lordfroakie said:
it's never too late to start, and you'll never regret learning it while I can promise you'll regret if you don't.
And I feel like this kind of attitude (like when they basically tell students in college "Oh don't worry about efficiency, computers are so fast and compilers are so smart everything will be taken care of") results in the sort of bloated software or game performance problems that can show up today.
And I feel like this kind of attitude (like when they basically tell students in college "Oh don't worry about efficiency, computers are so fast and compilers are so smart everything will be taken care of") results in the sort of bloated software or game performance problems that can show up today.
Spending too much time optimizing too early is generally bad, certainly. I would say in particular that you want to find a way to identify a bottleneck by statistics, profiling, or whatnot, and then focus on optimizing that area.
And I feel like this kind of attitude (like when they basically tell students in college "Oh don't worry about efficiency, computers are so fast and compilers are so smart everything will be taken care of") results in the sort of bloated software or game performance problems that can show up today.
Spending too much time optimizing too early is generally bad, certainly. I would say in particular that you want to find a way to identify a bottleneck by statistics, profiling, or whatnot, and then focus on optimizing that area.
Sure, but I wanted to address the particular point that I bolded, and try to clarify that it's not like optimization in general is bad.
I recall even in the UDK I was getting performance improvements on things that most people seemed to just immediately say oh that's impossible, or oh don't do that with UDK, or oh you need a better computer.Makes sense to me. It's not true of everyone of course. For example, I seem to recall a presentation from Naughty Dog a while back about how they were able to do so well on the PS3 hardware, and (I think?) it was largely because they understood the hardware really well and optimized accordingly. BUT, again that's presumably after identifying bottlenecks. As Skinpop is pointing out, a few polygons might not make a difference, but other things might, so you don't want to spend a lot of time optimizing when it's not important.
Sure, but I wanted to address the particular point that I bolded, and try to clarify that it's not like optimization in general is bad.
Then one crazy person on IRC was like, "Hmm, maybe if you try this trick, and this trick, it could actually work" and bam, way fewer draw calls, and way better performance, and I was able to do the crazy test I was trying.
Sure but that's more of an engineering/programming/workflow kind of issue, as an artist you're gonna have set limitations and guidelines - there's not gonna be much impact in terms of optimization from your end.
I think Freenode has the main #unrealengine, #opengl, #blender, and maybe #c++ channels depending on what you're into.
-- Unreal 4 doesn't support custom lighting models so this might be harder to pull off in the same way.
Did you have to do anything special or did it just fix itself after being stuck eventually?Ugh... since this morning every time I boot up the Unreal Engine, there's an update that gets stuck at 81%. What the hell, now I can't load the engine and I wanted to try some stuff before starting work
Edit: Even if I try to boot up the engine via the 4.0/Engine/Binaries/Win64/UE4Editor.exe it actually goes to the update window and gets stuck there, so I can't get to the main portal window.
Edit: Got it to work. It updated fine now.
I just waited. Apparently they released a second patch fixing that issue (it was a server timeout that was leaving people waiting at 81% or so).
