Overall Game Perfomance state - and why GGG is in silenced mode about it?

"

UE's SSGI implementation based on advanced lightning probes (a set of cameras which produces cubic maps of lightning levels). They called it Lightmass. And the program which generates the probes - Light Swarm (precomputed lighting). It's completely different algorythm than Sasha proposed to use. GGG uses hierarchical radiance cascades. They have only the same technoloy name.

I suppose this is one of those features that the experts are very excited. They show you.
- You see! the borders of the shadow follow the normals of the stone! And you can see the reflection color of the wall that is why there is an increase of the saturation by 1%
- Players... ok... If you don't tell me I won't notice the difference. Can I play in my mobile phone?

"

Silverpelt wrote:

What's the hate with the Unreal Engine?

No hate. Dude proposed to use it instead of proprietary GGG engine not understanding the consequences of this decision.

I prefer original UT (1998).

"

Dxt44 wrote:

Tool like tool, can be used in good way or wrong way.

Game engine is not just a tool. It's infrastructure. You definetly don't understand what you're talking about.

"

B00b wrote:

I suppose this is one of those features that the experts are very excited. They show you.
- You see! the borders of the shadow follow the normals of the stone! And you can see the reflection color of the wall that is why there is an increase of the saturation by 1%
- Players... ok... If you don't tell me I won't notice the difference. Can I play in my mobile phone?

I got your point. Sure you won't see the difference on low res or Nintendo. But it targets the systems which will be made in 2030 or even 2040. It something fundamental, a core part of their engine which should be a scalable feature for long term use.

"

cursorTarget wrote:

No hate. Dude proposed to use it instead of proprietary GGG engine not understanding the consequences of this decision.

I never said that. And stop copy-paste my posts without context. It's not serious what you are doing.
And it's my last to post to you. I won't talk to anyone who can't read with understanding and made things up.

"

cursorTarget wrote:

https://dpaste.org/M7dsP

Why boast with such poorly optimized code? It's runs slower than my grandma, which isn't surprising when you advocate for throwing more hardware at unoptimized software. And passing in data directly as a format string, when you could just use %s or a simple puts call, SMH.

did you even run it?

My teacher would have hung me on the wall with a donkey hat if I wrote a small function with two returns wrapped in a if else like this one.

"

inline char Value(char c)
{
const char* p = strchr(alfa, c + 1);
​
if (p)
return (p - alfa);
else
return 0;
}

"

return p ? (p - alfa) : 0;

Wouldn't that work? I forgot I haven't coded seriously in a long time.

"

Silverpelt wrote:

Wouldn't that work? I forgot I haven't coded seriously in a long time.

Yes. Moreover it will work even without that pointless Value() function.

"

Have to agree with this one. Even Star Citizen is better optimized than PoE.
Performance went to dogtier when they introduced the new shaders around Ultimatum league.
One thing that makes the engine shit itself are corpses and items laying on the ground in juiced(& semi juiced) maps so they as a crutch implemented not showing items on the ground that are filtered by lootfilters (1-2 years ago)

I dislike to blame entirely to the tech team. I know by experience that new features are usually demanded by the CEOs who have only a background on marketing and sales, but they have seen a show in one conference of cool features and they demand the techs to incorporate, because they are cool.

While the project is small and your team is small everything works and communication is flawless and there are no much dependence on external libraries. If there is a problem anyone in the tech team knows how to fix it.

But as a project grow they rely more on third party tools they have little control. The team become more specialized on different parts. As the number of team members grow, more time is lost in unproductive meetings. At some point you have a broken system and no body exactly knows how fix it.

Maybe this is why Chris sold his shares. He saw this coming. And it is time to go back to the basis.

100 lines to print a statement that can be done in 1 line. Straight out of GGG's coding book

lol very cheeky

In my last project, we develop a simulator from scratch and was the fastest in the industry, it was done in plain ANSI C. Then, the managers wanted more capabilities, an interface, they didn't want something simple like FLTK, no, must be Qt for who knows.
Also, incorporate the development of other research teams to "unify platforms", managers might read in some book of strategies.
It became a behemoth of C, C++, fortran, .net and python scrips it took 8 hours to compile during the night and cross the fingers there is not a syntax error.
In that conditions development was so painfully hard and slow and it was always broken and something to fix. But that was a good thing. We intentionally broke it from time to time to justify our budget.