Wednesday, February 1, 2012

4x4

So let's see if we can give this initial screen a go. And let's not focus on the artsy bits. But just the sheer html 4x4 layout. Recall our baby.
4x4 (approximately)
html is easy to write. So is javascript. The harder part is getting it to work. Especially on a wide variety of devices and platforms. Let's start by getting things up and running on one browser on one platform. And then hope for the best. And check it. And rinse and repeat.
The nicest html document that one can write, is, according to the syntax validator So let's try to get 4x4 divs into this html.
...
Hmm, I stopped writing this blog entry a few days ago and instead starting coding the thing. Sorry. Try a version at http://killard.herokuapp.com. Kill lard by clicking. Inspect it by double clicking.
It should work somewhat on mobile phones and tablets. It uses webgl, when available.
A lot of extensions could be done. Add additional formulas. Make it work better on touch devices. Focus on creating beautiful formulas. And a lot more.

Generation 0

A few clicks later

A couple of more clicks
I've included a couple of screenshots as proof of concept. I had fun.

Friday, January 6, 2012

killard

The first project of the month is killard. The name is a wordy play on kill, art and lard. Enough said.

The basic idea is an html app. That uses genetic programming and evolutionary algorithms to produce some hopefully aesthetically pleasing images. One of the pioneers of such imagery is Karl Sims. Nowadays SBArt can mimic this in real time. This real timeness is possible today as a consequence of the massive parallel capabilities of modern graphics cards.
killard will try to mimic the images of Karl Sims using the techniques of SBArt in an html app. Hope is that this app will be executable on standard browsers on standard hardware, eventually including the modern touchy feely mobile phones and tablets

This means we're probably gonna explore javascript, html, WebGL, mobile, ...
Let's look at my current mental picture of killard.

Initial Screen
Random images
A few ugly images have
been removed by the user
A few new images occur
spontaneously












That's it. Quite the simple app. The images are generated by some "formulas". The user "kills" the images he doesn't like. New images are generated from the "alive" ones by means of recombination and mutation.

One could imagine all sorts of nice interactions for killing, choosing, recombining. Later that is.

Hopefully nice images can be generated.

Thursday, January 5, 2012

habits suit me

Or that is actually just a pun on the danish word "habit". But what is hopefully going to happen is that I will get a small coding project done each month in the year of two thousand and twelve.

I have a couple of ideas for the first couple of months. After that I will just have to wait and see.

Creativity is desired. Learning would be good. Nothing is required.

Thursday, January 13, 2011

I'm no ventriloquist

Nope. I'm quite sure though, that I once read a short article about "how to become a ventriloquist". And even tried pronouncing danana without lip movement. I recall b as being very lippy indeed.

But if I were a ventriloquist I would definitely build or buy myself a gigantic doll. I guess buy. I'm really not very good at building stuff, especially dolls. And I really do mean a giant big doll.

Instead of the more common act "man has hand in doll" I would go on stage with the somewhat more appealing and appalling "doll has hand in man". For the less imaginative of us, this would be an ideal place to envision a giant doll with his hand in a much smaller man.

Wednesday, January 6, 2010

The Berg

Not that I just found out but I'm very much still disappointed. In so many ways. The germans are going to build a mountain, a tiny silly one, but still a mountain. http://www.the-berg.de/

What disappoints me is that a friend and I had the exact same idea some 5+ years ago. Why didn't we go through with it? We could have been known as the mountain kings or similar all over Europe. Sweet. And sour.

Actually it wasn't exactly the same idea. We would have made a decent mountain. 4 km high. At least. Making it possible to go skiing much of the year. And since we are Danes and Jutlanders we (with a bit of help) would have placed the mountain in Jutland, which is basically an enourmous, in danish terms, piece of flatlands. This would surely have put Denmark on the ski-map, if such exists.

Still, the ideas are pretty close. So why didn't we build the damn thing?

Well none of us knew or know anything about building anything. But both of us seemed to like the process of estimating. Especially estimating things that we know nothing of. So how do you go about building mountains? And how much time does it take to build?

The mountain we chose was the old school solid one. Build out of dirt, rocks and everything else you get when digging up dirt, rocks and everything else. Building is then accomplished by tossing more of this on top of the already build mountain. Problem solved. So the difficult part is really finding out how much time it takes. Here is one estimate. It might be close to our original estimate. I think so.

We need 1 mountain, 4 km high, with an inclination of approximately 10 degrees. That should hopefully make skiing ok. Mountains are sort of cone formed, but the volume of a cone is sort of hard to remember or recreate when you are uninterested. And finding the radius of a cone is sort of the same business. So instead we go for the good old cube. One guess, try waving your hands for inspiration, is that a cube with sides of 4 km might have somewhat the same volume as the mountain cone. That is 64 cubic km.

Having determined the volume needed, we need some truck drivers. And trucks to move them around. I only know a few truck drivers, if any at all. And even fewer trucks. But I guess you could easily get 1000 trucks down at the big truck station. And put 3000 truck drivers in them. Making them operable 24 hours a day. How big are trucks I hear me ask, let's say 5 by 5 by 20 meters. 500 cubic meters. Or 0.0000005 cubic km.

So, assuming we can dig freely (we might have to ask someone about this) somewhere nearby, Jutland, Funen or Schleswig, a truck might deliver 10 truck loads each day.

64 cubic km divided by 0.0000005 cubic km per truck yields 128,000,000 truck loads. I guess.
The truckers do 10,000 truck loads a day, yielding 12800 days of building. Or approximately 30+ years.

As I recall, we didn't build the mountain mostly due to the above, who wants to wait 30 years for anything. And I think we had other reasons to. I recall I had soccer practice later that evening. I don't think we did an estimate of the price of building the mountain, but I imagine that is not negligible.

In fact I think we ended with 80+ years in our original estimate making it even more hopeless. But I also recall we settled for just a quarter of a mountain. That might also be a viable option now.

I guess construction engineers and everyone else might have better estimates or methods for building.  But that's not the point. I guess.

Anything remotely interesting

A blog. Anythink of interest might appear irregularly. I hope.