@ChrisMcFarland I'm already at 2018 or 2017 I think

:)

Ah nice, should be easy then

should be ;)

I guess updating the game with smaller unity versions is the way to go? Only smaller things break and it's easier to fix?

yeah

Unity will periodically change stuff, and in the upgrade it rewrites code for you

if there's enough of that, things can break

also, its new code is often not very well-optimized.

we had problems with this on Love & Hip Hop.

@Almo Ah I've had that happen to me. Honestly it was kinda annoying that Unity re-wrote code for me, although it's much better than having something that doesn't compile.

Oh, yeah, I'd prefer to that it compiles and sorta work.

As long as you know where it changed stuff, you can go and fix it.

Most of Unity new versions don't even install on my machine.

:D

Has anyone experienced skipping a graphical frame when the game starts to lag?

e.g. try to keep the simulation running at say, 60hz, but skip a drawing the frame once in a while because the physics takes too long to compute?

@AlexandreVaillancourt Depending on the language/framework/engine, I've had some minor experience with this.

What seems to be the problem exactly?

Oh, well, we have different engines for rendering and for physics. We try to target a certain frame rate (e.g. 30 or 60hz). But currently, the physics and graphics are tied together: for one Gfx, we'll generate one simulation frame (physics).

I could add code to detect when a frame takes longer than it should have, and skip a graphics frame to make sure the physics run at the desired frame rate.

Say the user asks for 60hz, but in a situation, the graphics card is not good enough and takes too long to render the frame, I could detect that and render at 30hz for a few frames but keep the physics update at 60..

If in the settings, the user has chosen 60hz, I'd say stick with it, and optionally render at 30hz as you suggested. That seems the right way to go.

Mmm although, are the physics affecting the gameplay? Because in that case gameplay might change

I work on simulators, so yeah, physics is everything :)

Is the problem that you are not sure if that approach is the correct one?

I need a way to make my game more engaging... players can only craft a few items, harvest trees or chop them, kill a few static not moving mobs, build about 5 different structures and chat... Should i implement moving mobs as next ? Or a Quest System to give the players something to do ?

@TomTsagk I wonder what is the effect of a "smooth" simulation, but a frame not shown once in a while

@genaray You are working on an MMO?

Do you have super hard bosses that force players to play together?

@AlexandreVaillancourt Well... its not a mmo yet... more of a rpg :) And it will probably never reach that scale... its just a hobby project of myself... thats the problem, the mobs dont move... They actually only hit people in range when those are attacking... bosses require more mechanics first ^^ The game is really basic at the moment

And so your mobs are not really mobs :P

Lets say a player enters a new zone... and there 50 mobs around... and we arent talking about 1 player... there 20 players online at the same time... 20 x 50 = 1000 visible mobs... How to make them move ? I guess one single loop causes massive problems... on different threads ? Every second instead of every call ? ^^

ups... well... then my monsters are dummys instead of mobs ^^

Why are there 50 mobs per player? Isn't 50 mobs in the zone?

"Mobs" is a short for "mobile" :)

That comes from the MUD era, if I recall.

@AlexandreVaillancourt Thanks :D I picked this word up and thought it just symbolizes a monster or animal...

@genaray Mobs is the correct term to use, don't worry :) (I just found it ironic that they did not move, but the term originally implied that they did.)

Ok, so you have "instanced" zones; it's not multi player, it's just online?

Its multiplayer. If two players get in the same zone, they can see each other and see the same mobs :D But that doesnt happen this often... so we can just use the guideline : One Player One zone. I stored the zones and can "acess" every single one... so they are bascially instanced at runtime :)

Oh, ok, so the total 1k you got was from players in different zones. Makes sense!

Then you lower your tick rate to 20hz and put each zone on its own server :)

So i guess i would only put active zones on its own server ? :D Not 60k x 60k in total... right ? ^^ I never thought about this... Not even sure how this would work ^^

No, only active zones. You start an instance only when there is a player in there :)

I don't know either :)

No one tells you when you start to game dev that is going down an infinite rabbit hole.

@AlexandreVaillancourt Yeah i see... im doing my bachelor of computer science and i hoped i would learn about such techniques... fourth semester... tons of usefull and unusefull stuff... but not a single lecture about such topics :D

@genaray They don't teach you that at schools. (Most) Professors are interested in research and topics not related to what is done in the industry. School is good to broaden your horizon, and to teach you to learn and to work hard, but don't count on it to learn such advanced topics.

^ Most academic professors haven't worked in industry. It's not even so much a matter of caring about what industry does, but if you're not there you don't know what goes on in the trenches. And academia has plenty of other expectations that are expected to take priority over figuring out what happens in places that you aren't.

^

That's a better way to put it, yes.