Thread 106545360

>>106545360 (OP)
>optimization is bad
Some times this is true, there is a reason games pick 30fps everything else is cope, it's about optimization.

>>106545360 (OP)
So how do you then make good software if you aren't supposed to optimize, does the compiler do this?

It's always crazy to me how programmers give so many talks which are just basic shit you should learn in nearly any book you could pick up at random or any university course if you paid attention to anything, and then they make a name for themselves. This has to be the only field in STEM which actually has people regularly regurgitating the most basic stuff and making names for themselves as if it's a big intellectual revelation.

>>106545591
People don't read books, it's not the stone age anymore.

>>106545615
That's the problem, they'll be proudly semi-illiterate yet spend twice the time and effort watching youtube videos by someone whose parents unironically named him "Joe Blow" with straight faces, and in the end get half the payoff, then come to /g/ and ask why technology feels like it's going backwards.

Part of why I make games is so I can become a better programmer. Following Jon's logic, I shouldn't use an r-tree for static neighbor checking on a game server because just checking distance each event is "fast enough". Which is reasonable if your only goal was to make a game. But is it really?

Why should I ever listen to Jonathan Blow? Why are you so obsessed with him?

>>106545787
>Why should I ever listen to Jonathan Blow? Why are you so obsessed with him?
He is one of OUR lolcows.

He wants to be artsy fartsy but has absolutely plebeian taste and wants to pay people pocket change to work on his game while treating them like trasha nd living in the second most expensive place in Colorado.

I like Jon Blow, Terry Davis, and Tsoding mindset: Just don't overcomplicate your slop. Simple as.

>>106546094
do you actually need to attach names to such a basic concept?

>>106546231
No, but it makes you seethe, so I still won at the end.

>>106546237
you think you "win" what, exactly?

>>106545360 (OP)
His loss either way.

>>106545733
Joe is not short for Johnathan you dunce.

>>106545615
Tate pls go

>>106545360 (OP)
literally who

>>106545591
programmers are retarded and getting worse by the day

This guy is stupid as hell, mang

>>106545591
yeah you would never get away with having the absolute audacity to give a talk in science or engineering about actual baby-tier concepts like this
software "engineering" is the only field where people regularly do this and it makes the whole field look stupid as fuck by association

>>106545591
Nothing to do with the 2 minute video. Braindead post.

>>106545360 (OP)
>giving a single solitary shit about what jonathan "vulkan is bad because i am stupid and things should cater only to the most retarded of the population, even OpenGL is too much, the vendor published graphics APIs should be like raylib or SDL's builtin drawing library, having power and control over GPU hardware is bad and multithreaded SIMD is scary" blow says

>>106546727
John Blowjob and the video is absolutely an example of that, it's literally Knuth's quote about premature optimization but dragged out into a talk that just doesn't need to exist

>>106546769
>>106546969
Why don't you do it then? Go give an easy talk and make some easy money.

>>106547000
you don't get paid to give talks at conferences. it's the opposite, actually, you have to pay (on top of the attendee fee). what are you, 15?

>>106547027
Arguing in bad faith, I stop here.

> you don't get paid to give talks at conferences
shitty bait

>>106546094
what about ebassi

>>106545547
early optimization is bad

>>106547032
what part of that was bad faith? or incorrect? how is it you are able to breathe without grievous bodily injury?

>>106545615
I read books, am currently learning javascript with a book

>>106547034
depending on the conference the plenary speaker may recieve a small honorarium or get travel and lodging covered. all other speakers have to pay for the privilege. what the hell conferences are you going to?

>>106547078
use the durgasoft, saar

>>106547141
i make good javascript for good looking website saar

>>106547201
ahh very good saar very good

>>106545360 (OP)
>be extremely effective at getting things done
>makes custom programming language for a decade in order to make sokoban game
what did jon mean by this?

>>106545773
>But is it really?
So you admit that you don't even know which approach is better, so you default to the harder to implement, harder to understand approach, might not be a big deal for one system but if you approach every problem the same way defaulting to the most difficult approach because it might be faster then you are sinking a lot of time

>>106548642
>So you admit that you don't even know which approach is better
No, the question is referring to your goal

>>106545360 (OP)
>>106545591
Keeping things simple but performant is better advice. Minimum amount of features.

>>106548849
dont bother arguing with a woman, fren

>>106545360 (OP)
Braid looks like ass. Terrible color palette, terrible character designs. Belongs on a fucking Amiga.

>>106545360 (OP)
>muh life
I think anyone who ever gets into game development can tell you without doubt life fucking sucks and spending decades holed up reading about abstract transdimensional geometry beats being a failed normie every time. Is Blow even married?

>>106548925
>Braid looks like ass. Terrible color palette, terrible character designs.
Objective true but he still coded everything from zero. Jon certainly knows what he does.

For now, I am nowhere near his level, and I can just barely imagine the level of skill needed to program your own engine.

>>106547000
>Why don't you do it then? Go give an easy talk and make some easy money.
The correct reply to seeing degenerate conduct isn't to participate yourself and get on the grift, it's to condemn it as a grift.
When I've given seminars on things it's because actual new work was done that requires the time to present. I'd be embarrassed to give a talk that should've been an email.
>Hey guys, turns out you should only focus on performance once you've got everything else sorted!
A true revelation. I don't know what we'd have done without that presentation. Thank fuck we've got influencers to tell us to come and tell us all the blatantly obvious.

>>106549919
linear algebra and vector calculus. 19th century mathematics.

Anyone here used Jai? I saw his recent talk on it and I'm interested in how it is, especially the meta-programming.

>>106549919
It's really not that hard. The only technically impressive part of Braid is the time rewind feature and that's just an extension of what games like Bunny Must Die already had taken to the extreme of memory constraints for consoles at the time.
Other than that making your engine is pretty easy. Just boring.