pouët.net

Hi everyone,

i'm doing some work that isn't REALLY a demo but its the same mindset. I want small code for my game, and I want it to run very efficiently on my current hardware.

The art-style for my game is very... "cybery". Now, I have a bit of an issue. To make it look like "the way I want" is HARD, for me. I'm not like inigo quilez or other super-coders... graphics is hard for me. There's nothing I can't understand but its a struggle. I can spend months trying to get a simple part of my code working. Meanwhile I see other people make amazing things in a few days.

So I want a cybery game, to run efficiently, but its hard for me.

I'm thinking... maybe part of the reason Its hard for me, is IM DOING EVERYTHING ALONE? Is that it? Maybe someone else knows? I had some bad experiences with the development community (not here... it was full of very stiff and incompetant programmers in business suits, and I hate business suits)... so after that I learnt I was better off to avoid everyone... and just do my coding alone. I really was.

But maybe it's gone too far? I had some (many?) things better coding alone... but maybe the graphics stuff ISN'T it. Maybe I SHOULD try to share a new technique with people?

Maybe that's the problem? Like my heart (or whatever part of my body that is feeling stressed) really wants to tell me that "You can't just make a major new technique and not let the world know about it."...

I don't know. I don't know how these things work. Life is complex.

Anyhow... if anyone wants to take a look at pictures of my "Cybery technique"... let me know... I can grab upload them.

...

The problem I'm having, is that it runs too slow! It needs to be faster. It looks good, but it just needs to be FASTER.

Making it faster is stressing me out.

That's where I want help. Big halp needed plz :`(

How much experience do you have? Everything is tough for the first time, but it's growing pains.

Definitely avoid the people in business suits.

there are books and blog posts on optimization which you can pick up. depends exactly on what you're doing ofcourse.

but yeah i would say in general sharing what you're doing and asking for feedback is the best way to learn new tricks.

if you're doing it on pouet i would suggest you make sure it's a summarized and straight to the point post, as opposed to an uncomprehensible wall of text no one will want to read or have difficulty understanding.

and keep an open mind when people start dissecting it and asking questions that challenge the whole approach. if you post on pouet you'll likely get a couple people giving explicit feedback and optimization tricks on how to improve or rework it from scratch, a few people telling you it's been done before and you just didn't notice, and a couple trolls making references to things that don't really matter to the case at hand. if you focus on thanking and answering the first two types and ignore the rest you should get some decent value for the exposure out of that.

I've made my own compiler, language, and some major new advances behind this kind of thing. And a lot more.

In terms of graphics... I've got the shell of a very basic 3d game, with very cool postprocessing effects, but... it doesn't even have hit-detection yet.

So im good in one area, and bad in another area. The old area kind of... well the "business suit" people ruined that for me. There is no future for me there...

People work alone too much...

Hello Wizz and sceners,

my name is cosmic and i´m a 3d artist.

from my point of view and concerning this thread (among others):

http://www.pouet.net/topic.php?which=11027&page=1

i don´t want to troll, or offend people here but i am following
these discussions for a long time and i just have to add my five cents.
i want you to think about it and maybe it helps.

(from a 3d artist view and how i would approach things)

I have no clue about how to code. i did some assembly coding
in the 90ies on amiga but what it means these days - i have no
clue how much work that really is.

As is said, this is all my personal opinion and how i would try
to „get things done“.

Learn 3D: Blender, zbrush, maya, max whatever you like.

make your first camera movements, learn about photography (everyone
- me excluded has a cell phone in his pocket) and learn about staging
and cinematography and shadows & lighting. traditional principles.

THEN / or parallel

try to get this to „realtime“ somehow - the easiest way is to press the p
button in blender and enter the game engine.

ok, i dont want to start another discussion about 3d engines or the own
coded engines. i want to help to get to a result quick and without a frustration
time that all you coders went trough.

again: i have the deepest respect of what you guys do with only a few bytes.
and when you know or learn with a engine how this works - nothing´s wrong
with coding your own one right ?

but as a beginner you want to have results and not only frustration right ?

when you start coding (whatever that means - i did some coding / and tried
some lua scripting on the pico (as a 3d artist ;)

here and here:

http://www.lexaloffle.com/bbs/?search=first+intro&cat=7&sub=2&mode=carts
https://twitter.com/CosmicOllie/status/808744193367674881

so, if you are into 3d - no matter what level.

then why don´t you ask yourself: what do i need to learn to get this in some
kind of a 3d engine, how could i find a way to get all the lighting baked / or
for real and all the models and the camer stuff to work on my gfx card.

THEN ! guys please think about it. there would be a goal. and lots of motivation.
and i´m sure all the experienced coders here will help you find the right way
or even point you in the direction where you have to search for on the web
and what you have to learn from coding to reach your goal.

the question how things get then packed into a 4k - during this time you learned
a lot right ? would be the next step.

what do you think ? only my 5 cents. maybe this is an „easier“ - no - more productive
and more fun way than starting the „normal“ way.

all the best

yours cosmic

This could be my problem.

Although I think the real "Thrust" of my message probably would have been better understood.... if I directly sent it to someone who has created new cool techniques and shared it with everyone...

can be wierd or awkward doing that kind of thing though. These people normally prefer to talk about very... concrete things... than the emotional side behind it, and any kind of deeper life purpose talk.

I'm not sure...

If anyone ends up taking my message seriously, we could see a new style of demoscene prod emerge (So 101 styles instead of 100 styles... just +1 style, but every new style is cool!!)

A very cool... magical and cybery style. That sort of.... sucks you in... softly.

Trust me, you have not seen anything like this postprocessing technique. It really turns boring images into something special.

Just needs opts to make it worth it.

by "taking serious" i mean if someone chose to help me optimise this work... then of course I would share the technique with everyone. Depends if anyone wants to learn/help this.

I'd like that, because then I'd be able to see my technique used by many people, and not just me, and it looks so nice, so I'd have more cool stuff to see!

But who knows? Maybe its just not the right thing. Life is wierd that way. I don't know the rules of life, I just... bump into walls I didn't know were there.

Well, there´s an easy approach to this i´d like to advice to you:

As you say you have some nice new postFX going, i guess this is shaderBased...
...so you should simply go to Shadertoy, make an account, redo your postFX there (it is able to have several passes by now! yay), publish the Shader and ask for help getting this faster in the Description (or even inside the code with comments at the lines you think need better performance!).

You´ll get help there for sure...and if it really is as nice as you say you´ll collect some hearts aswell! ;)

If you do so, please don´t forget to come back here and post a link to your nifty PostFX-Shader!
We´d like to see it, you made us curious now! ;)

Another biggie is the NIH (not invented here) syndrome and the habit of many coder-coders to reinvent the wheel constantly while running in circles and mad-scientisting around all the time with little to no focus on their actual goal. And because they are doing it all alone they lack a counter-balancing voice that would moderate said madness.

Wait till the emotional burden turns into the "fuck there might be someone who's already patented that" burden :P

@whizz: I always forced myself to work too much when I was trying to please the demo scene. Too much coding, too few releases etc.

Yup and one I haven't seen mentioned. Thinking SO hard about optimising, that you never get anything going..


If you want a development mindset related reboot - read this book. It's changed my entire outlook and it can be applied to anything. You can pick up and English language copy dirt cheap, other langs may vary.

https://www.amazon.co.uk/Lean-Startup-Innovation-Successful-Businesses/dp/0670921602

I'll get banned for this, but the truth is that pouet also makes you lonely. People don't even have email addresses on here. Example: wizzz.

If you have some cool new tech, post a screenshot at least. If it's cool new tech, people will be interested, and you might find people to work with. As it is, it sounds like the same thing we've heard a ton of times before, that turned out to be stuff people already did, or stuff people know about but don't do because it's not practical for whatever reason, or just some mad idea that'll never actually work :)

Speaking of emotional burden though... I've been doing some realtime path tracing stuff (examples on shadertoy). I find that a weird mix of pure joy (because this stuff is challenging but looks soooo good when it works) and depressing.

Why depressing? Because for so many years, a big part of the pleasure of the scene for me has been watching the progression year after year. The new demos with new techniques that blow away the stuff released just a year before.

Path tracing is going to do that again as we get our heads round the challenges, sure. But... what about the year after that? I don't see much beyond path tracing :/

on the emotional part, the way i see it, is that you had a bad experience once or twice and then you ended up believing that things will go wrong every time you coded something seemingly resource demanding. i'd like to put it that way, think of it like this, all these super coders you have in mind probably failed around say 20 times and after this they succeeded the next say 5. you probably have seen only these fives times and you are like i will never be like them. it is all in your mind. try do something and if you 'fail' don't give up. take a step back and think 'what change could i do to make this work better?' think of the change then apply it, see if it works, and if doesn't then try again and so on. never say i cannot be like xyz guy. to me it is all a matter of psychology and attitude towards your work. also sharing your code with people who are positive to help you out with it and not just criticize it for the sake of criticizing it, is going to help you make the appropriate changes to your code and succeed.

You are in the good path. Don't desperate. Creation requires time. A lot of time. Especially when you are doing all alone. Just continue what you are doing and you will improve each day more. I can only advice you a thing... avoid perfectionism.

Gonna have to back out of this discussion for the moment. Thanks for all the replies, even if I don't agree with all. I know a lot more than you might realise...

A few responses though:

ham: Thanks... yeah that's what I've found. The answers, or best results, always just came when I followed my own path...

hardy: shadertoy is so cool, but I don't have the time to put my stuff there. I hope however someone else can help me put it there.

My effect is not some ground-breaking thing, its just like "my personal style that I really like".

Everyone needs their own style... thats the most important thing. A "Signiture technique".

Sorry for anything I didn't reply to. I have to make my own decisions.

(and it's wizzz first post on pouet ever)

Not really my business, but something is fishy here.

Previously I was a co-founder startup where I had to learn pretty much everything, and I was the only coder as well for most of it's life. So my current job feels really boring and repetitive in comparison. I'm thinking about quitting and finding something else.

TLDR; hard is fun. easy is boring.

@wizzz: please post pics & video