@DMGregory thanks

We wanted to add some interaction to our game... because its on mobile we thought about a dynamic popup system. Each entity should have a different set of popups with different interaction options ( buttons ), the client only knows the different id's for the visuals and the server defines the structure and the logic behind those... the only problem here is the chaining of popups :

Trader NPC PopUp -> Click on Trade -> Trading Popup with text input -> Done -> Previous Trader NPC PopUp -> Close Trade... Any idea how we could chain those popups after another in scenarios where later popups define the outcome of previous ones ? We would be glad for any tipps we could get ^^

Time travel. Easy.

I dont think that this is time travel ^^ because both popups still exist till all are closed

Another question... how many systems in a entity component system are a good amount ? I have the feeling that i actually have too many systems in my ecs and i have no clue if this decreases the performance... For example the condition mechanic alone already consists of plenty of different systems all running over entities to check and insert different stuff.

Actually I didn't really understand your example. You can probably make a stack of popups or make a "Clear popups" message that you send when someone talks to an NPC because it doesn't make sense to have multiple NPC popups open (and you can still choose not to send the clear command should it become necessary).

Profile and benchmark.

I hope this is not C++.

Java ;) that are all current systems required for the "basics" of my game... moving around and showing stuff at the client, thats all... but it feels like a real game requires hundreds of systems processing different stuff... i didnt even implemented popups, combat and co and each of those require some more systems aswell :o

Currently it runs fine... using about 1% of my cpu... only gets processed 60 times per second. But i dont want to make the mistake to use too many systems

What's the alternative? The work that the systems do must be done, there is no way around that. The only way I can see many systems becoming an issue is that caching becomes worse if you keep iterating multiple times, but even that is a stretch. You might be able to organize it so that you iterate once and pass fitting entities to the appropriate systems, thus eliminating even that drawback.

I wouldn't worry about the too many systems issue. Then again I have little experience using an ECS.

Thanks a lot @nwp @Tyyppi_77 probably im just to worried... thats the 3rd time that i refactor my project... ( first time it was a mess, second time i was overusing oop and the flexability was missing ) and i dont want to get forced to change the entire structure again at some point... Currently theres no system that processes all entities, just spatial stuff ^^ @Tyyppi_77 Thanks, i think i get the point... probably some kind of stack is good to use to store the payload

I want to give a course some day.

@genaray I've never observed worrying about future refactoring to actually successfully prevent future refactoring.

So I'd say: trust yourself, make the best judgements you can with the knowledge you have now and the tests you're able to perform. If it later turns out you erred and you need to re-do something, yeah, that sucks, but that's game development.

Clearly you first need to fix their IT infrastructure and teach their staff programming before you can start to study.

@Tyyppi_77 What new will you be learning?

None. new is evil.

haha

@Tyyppi_77 hehe

@Tyyppi_77 woa that's annoying

@Tyyppi_77 I put film classes there

Took 4 of them.

one of the best professors I ever had. Took literature from him, then found out he did film so I took 4 from him

We learned that when he said in a lit class one day "Star Wars is a movie about Vietnam" when making the point that it doesn't matter what an author thinks his work is about, other subject important in his culture at the time may pop out. So the argument wasn't "Lucas set out to make a vietnam movie", rather it is "Star Wars was made in 1977, a time when the US was still very upset about what happened there.

A computer science major I knew was there, and called bullshit. The prof explained it. The guy switched majors to english literature.

Unfortunately, the prof died at 55 of Lou Gherig's disease. :(

I teach my game design students about this too. Meaning is constructed in collaboration between creator, audience, and their cultural context. Doubly so in games. How players play our games transforms what meanings that play experience communicates.

That professor sounds amazing! :D

yeah he was :)

Soo... what is the Mario Bros series of game really about?

Heteronormativity. 😛

:D

The famous 1951 western "Shane" had a lot of concerns about the atomic bomb.

I wrote a paper on that one, was pretty interesting.

@DMGregory Interesting!

Good GDC talk on the how of meaning in games: gdcvault.com/play/1014597/Dynamics-The-State-of-the

Reminds me that we never really got a strong consensus on how to tag questions about rule modifiers and buffs. If anyone would like to add onto that Meta question linked in the comments, I'd appreciate your input!

Can't vote more than I did :P

Well... I could, but that wouldn't reflect what I think, or that would be going around the rules of the system :P

😂 Yeah, just wanted to float it for other chatfolk to weigh in.

Huh - just noticed that I can't change my vote on meta unless the answer has been edited? In this case, something that seemed preferable to me 2 years ago no longer seems strikes me as the best fit, but I can't retract my vote? Probably a rationale for that, but it seems weird in this instance...

I think it's supposed to discourage revenge-voting.

You can edit the answer yourself and then change your vote.

@Pikalek You can disclose your vote and tell the world what's your new desire in the comments ;)

Both of the strats would work - maybe it already reflects my opinion though - either solution is probably better than nothing.

I could also go edit both answers and give you a chance to adjust if you need. ;)

There, found something I was less than satisfied with in both, so they're both freshly edited if you'd like to adjust your votes.

Here we go again

Guess this is getting my daily routine now... our last lecture was about light models in opengl and how you calculate ambient, diffuse and spectular light. I understand the math behind it...

But what happens with those calculated light informations ? Lets say we calculate the ambient light, basically the vertex color + our ambient color... How does this get applied to our vertex ?

Does a shader calculate the light for each vertex or for each pixel ? Lets say we wanna calculate the diffuse lightning instead, we basically end up with a bunch of calculated vectors and their color... And now ? Does opengl project those vectors to the surface of our object internal ? I just cant find any informations about the link between calculation and actuall rendering of that shader light.. anyone having experience with that topic ?