo gamedev with one commercial title under my belt (if you've played it, apologies in advance) and have been doing gamejam stuff mostly over the last few years. I've never worked in the game industry for a AAA/AA/III company or anything like that so I have a question for those who do. What's a designer's day to day like? I get the prototyping/hashing out different concepts and mechanics bit in the pre-production stages, but during production what role does a designer fill? Do most designers tend to do additional stuff when there's less gameplay-specific design work to do on a title (e.g. filling in for level design). Same thing for sound (assuming you're a studio with a sound person, rather than a contractor) - are you continually tweaking stuff, or is it a "once it's done you do other things" type scenario. I imagine it's going to be vastly different from studio to studio but I'm interested in your day to day, in case I ever get sick of making dumb jam games from home in my underwear
|
# ? Jan 26, 2018 09:17 |
|
KRILLIN IN THE NAME posted:What's a designer's day to day like? I get the prototyping/hashing out different concepts and mechanics bit in the pre-production stages, but during production what role does a designer fill? Do most designers tend to do additional stuff when there's less gameplay-specific design work to do on a title (e.g. filling in for level design). Game design - or good game design, at any rate - is inherently an iterative process. In preproduction, you hash out rough ideas about what the game will be about, what the basic gameplay loop will be like, and get a rough idea of what features you want. Then you start making a prototype, and whoops: half the poo poo we came up with is bollocks. So we change it. And that's bollocks too, so we keep changing it - that part wasn't fun, this part needs tweaking, the numbers on this weapon aren't balanced at all, this feature is only a shell of a design and needs fleshing out before the programmers can even start implementing it, here's an idea for a new feature that we think would make the game more fun... The rest of the team are working hard, meanwhile. You don't need a fully complete design to start writing code, or producing artwork. Sometimes that means a designer has to come in and ruin someone's day by announcing that a feature has been cut, or worse, that they want to try a new feature that has no prepwork done for it at all. But that's life, you know? No one in the history of game design has ever nailed the final design of a game in pre-production, usually not even close. In SWD2 our designers were cramming new features into the game way, way late in the development cycle, sometimes to the consternation of the programming team (including me). (Now the foot is on the other shoe and I get to be the one who makes up features for programming to do! ... except, oops, I'm still the programming lead so I'm just giving myself more work. Oh well.) Hyper Crab Tank fucked around with this message at 12:51 on Jan 26, 2018 |
# ? Jan 26, 2018 09:51 |
|
There’s also a huge chunk of programmers/QA going “what do you want to happen if they have zero of that/someone else does it at the same time/the player is on fire?”. You think you’re being specific until it needs implementation and then there’s a million edge cases you didn’t cover. Especially when interacting with other systems.
|
# ? Jan 26, 2018 10:33 |
|
KRILLIN IN THE NAME posted:Hello gamedev goons - I'm an indie solo gamedev with one commercial title under my belt (if you've played it, apologies in advance) and have been doing gamejam stuff mostly over the last few years. I've never worked in the game industry for a AAA/AA/III company or anything like that so I have a question for those who do. Design is writing proposals and mucking with excel. Also fielding questions from anyone confused about how something is intended to work.
|
# ? Jan 26, 2018 12:47 |
|
There are also content designers, people who make puzzles, levels (greyboxing), story, quests, texts, ui and so on. Some designers can code and they do some prototyping themselves.
|
# ? Jan 26, 2018 14:42 |
|
KRILLIN IN THE NAME posted:Hello gamedev goons - I'm an indie solo gamedev with one commercial title under my belt (if you've played it, apologies in advance) and have been doing gamejam stuff mostly over the last few years. I've never worked in the game industry for a AAA/AA/III company or anything like that so I have a question for those who do. A lot of it depends on the size of the team. My current team is a large mobile game and has 3 designers. At that point all of them are what in AAA would be a level designer and a systems designer. So creating levels, balancing difficulty of content, iterating on tens of thousands of pieces of data, ensuring the scale of the game feels right. Plus as engineering works on their task, design will be iterating with them on it finding the fun. Once the engineering task is done then design will be integrating it into the rest of the game. IE imagine a FPS game and in month 4 engineering finish a table flip mechanic. Any level done up to that point has reserved spaces for tables to be flipped but now they need to be hooked up play tested, move a bit, maybe make the table bigger, maybe smaller until the combat loop feels tight. Doing a retrofit like this could be an entire small strike team of level design, world art, combat designer, gameplay engineer in a huge AAA game, or “A Designer” in a small team.
|
# ? Jan 26, 2018 15:03 |
|
What are the favorite engines to work with for the devs here, and why? Which do you think are overrated or underrated?
|
# ? Jan 26, 2018 23:43 |
|
Unity... because that's the only one I know and it does everything I need and I was able to learn it without any prior coding experience.
|
# ? Jan 27, 2018 00:19 |
|
Bushmaori posted:What are the favorite engines to work with for the devs here, and why? Which do you think are overrated or underrated? Depends on what I'm doing. Generally speaking, I don't care what an engine's reputation is; if it gets me from start to finish faster it's the one I'm going to use. Prototyping: Unity or Game Maker: Studio (majorly dependent on 3D vs 2D, with a bias towards Unity if it's menu heavy) RPG development: RPG Maker MV or ORK Framework with Unity Adventure: AGS (Adventure Game Studio) or Unity -- depending on what I need in the engine. Simulation: Game Maker: Studio (low level prototyping), Unity (high-level or very systems-complex prototyping) Visual Novels: Presently 'Visual Novel Maker' unless I need a very gameplay-element-heavy VN, in which case I use Unity. I also pull out various subtools depending on what I'm working on. I've got some $2000 invested into Unity add-ons to make my life easier in a variety of regards. If you have questions about engines that aren't super-new or haven't really hit 'commercial-grade adoption' yet, I've probably messed around with the engine you're considering so feel free to ask. Essentially: if it has a well-developed and stable (no longer receiving major updates) IDE I've probably played with it -- including smaller engines like Visionaire and 001 Engine. If the engine's very code-heavy (e.g. PyGame, Ren'Py, MonoGame) I probably haven't worked with it by contrast as I don't like particular enjoy engine development. Firgof fucked around with this message at 07:36 on Jan 28, 2018 |
# ? Jan 28, 2018 07:32 |
|
Downs Duck posted:More "games" than "games industry" question here, but maybe it can set off some good talk points about game development and the industry anyway. Super Smash Bros Melee has had 17 years to be utterly dissected and yet not only do people still play it but the metagame is still evolving
|
# ? Jan 28, 2018 08:52 |
|
Firgof posted:Visual Novels: Presently 'Visual Novel Maker' unless I need a very gameplay-element-heavy VN, in which case I use Unity. Can you recommend any Unity add-ons/tutorials/etc. that you've used for VNs? I'm currently writing a VN-style adventure game where the player interacts with other characters not by directly choosing from a list of appropriate responses, but rather by making short phrases from an inventory of words that they always have available. (I'm also considering other gameplay elements for some scenes but that may or may not be within scope.) I'm definitely not married to using Unity if there are better tools for the job - it's just the engine I'm most familiar with (admittedly this is not saying much, relatively speaking) and I know it'd be flexible enough for what I want to do.
|
# ? Jan 30, 2018 04:30 |
|
I'll shill this once, and only once here, but if anyone is interested in 3D art, I've started a twitch channel where I'm livestreaming work on an Overwatch-themed large scale environment art project. I usually stream Wednesday nights, but I'm trying to find other nights of the week to add to the schedule. https://www.twitch.tv/mutatedjellyfish Going live in 10 minutes.
|
# ? Jan 30, 2018 04:51 |
|
I have a question about Hackers. I love FPS games like Playerunknown's Battlegrounds, CS:GO, Battlefields, etc. It seems like every popular twitch multiplayer gane like these all attract hackers as they grow in popularity. Cheating players will pay hundreds of dollars per month to access to these hacks (which is just insane). It then turns into a never ending race to the bottom as developers release new patches and hackers develop new hacks. Valve has been trying a community rating system, but that's far from perfect. Is there any hope? daslog fucked around with this message at 21:44 on Jan 31, 2018 |
# ? Jan 31, 2018 21:36 |
The best protection against cheating in multiplayer games is to have a zero-trust model, where the server assumes every client packet is malicious and verifies that everything follows the rules, and additionally gives the clients as little data about the world as possible, so the client can't know things the player shouldn't (e.g. what's behind a solid wall.) The problem with doing that is latency, you simply can't implement client prediction in a good manner if you don't trust anything. The other alternative is to force players to only use consoles with strong hardware DRM, and encryption on the network data to prevent spoofing from external devices. One other possibility could be to add heuristics to the servers that verify actions against what should be reasonably possible, e.g. a player with a ridiculous turn-rate that hits headshots every time is suspicious, even more so if it's against cloaked players. But any heuristics like that massively increase the risk of false positives. I'd say it's not a battle that can be won with technology alone. But the right mix of less trust in the client, together with community voting reinforced by server-side heuristics, could help somewhat.
|
|
# ? Jan 31, 2018 22:36 |
|
daslog posted:I have a question about Hackers. I love FPS games like Playerunknown's Battlegrounds, CS:GO, Battlefields, etc. It seems like every popular twitch multiplayer gane like these all attract hackers as they grow in popularity. Cheating players will pay hundreds of dollars per month to access to these hacks (which is just insane). It then turns into a never ending race to the bottom as developers release new patches and hackers develop new hacks. nielsm posted:The best protection against cheating in multiplayer games is to have a zero-trust model, where the server assumes every client packet is malicious and verifies that everything follows the rules, and additionally gives the clients as little data about the world as possible, so the client can't know things the player shouldn't (e.g. what's behind a solid wall.) The problem with doing that is latency, you simply can't implement client prediction in a good manner if you don't trust anything. The other alternative is to force players to only use consoles with strong hardware DRM, and encryption on the network data to prevent spoofing from external devices. peer based systems are also cheaper to run than an authoritative server. so no, there's no escape.
|
# ? Feb 1, 2018 02:52 |
|
daslog posted:I have a question about Hackers. I love FPS games like Playerunknown's Battlegrounds, CS:GO, Battlefields, etc. It seems like every popular twitch multiplayer gane like these all attract hackers as they grow in popularity. Cheating players will pay hundreds of dollars per month to access to these hacks (which is just insane). It then turns into a never ending race to the bottom as developers release new patches and hackers develop new hacks. I think there is a nonzero chance that CPUs become robust enough that keeping runtime memory safe might be possible. If a CPU has enough time to obscure important data in memory so much that even it can't decode it fast enough to use it (without a key of course!), that'd be kind of the dream. The nice thing is that this is possible on paper and I honestly don't think it's that far away. In fact, with Sufficiently Clever Algorithms™, we might already be able to theoretically do it. Attacking the client/server comms will always be a thing in multiplayer games, but games can do a lot of stuff to make it really hard to do it effectively, but it's also something people have been working on for 20 some odd years, so even today, we are pretty good at this, but not perfect. Finally, attacking the software at a 'high level' will probably always be some variation of possible, but usually unreliable for a hacker. Instead of attacking data or the memory, imagine instead that you just capture the data on the frame buffer and then write an algorithm to explain to a computer what a 'head' is in a game, and how to move the mouse there and click. Since you aren't dealing with absolute states, an attack like this won't be consistent, but it's also very difficult to prevent.
|
# ? Feb 1, 2018 03:19 |
|
Canine Blues Arooo posted:Finally, attacking the software at a 'high level' will probably always be some variation of possible, but usually unreliable for a hacker. Instead of attacking data or the memory, imagine instead that you just capture the data on the frame buffer and then write an algorithm to explain to a computer what a 'head' is in a game, and how to move the mouse there and click. Since you aren't dealing with absolute states, an attack like this won't be consistent, but it's also very difficult to prevent. There's already cheating tech that works based only on what is rendered to screen. There was actually stuff that did this way back in the Quake days -- IIRC, this required that you delete / replace textures on the client so opponents would be rendered in certain colors and then the system just looked for pixels with a certain RGB range. This kind of thing means that even the "Holy Grail" of anti-cheat -- all of the game running on a secure server somewhere and every player just a dumb terminal sending up commands and streaming down frames -- has vulnerabilities even if you can solve the (non-trivial) cost, bandwidth, and latency issues required to make that work. Heuristics can certainly be improved but when these first started to appear, the cheat providers just added knobs to their tools so that they could be tuned -- don't always insta-headshot, instead only shoot if enemy is within a certain proximity and then add a random amount of delay for reaction time and then add some inaccuracy to the shot, etc. And, no matter how good system like this get, they generally create a lot of customer support calls arguing their validity. ...which is the problem with the community rating approach (at least one maintained by the dev) -- that's a massive customer support and community team burden. Nobody who is banned, even people recorded on stream using hacks, ever says “yeah, you got me, I was cheating.” They send emails and call and show up at your office. Like an episode of Cops where they’ve pulled over a driver caked in cocaine, the cheater will just repeat over and over that they didn’t do anything or that someone else was playing their account (the Cops-episode equivalent of the “these aren’t my pants” when something is found in a pocket during a frisk). If you maintain the ban, they then thumbs-down you on Steam and visit every place on the internet to post how awful your game is. It sucks. The arms race of trying to keep this even slightly under control is very costly and time consuming.
|
# ? Feb 1, 2018 15:25 |
|
Mother posted:There's already cheating tech that works based only on what is rendered to screen. There was actually stuff that did this way back in the Quake days -- IIRC, this required that you delete / replace textures on the client so opponents would be rendered in certain colors and then the system just looked for pixels with a certain RGB range. This is still very much a thing—not just messing with enemy textures to make them easily detectable by something like AutoHotkey, but also making it possible to see through level geometry and poo poo like that. There are a number of approaches to detect that kind of thing, though. It does seem like any game that a computer is capable of playing better than a human is going to have to deal with cheating until the end of time. It's sort of like all of the goofy rear end websites that prevent you from right-clicking on images in an effort to stop you from copying them (this seems to have become less popular, at least); if a player can see it, there's really no way to prevent them from showing it to a computer that has much better reaction times and accuracy than they ever will.
|
# ? Feb 1, 2018 15:49 |
|
Just make all multiplayer games turn based affairs with no hidden information and server side RNG. Job done.
|
# ? Feb 1, 2018 17:29 |
|
As long as the user can tamper with the operating system, install third-party drivers, and run in a VM, there's no way to get rid of cheats. A distrusting server can prevent a lot of cheats, but it can't prevent any cheat that mimics superhuman reflexes. Hiding information from the client is also easier said than done, like yes you can stop sending information about players that should be blocked by LOS, but then they stop making sound. The upshot is that detection only needs to work once to ban a cheater, and going to higher privilege levels in the OS or even hardware to stay hidden always means a drastic increase in complexity.
|
# ? Feb 1, 2018 17:33 |
|
Wallet posted:This is still very much a thing—not just messing with enemy textures to make them easily detectable by something like AutoHotkey, but also making it possible to see through level geometry and poo poo like that. There are a number of approaches to detect that kind of thing, though. When I saw this setup, anti-cheat had already moved into the realm of interrogating the OS to see if you had anything suspect running. The enterprising cheaters started using two boxes, one that was just running the game and would report back "clean", and a second that was networked and (essentially) just looking at the screen and managing the actual cheating. This was early 2000s and already at that level of effort / sophistication. No matter what you do to prevent it, someone is going to find a way around it.
|
# ? Feb 1, 2018 17:50 |
|
MissMarple posted:Just make all multiplayer games turn based affairs with no hidden information and server side RNG. Job done.
|
# ? Feb 2, 2018 08:31 |
|
ShadowHawk posted:People already cheat at online chess, adding a random element doesn't make it that different unless the randomness determines the game almost entirely. And then you've just re-invented online poker. The random element determines individual hands, but not sessions or tournaments in poker. There is a nontrivial amount of skill in the game. Not the least of which involves vigilance, since it’s such a boring game to play well.
|
# ? Feb 2, 2018 13:04 |
|
Cheating is always going to be with us one way or another. What we can do is mitigate as much as we can. * Secure what can reasonably be secured. Don't trust clients if you don't have to; a lot of the time it doesn't even cost you anything, it's just a matter of recognizing when you're about to trust something you shouldn't be trusting. * Make it difficult enough to cheat to deter casual cheaters. Determined cheaters will still do it, but most would-be cheaters aren't determined, they're lazy and don't want to put a lot of work into researching how to cheat in your game, let alone spending money. You can go a long way by making it hard enough that Li'l Bobby Tables can't trivially do it. The bike lock principle, basically. * Design your game so cheats don't have as much impact. If your game has 15 minute matches, then if you run into a cheater then at least it will be over soon. Same if your game has come-and-go gameplay where honest players can just leave games with cheaters in them at-will. * Make it possible to detect inconsistencies after the fact, or by human inspection. If you can't detect it while it's happening, then being able to detect it later on will at least make it possible for you to punish/ban people who do it a lot.
|
# ? Feb 2, 2018 14:50 |
|
The more you can put it in player hands, the better you are, IMO. The best FPS multiplayer experiences I've had, and pretty much the only ones I want to play, are community-run affairs. A publisher might have to be reasonable and careful and fair with how they treat their players but a community doesn't - if an admin on a private server wants to ban you for whatever, it's not a problem. Of course, from the publisher perspective, this isn't the best since it isn't mass-market friendly and since putting things in player hands has all kinds of implications for F2P or loot crate-type models.
|
# ? Feb 2, 2018 15:52 |
|
leper khan posted:The random element determines individual hands, but not sessions or tournaments in poker. There is a nontrivial amount of skill in the game. Not the least of which involves vigilance, since it’s such a boring game to play well.
|
# ? Feb 4, 2018 01:53 |
|
I often see devs ignore certain issues that seem relatively simple to fix or straight up make some weird design/balance decisions. I have to imagine there is more to the problem than the players can see. Perhaps it is a tech issue, the company has too much bureocracy slowing everything down or it just doesn't line up with the business interests of the company. Am I really far off? How common is this? Particularly in AAA. Also, I do know that sometimes a "simple" solution is not actually simple at all or it just isn't a priority. Something minor such as adding a button or adjusting a number could involve many people and impact other areas.
|
# ? Feb 6, 2018 18:44 |
|
It would help if you could give an example of the kind of thing you mean, but usually when an issue goes unfixed for a while it's some combination of:
|
# ? Feb 6, 2018 22:24 |
|
|
# ? Apr 20, 2024 04:07 |
|
what about 'the bug is really funny and doesn't detract from the game'?
|
# ? Feb 6, 2018 22:32 |