Greater HP vs. More Enemies

[LannFyre]

I'm having an issue in balancing combat in my first test map. On one hand I want to place at least 2-5 enemies in certain areas, but I'm finding that I will hit some lagging framerates very quickly (3 commandos and I'm getting lag). So on the other I'm thinking of reducing enemy counts to 1-3 and just jacking up their HP, but then I've read on MH threads for UT99 that people reeeeaaallly don't like bullet sponges. Not to say I'd bullet sponge any basic enemies, maybe just change their HP by a degree of like, 1.5-1.8 (so if they were 100, 150 would work).

So before I continue working, what would you suggest?

I should also point out that the weapon the player is using fires projectiles that scale off of stats they carry.
> Base damage is 20 + whatever their strength stat is multiplied by two;
> base strength = 21;
> at level 5 = 26;
> so 20 + 26*2 = 72 damage per shot, fires as quickly as you can press fire, very small splash damage, very fast ammo regen but slow enough to prevent spam.

[gopostal]

3 commandos and I'm getting lag

That's the problem you need to address. Gameplay first, everything else second. It doesn't matter if it's gorgeous like some of the lush jungle MH maps, if the frames stutter it ruins the entire thing.

Also I used to play a ton of Invasion RPG in UT2004. In fact my first online mods were monsters and artifacts for my server. In Druid's RPG stuff there is a lot of weapon damage/fire rate adjustments tied to leveling. Player's almost entirely either ignored it or disliked it. We as coders tend to like the mathematics of adjustments because it plays into our interests. Players just don't. If your rifle does 22 damage then suddenly does more or less there is a subset of players that will notice that and will not like that things have changed. These also tend to be the careful players who will understand the level system and care about it, not the charge-in-and-kill-everything guys. Players care about gaining levels (my e-pen!s is bigger) but really don't care if their 'stats' do much of anything. If you script a leaderboard for your server then that will give you almost everything you need that will truly interest Joe Player.

(Aside: we can't use the word 'peni$'? Really, that's off the table?) 

[LannFyre]

Honestly, I don't even intend on exploring what all I can do with stats until I can work on models and junk (different combat styles and junk that would make stats actually matter).  But I see what you're saying, I'm trying to strike a balance in the middle where the player wont feel worse for their choices, but instead choose classes based on their play style, stats only making a pawn play more to the play style of the player.

. . .  Or is that what you guys already tried?  D:

Also apparently I can have more than 3 enemies in a room, but they all have to be different pawns (Brute, Commando, Skaarj Scout, and a Slith).  Anyone know why this is?

edit:
Think I'll drop the projectile-stats connection for now, see how that goes.

[gopostal]

Something is really wrong if you are lagging with so few enemies. I haven't looked at your code but I'll bet you a nalifruit smoothie you have tick functions that are flooding your execution. If you don't see right off what's up you are welcome to email me your source and I'll take a look with fresh eyes (then delete it ofc).

[LannFyre]

I might have to do that. While I have tick functions running, I try to make them shut off asap. . . But I really gotta minimize them. Yeah I'll e-mail what I have, I'm pretty sure I see a spot or two where I can cut down.

[gopostal]

I took a look and not much stood out to point at directly. Yeah, there's a lot of camera update stuff but I'd be surprised if that's the problem. Why don't you send me the map too, let me run it here and see if I also have the same spot lag. I'd like to see the log anyway. As far as I can tell it shouldn't be having an issue when enemies are there.

BTW, very nice coding format. You completely relaxed my OCD and that's saying something.

EDIT: I think I see part of the problem. It's the way you are saving stats. It's too often. You are running DataHandler 3 times a second in the mutator. You really need to find a way to do this either when called (via some player controlled event like 'save') or design some logical game stops where this will occur (like a door frame or section divider). Running it all the time like you are isn't a lot on the engine but the writing to ini is what's so slow.

[LannFyre]

Check your inbox, just sent links.

[gopostal]

Got it. Have to work today so it will be in the morning before I can fire up a test.

[LannFyre]

Hey mate, when you get a chance to look at the map just E-mail your response or PM me either here or on UT99 so that way we aren't constantly bumping this thread. I'll just edit this post and post the results of why the map was lagging so badly. Ironically this post will bump the thread, anyone else viewing this ignore this bump post pls, it'll be edited when this issue is resolved. I wasn't expecting the thread to get bumped quite as much as it did. Still not sure what the issue is.

As a side, you may have noticed in the map I sent you there is an area that has like, 10 enemies. Play the map with God mode activated and either kill them or watch them kill each other and observe how much the lag scales down. If you want to pop the map into the editor and change the amount of enemies down to about 3-4, watch how much your frame rate STILL dips. I've noticed that even if there are only 3-4 enemies in the biggest room on the map (you'll know what I mean), there is a massive frame spike that knocks the framerate down to about the 30fps range. There are maps in Unreal that have rooms bigger than that (I think) yet you don't see anywhere near as ludicrous of a frame spike.

[gopostal]

Honestly it'll likely be Friday before I even open the map. It's voting time in Oregon and we are a ballot-by-mail state so it's a little crazy for me. I start a week vaca on Friday though so for sure I'll be on it then.

You are always welcome to call me. I'll answer if I'm not on the route or running.

[LannFyre]

Hey mate, had a chance to look into the map yet?

[gopostal]

Sorry, it was my daughter's prom last night. It consumed our weekend.

So I loaded the map and the associated files and ran some tests. It's super laggy when you see enemies like you said but you'll need to do some stuff to narrow down the cause. My last test I was in-game about 25 seconds and the log was already half a meg. You ought to add a debug option to all the logging so you can isolate it from causing any of the stutter. That's a lot of logfile creation going on, it's bound to be causing a small hit.

There are smaller but important error lines like this

Code: Select all

Log: RPG_Gunner RPG_E1M1.RPG_Gunner0 (Function RPG_Game_dev.RPG_ClassSystem.PlayFiring:0083) PlayAnim: No mesh

and

Code: Select all

Log: RPG_Gunner RPG_E1M1.RPG_Gunner0 (State RPG_Game_dev.RPG_ClassSystem.NormalFire:000B) LoopAnim: No mesh

You'll need to fix those. They repeat over and over. Also I see this a lot:

Code: Select all

ScriptLog: RPG_Camera: Error, Return.

Is this something else that needs cleaned up?

Now my personal thoughts concerning gameplay: Sadly I really dislike it. I can't control the damn guy and it's really frustrating. I get you are going for a 2D/3D hybrid but it's clunky and aggravating (to me). Then again, I'm a coelacanth. Younger kids can probably adapt to this and do very well. It's also a negative that I had to either swap out your ini's to make it work or make a bunch of changes to mine. I ended up just hot swapping yours for the test.

Thought concerning the mod: It's pretty interesting as a concept. You've done some neat work on the camera controls and it has intriguing potential. It's just not at all my cup of tea but that's fine. We each like what we like.

When you get your build edited to fix the above stuff let me know. I want to take another test run and see how much that helps. I think there's a lot more that needs some attention but I think it's mostly going to be code optimizing and sanity checks to exclude dead ends in the code. That stuttering can be fixed for sure.

Hope that helps and thanks for letting me give this a spin. It's always cool to see unique efforts like this. You've for sure went all out in trying something new.

[LannFyre]

I hear you on the gameplay myself, honestly this isn't even 10% near where I want it to be. In fact, the combat isn't at all what I am looking to create but is instead a placeholder to see what can be done with character movement. I'd like to work on something that uses the full axis of a joystick for turning to help fix aiming/movement issues as well as clean up a lot of the camera angles, but right now literally everything is in an experimental phase. I would much rather be working on the rpg/fighter aspect, but what you've seen now is me getting my feet wet with weapon coding and to see how AI works with certain variables.

I'll send the new code likely some time in the next week or so, been applying what I could to get into Java (and thereby Android) coding so I'll have my attention somewhat divided.

I do have a question though: in what ways were the player difficult to control for you? Any specifics you could draw out? I already know the directional movement/shooting is a big issue, since movement and direction changing are tied to the same buttons now.

As a side, honestly not too sure what I can do about the .ini or if anything could be done about it. Likely some console stuff I could do but I'm at a loss on that.

[gopostal]

I'm not dumping on your work, this is just pure criticism without any malicious intent. I always think it's stellar when someone does something new and I want to support that. I also don't want you to do a 'Survival'* on yourself if I can help save you from that because that's where I think this is heading.

About controlling the player...It's likely that it's one of those things that needs time to get used to. You picked an odd angle to use as the home, it's above and offset. Once you said joystick though I can immediately see that using one of those would be superior to keyboard. This is a controller game 100%. Keyboard is the clunky part. I'm going to dig out my stick later today and see if I can install it and try that out. My guess is that will fix most of the weird feel.

About the ini...I can almost promise you that this will be the kiss of death if you don't find a way around it. Too much has to be remapped to work right. I'm not sure this could be compatible with a normal installation because of the changes you need to make to default movement keys.



*- Years back Dane and I spent at least a thousand hours building a new gametype called Survival. It was rounds-based like Invasion but we added stats, did all the waves, created a ton of backend stuff. It had custom everything (including a ton of maps) but we made a hugely critical mistake. As coders we spent WAY too much time making things work and doing the nuts-and-bolts coding and spent almost no time playtesting the fun factor of the game. It works very, very well as a unique gametype but frankly it sucks as far as fun is concerned. Because of this it got nobody to use it more than a few rounds of playtesting. Feedback was....humbling. Big lesson learned.

[LannFyre]

Ah, that makes a lot of sense.  Yeah, I should point out even on my project page that this mod is almost entirely being made to be used with a controller or joystick, more specifically Xbox or Playstation controllers.  After we get this lag issue figured out, I'd like to post the first map to get some more input from other players as it would really help with the future direction of the project.

I wonder if it is at all possible to check keybinding slots through script and fill in empty keybinds with the control binds needed to play. . .  I haven't read anything like that though.  I'll be sure to ask around when I have time.

EDIT:
Also don't worry, I can tell between incendiary comments and potentially mod-saving advice.  I would never shut down good advice. 

EDIT 2:
I removed all of the log b.s. but I am still getting some lag. I'm doing some quick bug hunting right now, fixed some issues I was getting from animation calls in the weapons. On top of that, the room with the ramp I think is going to get edited severely as I've noticed that it really breaks the flow of the map so I'll be sending a new version of the map as well as the updated code.