Announcement

Collapse
No announcement yet.

Hit registration/Netcode in this build

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

    Hit registration/Netcode in this build

    It seems that this build has brought a significant decrease in hit registration across the board. In previous builds, on occasion you'd have shots that appear to fly through the enemy hitbox, but deal no damage. In return you'd also actually see shots that hit you, but you don't take any damage. In this build this seems to now be a frequent problem. Example:



    Shot 1: Flies through the centre of the model without dealing damage.
    Shot 2: Hits the model on what I can only assume is the highest part of the hitbox, but doesn't register a headshot.

    Both players in this video had around 25ms latency (totalling 50ms difference).

    In my opinion, in this stage of development, there are only a few factors that should be of critical priority: performance, networking and the best gameplay balance that can be achieved during testing. Performance has improved, gameplay balance is reasonable with a few exceptions, but networking is really not hitting the mark. It's really quite discouraging to be involved in league/competitive play when these kinds of issues are present. It makes the game feel unfair, and worse than that - not enjoyable.

    #2
    I haven't played this build that much but netcode seems to be about the same as last build. It feels better, even, but that's probably just due to better performance.

    Did you actually hit those shots on your screen? The video is from a replay, and those can be very inaccurate. There is definitely a bug where you can sometimes hit someone and the shot stops at their body but does no damage, however this doesn't seem to be happening any more often in this build than it did in the last (still pretty often though).

    Comment


      #3
      Originally posted by Barktooth View Post
      I haven't played this build that much but netcode seems to be about the same as last build. It feels better, even, but that's probably just due to better performance.

      Did you actually hit those shots on your screen? The video is from a replay, and those can be very inaccurate. There is definitely a bug where you can sometimes hit someone and the shot stops at their body but does no damage, however this doesn't seem to be happening any more often in this build than it did in the last (still pretty often though).
      Yes, those shots hit and they hit on replay. This isn't an isolated example. Looking through that demo, about 30-40% of onscreen hits didn't register. I see these shots fail multiple times per game, and it genuinely feels horrible to play with.

      Comment


        #4
        Originally posted by Tamerlane View Post
        Yes, those shots hit and they hit on replay. This isn't an isolated example. Looking through that demo, about 30-40% of onscreen hits didn't register. I see these shots fail multiple times per game, and it genuinely feels horrible to play with.
        Weird.. for me I usually get no more than 3-4 shots per game that hit and don't register (with hitscan, not counting the direct rockets/flak balls that don't register) , so 30-40% seems excessive. Maybe that server was lagging out?

        Either way Epic needs to look into this. And the direct projectiles not registering bug has literally been in the game for years now, so please fix that too.

        Comment


          #5
          To be honest, you HAVE to take this video with a grain of salt. It's the replay system which is janky as hell right now. I would rather see actual game play footage recorded live and not from the replay system. I'm certain we could all find super funky looking things in any replay.
          ON-GOING PROJECTS: DM-BloodCovenant, DM-Campgrounds (Absolute)
          CONTACT & TWITCH: Absolute Discord
          , Twitch

          Comment


            #6
            Yeah there is something about the movement registration that severely effects hit registration. I don't know how to explain it but there definitely is an issue somewhere that has been overlooked.
            X58 i7 970 4GHz HT off, GTX 1070, 144Hz 1ms, 1080p in-game res.

            Comment


              #7
              Originally posted by 213 View Post
              Yeah there is something about the movement registration that severely effects hit registration. I don't know how to explain it but there definitely is an issue somewhere that has been overlooked.
              It's because of the movement prediction, which consistently puts players in positions they never were on the server, so you can line up a shot and directly hit the enemy on your screen but if they changed their direction (which happens quite often), it won't register. All these issues with hit registration... I predicted over a year ago. And it's really frustrating from my perspective because I play (well, I used to play) the game so I'm painfully aware of the frequency of unregistered shots, and I also know that this stuff is relatively easy and should have been top priority and completed over a year ago. I had also offered to implement the hit registration myself but was turned down.
              Creator of NewNet for UT99

              Comment


                #8
                Originally posted by Trinatek View Post
                To be honest, you HAVE to take this video with a grain of salt. It's the replay system which is janky as hell right now. I would rather see actual game play footage recorded live and not from the replay system. I'm certain we could all find super funky looking things in any replay.


                Cmon man everyone has seen this ****.

                Comment


                  #9
                  Originally posted by timbur View Post
                  It's because of the movement prediction, which consistently puts players in positions they never were on the server, so you can line up a shot and directly hit the enemy on your screen but if they changed their direction (which happens quite often), it won't register. All these issues with hit registration... I predicted over a year ago. And it's really frustrating from my perspective because I play (well, I used to play) the game so I'm painfully aware of the frequency of unregistered shots, and I also know that this stuff is relatively easy and should have been top priority and completed over a year ago. I had also offered to implement the hit registration myself but was turned down.
                  Honestly I can't say I'm surprised, the issues with the netcode have been brought up so many times just to be kicked back with 'its fine at the moment'

                  It's clearly not fine. there are constant complaints, constant raging in-game over it, people are leaving, and it seems to be over ideology on the devs side.

                  Just recently another game - called Overwatch - was released/went through open beta, whose net code favors the shooter, this game feels so incredibly smooth to play even when your pings are approaching 200ms simply because the devs there feel that there is more anger in missing a shot you know connected vs being shot slightly around a corner, maybe sometimes. the game is extremely easy to play for 10 to 14 hour sittings nearly non-stop and it's simply because I have never felt like I'm being screwed over by the netcode (oh, and it's actually fun and flows very well which is a huge bonus). their servers even run at 20 tick and I have no noticeable problems.

                  Overwatch even has what I would classify as perfect input.


                  in UT, the input has been inconsistent for a very long time, seemingly framerate linked.. I used to dispose of what people said quite often back then but after going from Overwatch to UT back to back I can reliably say that something is very, very wrong here, and I owe those people I shut down an apology.

                  I have had very good hardware, both for my computer and for the servers I run, so for those of you I did shut down or mock that were attacking UT's input, I'm so very sorry.


                  We've had to run servers on this game at 120 tick just to compensate for the feel and it's honestly beyond a joke, and because of this servers that are running several blueprint mods/mutators are getting trashed because of how inefficiently certain things have to be done just to get what they want to work. If we had access to C++ (don't bother trying to say that you can manually share files, you know full well what a copout that is when trying to get exposure to something), we would be able to do what we want without having to constantly chase someone to make something blueprint readwrite/read/whatever. - even further, if we had access to C++, people like timbur could have fixed the netcode. Sure, it brings some security risks, but if you think that the players would say 'no' to fixing stuff that you don't seem to want to fix, you're sorely mistaken.


                  The fact that running a server with paks is several orders of magnitude slower than pakless (10x) is another major issue that is yet to be resolved, those of you on the dev team are well aware of how many times I've brought up major failures within how the pak system operates (massive memory leaks, recursive scanning being the 2 huge ones) - even now after the fixes, a pakless server will consume only 60MB of ram vs 300 for one with paks on it. wtf?!

                  We even had a redirect snafu (your code) that was causing rushbase to serve up over 30TB a month in custom data because everything was being downloaded twice - and it wasn't even fixed for the next **** build! I should have shut rushbase down back then, god forbid it happened to someone else who didn't have unlimited data transfer...


                  I guess what i'm trying to say is you've gone down the rabbit hole of features that have no gameplay significance and left stuff alone that actually makes players feel like they're being screwed. The game felt better back in november last year, cleaner to play, not so much **** in your face the whole time.

                  lets list it out here:

                  Netcode not working/completely unreliable (slightly more reliable if server is at 120hz)
                  Movement inconsistent
                  Movement incomplete (Still can't aircontrol when against a slope)
                  Servers crash every 3 hours because joe can't fix something, somehow killing itsself is a better choice...
                  Servers must run at 120hz to keep some resemblance of responsiveness
                  Content system is fundamentally flawed;
                  1) massive pak files consuming N*N for shared assets/small map modifications
                  2) causing excessive memory usage when it is not required
                  3) content system on servers is flawed (no dynamic add on hub start to ini files / no dynamic content update on map change or hub instance start) - ie, add CTF-Face2121 in UT99 and the server would recognize it during the next map load - why could an engine that's now likely 20 years old do that and this can't
                  4) pak files are inherently slower to use than without them due to some of the recursive functions in the engine
                  Seems to be a lack of any QA on the builds at the moment as there are flaws showing up everywhere and have been since november 2015
                  Things are added to the ui that seem to only be designed to add a massive amount of clutter/make it very hard to see things if you're doing well (picking an item up, have a double kill, got a killing spree for the double, etc).
                  There are more and more things being added to directly around the crosshair or being moved closer to it, normally the area of the screen where you should be able to see whats around it (you need a crosshair to aim guys, this isn't playstation 4)
                  Many things within the game have been setup so that changing them requires C++ - C++ cannot be shared dynamically, so changing things just doesn't happen, UI is a pretty good example here.

                  I'm sure there are a heap more but I simply don't play enough anymore to see the smaller ones that really annoy people, but I still do (was) supporting Wail with UnrealPrime so the server stuff is still extremely relevant.


                  On a more personal note, a ways back we were getting ready to push out an out-of-game server browser and we were in the middle of reversing MCP, we checked in with you guys and you asked us to stop - you advised us that the launcher was going to have the ability to have things like this added to it by us - This never happened.

                  If we had known this, we'd have just ignored the request to stop and continue to get an out-of-game browser running, pretty much simply for the reason that the time it takes to:

                  Open the launcher
                  Launch the game
                  Wait for game login
                  Load server list


                  is well over 90 seconds (seems to be especially bad if you're not close to the MCP source databases), these things just end up being put in the 'why bother' basket when picking a game to play, and considering that I can get from the battle.net launcher into a game in Overwatch in 15 seconds, you've got a problem here.



                  It's been a community project up until a certain point, but the thing is, it's been more or less a 'do stuff and maybe we'll talk about it', and I'm not perfect here, the people who I supported with servers/back end support/hosting/global access to players for their mods got a far, far bigger chance to get noticed than people who didn't come to me or who weren't willing to work with me to get around some truly stupid things that was built into the game (anyone remember when you had to download a file just to join my hubs? yeah..that was a thing required to get Speed to work.................... .... ... The game required the data to exist on the client BEFORE it would show it as 'available content' on the server. Good stuff.)


                  There are PR's that are still open, not commented on or brought in that do something relatively simple that speak volumes: https://github.com/EpicGames/UnrealTournament/pull/126 and another one with a useful feature: https://github.com/EpicGames/UnrealTournament/pull/145 "haven't got to it yet" cool, that was a year ago, maybe this year will be the year. Or maybe it might get closed.

                  Someone even built the entire DOM gamemode as a plugin (not as core code!) and it wasn't even COMMENTED on: https://github.com/EpicGames/UnrealTournament/pull/203

                  https://github.com/EpicGames/UnrealTournament/pull/218
                  no comments, not closed, not pulled

                  https://github.com/EpicGames/UnrealTournament/pull/264 same thing
                  https://github.com/EpicGames/UnrealTournament/pull/223 same thing

                  https://github.com/EpicGames/UnrealTournament/pull/284 dude fixes the editor for linux...makes PR, and you ignore it. for two weeks.

                  Perhaps maybe it's time to end the community game mantra going on here, there is never any discussion made about possible upcoming changes (look at the link gun.... wow.), there are very few changes made that originated from the community side these days (there could be none but I don't want to sift through 1300 commits to find one)

                  I still have some hope, but it's hard to maintain when we're where we are.

                  <3 raxxy.

                  edit: people have 165hz displays now as well, online play is currently limited to 150 as well. just a small gripe, but relevant.
                  Last edited by raxxy; 05-14-2016, 01:40 PM.

                  Comment


                    #10
                    Originally posted by raxxy View Post
                    Honestly I can't say I'm surprised, the issues with the netcode have been brought up so many times just to be kicked back with 'its fine at the moment'

                    It's clearly not fine. there are constant complaints, constant raging in-game over it, people are leaving, and it seems to be over ideology on the devs side.

                    Just recently another game - called Overwatch - was released/went through open beta, whose net code favors the shooter, this game feels so incredibly smooth to play even when your pings are approaching 200ms simply because the devs there feel that there is more anger in missing a shot you know connected vs being shot slightly around a corner, maybe sometimes. the game is extremely easy to play for 10 to 14 hour sittings nearly non-stop and it's simply because I have never felt like I'm being screwed over by the netcode (oh, and it's actually fun and flows very well which is a huge bonus). their servers even run at 20 tick and I have no noticeable problems.

                    Overwatch even has what I would classify as perfect input.


                    in UT, the input has been inconsistent for a very long time, seemingly framerate linked.. I used to dispose of what people said quite often back then but after going from Overwatch to UT back to back I can reliably say that something is very, very wrong here, and I owe those people I shut down an apology.

                    I have had very good hardware, both for my computer and for the servers I run, so for those of you I did shut down or mock that were attacking UT's input, I'm so very sorry.


                    We've had to run servers on this game at 120 tick just to compensate for the feel and it's honestly beyond a joke, and because of this servers that are running several blueprint mods/mutators are getting trashed because of how inefficiently certain things have to be done just to get what they want to work. If we had access to C++ (don't bother trying to say that you can manually share files, you know full well what a copout that is when trying to get exposure to something), we would be able to do what we want without having to constantly chase someone to make something blueprint readwrite/read/whatever. - even further, if we had access to C++, people like timbur could have fixed the netcode. Sure, it brings some security risks, but if you think that the players would say 'no' to fixing stuff that you don't seem to want to fix, you're sorely mistaken.


                    The fact that running a server with paks is several orders of magnitude slower than pakless (10x) is another major issue that is yet to be resolved, those of you on the dev team are well aware of how many times I've brought up major failures within how the pak system operates (massive memory leaks, recursive scanning being the 2 huge ones) - even now after the fixes, a pakless server will consume only 60MB of ram vs 300 for one with paks on it. wtf?!

                    We even had a redirect snafu (your code) that was causing rushbase to serve up over 30TB a month in custom data because everything was being downloaded twice - and it wasn't even fixed for the next **** build! I should have shut rushbase down back then, god forbid it happened to someone else who didn't have unlimited data transfer...


                    I guess what i'm trying to say is you've gone down the rabbit hole of features that have no gameplay significance and left stuff alone that actually makes players feel like they're being screwed. The game felt better back in november last year, cleaner to play, not so much **** in your face the whole time.

                    lets list it out here:

                    Netcode not working/completely unreliable (slightly more reliable if server is at 120hz)
                    Movement inconsistent
                    Movement incomplete (Still can't aircontrol when against a slope)
                    Servers crash every 3 hours because joe can't fix something, somehow killing itsself is a better choice...
                    Servers must run at 120hz to keep some resemblance of responsiveness
                    Content system is fundamentally flawed;
                    1) massive pak files consuming N*N for shared assets/small map modifications
                    2) causing excessive memory usage when it is not required
                    3) content system on servers is flawed (no dynamic add on hub start to ini files / no dynamic content update on map change or hub instance start) - ie, add CTF-Face2121 in UT99 and the server would recognize it during the next map load - why could an engine that's now likely 20 years old do that and this can't
                    4) pak files are inherently slower to use than without them due to some of the recursive functions in the engine
                    Seems to be a lack of any QA on the builds at the moment as there are flaws showing up everywhere and have been since november 2015
                    Things are added to the ui that seem to only be designed to add a massive amount of clutter/make it very hard to see things if you're doing well (picking an item up, have a double kill, got a killing spree for the double, etc).
                    There are more and more things being added to directly around the crosshair or being moved closer to it, normally the area of the screen where you should be able to see whats around it (you need a crosshair to aim guys, this isn't playstation 4)
                    Many things within the game have been setup so that changing them requires C++ - C++ cannot be shared dynamically, so changing things just doesn't happen, UI is a pretty good example here.

                    I'm sure there are a heap more but I simply don't play enough anymore to see the smaller ones that really annoy people, but I still do (was) supporting Wail with UnrealPrime so the server stuff is still extremely relevant.


                    On a more personal note, a ways back we were getting ready to push out an out-of-game server browser and we were in the middle of reversing MCP, we checked in with you guys and you asked us to stop - you advised us that the launcher was going to have the ability to have things like this added to it by us - This never happened.

                    If we had known this, we'd have just ignored the request to stop and continue to get an out-of-game browser running, pretty much simply for the reason that the time it takes to:

                    Open the launcher
                    Launch the game
                    Wait for game login
                    Load server list


                    is well over 90 seconds (seems to be especially bad if you're not close to the MCP source databases), these things just end up being put in the 'why bother' basket when picking a game to play, and considering that I can get from the battle.net launcher into a game in Overwatch in 15 seconds, you've got a problem here.



                    It's been a community project up until a certain point, but the thing is, it's been more or less a 'do stuff and maybe we'll talk about it', and I'm not perfect here, the people who I supported with servers/back end support/hosting/global access to players for their mods got a far, far bigger chance to get noticed than people who didn't come to me or who weren't willing to work with me to get around some truly stupid things that was built into the game (anyone remember when you had to download a file just to join my hubs? yeah..that was a thing required to get Speed to work.................... .... ... The game required the data to exist on the client BEFORE it would show it as 'available content' on the server. Good stuff.)


                    There are PR's that are still open, not commented on or brought in that do something relatively simple that speak volumes: https://github.com/EpicGames/UnrealTournament/pull/126 and another one with a useful feature: https://github.com/EpicGames/UnrealTournament/pull/145 "haven't got to it yet" cool, that was a year ago, maybe this year will be the year. Or maybe it might get closed.

                    Someone even built the entire DOM gamemode as a plugin (not as core code!) and it wasn't even COMMENTED on: https://github.com/EpicGames/UnrealTournament/pull/203

                    https://github.com/EpicGames/UnrealTournament/pull/218
                    no comments, not closed, not pulled

                    https://github.com/EpicGames/UnrealTournament/pull/264 same thing
                    https://github.com/EpicGames/UnrealTournament/pull/223 same thing

                    https://github.com/EpicGames/UnrealTournament/pull/284 dude fixes the editor for linux...makes PR, and you ignore it. for two weeks.

                    Perhaps maybe it's time to end the community game mantra going on here, there is never any discussion made about possible upcoming changes (look at the link gun.... wow.), there are very few changes made that originated from the community side these days (there could be none but I don't want to sift through 1300 commits to find one)

                    I still have some hope, but it's hard to maintain when we're where we are.

                    <3 raxxy.

                    edit: people have 165hz displays now as well, online play is currently limited to 150 as well. just a small gripe, but relevant.
                    Ditto...

                    something about 10 characters.
                    PayBack

                    Comment


                      #11
                      Somehow I feel this is the right place to add a few points to what raxxy has already (as I feel, correctly) pointed out.
                      My main issue is not with faulty iterations of netcode, gameplay mechanics or anything else (Although I must say that it often times feels like going 3 steps backwards rather than progressing.). My main issue is with Epic's lack of communication.

                      What's the current timeline? What are the priorities? What are you guys even working on? Do you realize whenever people are investing a lot of time, energy and in some cases even money to contribute to this community project?
                      What are the things you're currently working on? What can people expect to happen in near future?
                      When you are implementing a new feature or some change into the game, why did you implement that? (For example, new UI changes or new link gun beam graphics)
                      Answers to those questions are valuable because thats how the community can know what kind of feedback you are looking for.

                      But everytime I seek answers or any kind of clarity, I end up arduously browsing through past streams or forum posts that might or might not float around the forums. And even if I find some forum post by a developer or programmer, most of the time it's not gonna be official information.

                      "We are reading the forums." is a statement that is not sufficient for something that's called a community-project. I suspect that at some points Epic has given up communicating with the community altogether rather than finding ways to work with them, even if it's very difficult sometimes. I understand that it seems irrelevant or not really important to please a small crowd of say 200 people, and to a certain degree I agree with this. But if there was more clarity in your working process (especially in the programming side of things and especially when it comes to gameplay mechanics), there would definitely be less frustration and more progress.

                      Why the recent UI changes? Why did you think those are a good idea? Why the new link gun beam which seemingly only exists to obstruct the view? Why the new watermark on the lower right bottom?
                      Just by answering those questions you could avoid a lot of criticism and frustration. Because it's easier for people to disagree and move on than to be left completely clueless.

                      Anyway, I felt I wanted to add this whole point of "missing communication" to raxxy's list.

                      Comment


                        #12
                        Originally posted by raxxy View Post
                        <rant>
                        I have to agree. I'm sorry :/

                        The state of the netcode is beyond abysmal. I realize that it's still a work in progress, but knowing that it would be for now well over a year, it should have been put in as an option turned on by default from day one. We have been sitting around in IRC for hours upon hours, day, weeks, months, etc. on the UnrealPrime side trying to figure out how are we going to make this game better and attract more poeple and make playing custom content functional without tons of manual labor. Should we make our own launcher? Our own server browser? Replace all the weapons? Maps? Gamemodes? Movement? Blah, Blah, Blah. I can't tell you how many hours we discussed these things AND how much time we spent working on these things, too.

                        What did the discussion always go back to?

                        "It won't change anything."

                        I feel like I have to apologize to all the people I tried to ream about various issues in this game repeatedly as well. I was wrong about making this game as classic as possible. While classic game modes that didn't rely on modern features like classes and cosmetics would be a nice nod, UT is at a fork in the road. One path is "dead". The other is incorporate a bunch of features that the few hundred UT fanatics will absolutely hate to stay relevant. On the UnrealPrime team we cared a lot more about good input, fast action, and balance all things that are basically at "junk" status in UT right now. The input in particular with anything other than a G-Synced 144hz screen pushing 144fps consistently is horrid coming from so many other games, even other Unreal Engine games as sad as that is.

                        As much as Overwatch and soon Lawbreakers are allegedly different styles of game, this is where the fast arcade shooter is these days. Most of the classes in Overwatch feel lightning fast compared to UT. On top of that, overall movement crispness and responsiveness is so much better that it makes me feel bad to even think about it. If you think raxxy or I like the harsh reality of where this game is right now, we don't. I would love for UT to be better than Overwatch. Maybe in ten years it will be close.

                        I think it's time to just state the obvious. Epic needs help from community artists and it's nice to have UT out there as a sample of how heavily you can modify the engine, but Epic doesn't need the community to help them anymore with anything other than art. Shut down the master server, stop providing utterly infrequent builds that change very little or change all of the wrong things, and start giving community updates through Twitch and in blogs and keep the Github updated once a month or so.

                        I'm not lying when I say raxxy and I aren't going to undermine your plans for public builds anymore. We don't want to. We both agreed it was a mistake when we did it back in 2014.

                        Honestly, I believe this is the best thing the UT team can do at this point. There are so many problems with so many crucial systems (especially on the getting into a server front) that get maybe a cursory glance from build to build. The state the game is in right now is well within "oh, interesting" *uninstall* territory and doing more harm than good. Take a year to rethink this project and fix the foundation without releasing a single build. You know it needs to be done.

                        <3 brizz
                        Last edited by Sir_Brizz; 05-14-2016, 02:17 PM.
                        HABOUJI! Ouboudah! Batai d'va!
                        BeyondUnreal - Liandri Archives [An extensive repository of Unreal lore.] - Join us on IRC [irc.utchat.com - #beyondunreal]

                        Comment


                          #13
                          Ok so I don't want to be the cliché counterpoint guy but w/e.

                          This games community is it's biggest boon, and it's biggest problem; especially when it comes to communication. The cycle of demanding for, then complaining about a build is absurdly toxic and helps nobody. For example complaints that the watermark made the game literally unplayable are frankly ridiculous. Yes, it may be undesirable, off putting, or irritating but it does not render the game unplayable. The vitriol that some pour into their complaints is beyond unreasonable, and has resulted in some being banned. It's not surprising that the devs aren't raring to jump onto the forums and debate stuff in the open. It sucks and is a vicious cycle that helps nobody and keeps coming up from build to build; not to mention the almost daily occurrence of people poking RZE for new builds on IRC. If you expect the devs to communicate better, then we have to too. And by all means it isn't everyone. raxxy makes some good, and well put points.

                          On top of this it's clear that back in September/October last year there was a decision to focus on doing more and talking less, switching over to streaming twice a week for example. Personally I think this was for the best, the stream takes a chunk of time to set up, as well as to actually do, but it left us in the dark about a lot of things. There also seems to have been a shift in focus from maintenance, bug fixing, and larger scale improvements for prototyping (mostly gametypes and hud etc), again this leaves us somewhat in the dark about things like netcode and performance.

                          I see where Brizz is coming from, but I don't think I agree. The cats out of the bag now and personally I feel stuffing it back in looks worse than continuing. It certainly would feel to some like the last nail in the coffin of community development (even though it wouldn't actually be). I do however think they should stop entirely giving build estimates publicly; they're never correct and just make things look worse. On the flip side having a more public build objective would be great; i.e. say on stream what you're focusing on with the upcoming build, for example this one it was largely about getting 4.12 up and running.

                          None of this is easy and there's a lot of pressure on the devs to deliver; but communication is the key. If people know what's happening they're less likely to complain (in theory).

                          Anyway, tldr; this would be a lot easier if people calmed the hell down.

                          Comment


                            #14
                            I had really high hopes for UT, and once upon a time I believed that the game would eventually be good. But I no longer feel this way, and I know of many others who were unwaveringly hopeful up until a few months ago (at least), which is definitely not a good sign. Too much of the core architecture of the game (and possibly the engine?) is fundamentally flawed. For UT to stand any chance at a future, it needs to be remade from the ground up by a dedicated team who prioritizes not only player experience and gameplay above all else, but developer experience as well. It seems that priorities are backwards right now. When profits are prioritized over doing things the right way, this poorly cobbled together mess called Unreal Tournament is what happens. It might seem like a good idea in the short term, but it never works in the long term. And what's ironic is that if things were done right from the start (for both UT and the engine), all of this would eventually be wayyyyy more successful (and profitable) than it is now (i.e., not at all). That's how companies like Blizzard are able to make $1.7+ billion every year. They understand that doing things the right way from the very beginning yields the best results, regardless of how long it takes. They come up with a plan, put the necessary resources towards it, and they take their time to implement it properly, one step at a time, prioritizing gameplay and user experience above all else. No amount of hats, skins, or pretty graphics will keep people around if their experience isn't enjoyable. Blizzard didn't get to where they are by cutting corners. Blizzard's goal is clearly to create things that people want and enjoy, not to make a quick buck. Blizzard obviously understands that in the gaming industry, the business side of things comes after delivering a polished product. I mean, just look at the amount of time/effort/passion they put into Overwatch's sound design alone. It's incredible: http://www.gdcvault.com/play/1023317...sive-Goal-Play

                            Now, I'm not saying all of this to be inflammatory by any means. I would love for Epic (and the UT series) to be hugely successful, but the harsh reality is that things just don't seem to be headed in that direction, and the only way to change course is to speak the truth. There's a common pattern among technology companies where engineers take their time building some product that everyone wants which leads to the company being hugely successful. And then at some point, businessmen come in and make it all about short term profits. Next thing you know, the only thing keeping the company afloat is their initial success. And because of the fact that the MBAs (folks with business degrees) pressure everyone into cutting corners to make shareholders happy for the quarter, future products suffer so no one wants them, and thus, the company fades away into obscurity. If I had to guess, that's what's happening here at its core. So far everything seems to be a bunch of hacked together "features" thrown into one big pile. Yeah, we can call it #prealpha, but let's be honest here. There's only one thing really being accomplished right now: incurring technical debt and creating a ton of work for the future (which could have been completely avoided, mind you) because a lot of these "features" will need to be reimplemented. There are so many bugs occurring that should never have existed to begin with, all because of things being hacked together quickly and without any separation of concerns or long-term vision. The solution? Start over as soon as possible and put together a team of dedicated engineers/designers/artists who are passionate about properly implementing a clearly defined vision where player & developer experience is top priority. Don't let the idea of profits creep in until much later in the development process. When designed properly with concerns fully separated and features implemented incrementally (or in parallel where possible), it will be trivial to add whatever functionality necessary to generate revenue. And guess what? Since things were implemented correctly from the start, 1) it will be more successful because it was designed with user (i.e., customer) experience as a top priority, and 2) it will be much easier to improve and build upon in the future because you have relatively little technical debt. This is much easier said than done, but the end result is undoubtedly better and will come with a much higher degree of certainty.

                            Anyway, that's my $0.02 on the current situation, solely based on observations I've made over the past couple of years. This post will probably get buried and go completely unnoticed by the people who really need to see it. And who knows... maybe I'll even get banned for simply speaking my mind. But I feel like this needed to be said.
                            Creator of NewNet for UT99

                            Comment


                              #15
                              Originally posted by NATO_chrisjm View Post
                              ... If you expect the devs to communicate better, then we have to too. ...
                              I don't think that's absolutely necessary. As you have pointed out it's often times difficult or impossible to have public discussions about certain things. The information I feel is missing is no information that has to be conveyed via debates or discussions. It's simple information that the UT4-Developers should share with the community, in my opinion. Too often I'm left wondering "why they would do such a thing".

                              What's the reason for this new link gun beam? Why add even more clutter to the UI?

                              Why am I even playing if feedback seemingly gets ignored? What are they looking for, if not our feedback on how things feel and how things look? How can I help to give good feedback when I don't know what Epic's looking for? What's the point of some changes made.. whats the reasoning behind certain decisions?
                              Does it help when I voice my opinion about how something feels or looks, or is Epic just experimenting or doing god-knows-what?
                              What will happen with my custom content in near future? How should I code this or that so that it will fit Epic's plans?

                              The list goes on. For some reason they seem to have completely cut communicating these kind of things. Understandably so, given the behaviour of some people on the forums, but after all: that's something a developer needs to learn to ignore and move on.
                              Anyway, I don't want to hijack this thread further. I just wanted to voice my frustration about what I think of as lacking communication.
                              Last edited by rawlph; 05-14-2016, 03:40 PM.

                              Comment

                              Working...
                              X