As for the coding train, there's a few examples where I actually got directly inspired: the fluid simulation was a great one as I was always afraid that this might be too much work; this showed me I could actually do it in a day (but be vary that that his particular implementation has a few bugs).
If I look at one recent example, maybe raycasting is interesting. How would I make that look nice? Maybe I'll start by having a simple scene with a few regular walls but a circle of point sources that would rotate in a loop. Maybe I'll find a way to make the walls move in an interesting fashion. Or have the rays propagate through time and have a contest of flashing lights.
But what Daniel is good at is explaining some of the techniques used behind this type of work, it is not about the "how to make it look cool".
Now that I think of it, there's one thing that might be fun to do is to work though Casey Reas's Process series (linked here: http://reas.com/text ). It's a set of "elements" that have predefined behavior (but described as text) and a set of "rules" to represent how those elements interact. There's a lot of freeway since it's text; you might find something interesting by yourself here.
The point is that there is no right or wrong way to do things here, it's mostly a matter of "does it look good". That requires to think about all the points that make art look good, so you'll have to think about contrast, balance, composition and everything. But having something to show is half the battle/
The particular example in the PR is indeed based on the previous days before, there is a lot of overlap in the techniques used and it can be pretty opaque to understand but at its core it's a system that generates a tree by making elements grow/branch out according to some "energy" that is coming from the root (iirc). The other half of the issue is how to make this process look good.
As for the coding train, there's a few examples where I actually got directly inspired: the fluid simulation was a great one as I was always afraid that this might be too much work; this showed me I could actually do it in a day (but be vary that that his particular implementation has a few bugs).
If I look at one recent example, maybe raycasting is interesting. How would I make that look nice? Maybe I'll start by having a simple scene with a few regular walls but a circle of point sources that would rotate in a loop. Maybe I'll find a way to make the walls move in an interesting fashion. Or have the rays propagate through time and have a contest of flashing lights.
But what Daniel is good at is explaining some of the techniques used behind this type of work, it is not about the "how to make it look cool".
Now that I think of it, there's one thing that might be fun to do is to work though Casey Reas's Process series (linked here: http://reas.com/text ). It's a set of "elements" that have predefined behavior (but described as text) and a set of "rules" to represent how those elements interact. There's a lot of freeway since it's text; you might find something interesting by yourself here.
The point is that there is no right or wrong way to do things here, it's mostly a matter of "does it look good". That requires to think about all the points that make art look good, so you'll have to think about contrast, balance, composition and everything. But having something to show is half the battle/
The particular example in the PR is indeed based on the previous days before, there is a lot of overlap in the techniques used and it can be pretty opaque to understand but at its core it's a system that generates a tree by making elements grow/branch out according to some "energy" that is coming from the root (iirc). The other half of the issue is how to make this process look good.