pouët.net

:D

listen to the norwegians on transitions. they have oil ffs!

xteraco: usually overdesigning your system is a bad idea, especially at the beginning when you don't have too much experience. Most of the times you'll end up with something that took you 10 times more than doing it simpler or even hardcoding it. Then after using it a couple of times and facing the real problems you'll want to start again from scratch.

I'd rather make some very simple sequencing system first, and then focus on the effects, release productions and finally get more experience.
Also don't try to make the ultimate flexible demosystem, there's no such thing. You don't need it anyway.

i support ps. you should listen to us - we know how to fla... i mean transition.

we know how to HYPNOGLOW!

@ ithaqua

was wondering in what area i'm overdesigning? its hard to see from my point of view, but i'm definately open for advice...

I'm not really going for the ultimate flexible demo system.. but I am trying to go for something modular, and re-usable... i'm thinkin i'll add a few transitions, a few screen filters, and then make a few demo's, then toss the whole thing... I'm thinking about nameing the engine Kia Rio, the throw away engine.... lol

As far as sequancing goes, i've got something going on in the code for the sequancing... not sure i'm happy with it, but it works alright....

thats overdesigning. ;)

I'm afraid so :)

depends on his definition of modular.. a basic reusable demo engine + 'disposable' typicalities (mainly fx i assume) can be a modular approach as well. and possibly not a stupid one :)

the stupidest assumption you'll ever make is that you will not have to hack your engine to pieces while making your next demo.

(and you make it again and again no matter what)

Modular design: erase the parts of the code you don't want, code new parts using the same functions and global vars names ;)

You can check for a simple and effective (half hard-coded) way to do scene handling / transition here : http://www.pouet.net/prod.php?which=33 (source).

nice!

xteraco, don't feel bad if you disagree about the overdesign drill, even the best coders took ages to figure this one out (compare last_engine_ever.exe and very_last_engine_ever.exe, etcetera)

an exception goes for jumalauta penis, by the way, which has such excellent design that it's proven reusable and super-flexible indefinitely.

Skrebbel; your example (last_engine_ever.exe etc) in the same sentence as "the best coders" has to be a mistake on your behalf?

And about syncing, this thread has some interesting insights: http://pouet.net/topic.php?which=2542&page=1

most of the stuff you code "prospectively" wont actually work when you need to use it for real.
my advice - develop "on the job" - start making a demo and develop what you need for that. then make another demo, see what you can save from the last one and make the other new stuff you need for it.
eventually you'll see a pattern in what you need to save, and stick it in an "engine". :) and you'll see what jobs you do all the time and make a "tool" to do them. :)
after all, it's only a demo, it's not going to run a nuclear power station.

I just can second smash (and alot of others as well). it"s pretty unrealistic to have an overview of what you need for your demo before your actually code that demo (except for the very basic stuff, but even here you might be doing false design-decisions first).
the "this stuff finally should be put in a class/method/package for reuse'-method did quite well for me so far :)
still didn't put it in a demotool though ;)

slummy, it was a typo. i meant "best north-norwegian graffiti-style graphicians who also code"

I can see what you guys mean about over designing... i ended up practically re-doing my timer because i thought it would be more useful with floats and such... wrong ... but hey, works perfect now, so cant complain about that...

i'm not sure i have the best design style... but, i kind of code a little, mess with it, see what the new problems are, solve and start over... its like a loop. but to keep up with the blog, i might as well mention, the latest problem is, the 3d engine hangs when textures are being loaded, so i need a loader... I've got a pretty good plan though i think.... super simple, all i have to do is re-structure the entire code :P ... just kiddin... welp, i better stop bloggin before i get a "shut the fuck up" hehe

i think, overdesigned or not, it all ends up in the fact that it's probably your 3rd or 4th engine that will be really good anyway. at least, that's what i seem to see.

nothing personal... you asked for it :)

Gargaj: I'm probably somewhere around "engine" #11-12, and I still constantly change my current engine. And by the way, the difference between "last_engine_ever.exe" and "very_last_engine_ever.exe" is more or less a cleanup (ie thrown away things that didn't work out that well) and shader-support. It just makes it easier for me to differentiate between generations, I guess.

kusma: my point was that around engine #3-#4 you kinda reach a level where you finally have a good idea about what an engine should be about... of course, changing trends might add some more steps to it.

engines schmengines! *runs*