diff --git a/examples/flappylearning/.gitignore b/examples/flappylearning/.gitignore
new file mode 100644
index 0000000000..dc22e61c94
--- /dev/null
+++ b/examples/flappylearning/.gitignore
@@ -0,0 +1 @@
+game
diff --git a/examples/flappylearning/LICENSE b/examples/flappylearning/LICENSE
new file mode 100644
index 0000000000..87cfb6f0d1
--- /dev/null
+++ b/examples/flappylearning/LICENSE
@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2020 uxnow
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
diff --git a/examples/flappylearning/README.md b/examples/flappylearning/README.md
new file mode 100644
index 0000000000..57e288910f
--- /dev/null
+++ b/examples/flappylearning/README.md
@@ -0,0 +1,16 @@
+# flappylearning-v
+flappy learning implemented by vlang
+
+## get started
+
+```sh
+v run game.v
+```
+
+![flappy.png](img/flappy.png)
+
+## thanks
+https://github.com/xviniette/FlappyLearning
+
+## license
+MIT
diff --git a/examples/flappylearning/game.v b/examples/flappylearning/game.v
new file mode 100644
index 0000000000..65c10625de
--- /dev/null
+++ b/examples/flappylearning/game.v
@@ -0,0 +1,275 @@
+
+module main
+
+import gg
+import gx
+import os
+import time
+import math
+import rand
+
+import neuroevolution
+
+const (
+	win_width  = 500
+	win_height = 512
+	timer_period = 24 // ms
+)
+
+struct Bird {
+mut:
+	x f64 = 80
+	y f64 = 250
+	width f64 = 40
+	height f64 = 30
+
+	alive bool = true
+	gravity f64
+	velocity f64 = 0.3
+	jump f64 = -6
+}
+
+fn (mut b Bird) flap() {
+	b.gravity = b.jump
+}
+
+fn (mut b Bird) update() {
+	b.gravity += b.velocity
+	b.y += b.gravity
+}
+
+fn (b Bird) is_dead(height f64, pipes []Pipe) bool {
+	if b.y >= height || b.y + b.height <= 0 {
+		return true
+	}
+	for pipe in pipes {
+		if !(
+			b.x > pipe.x + pipe.width ||
+			b.x + b.width < pipe.x ||
+			b.y > pipe.y + pipe.height ||
+			b.y + b.height < pipe.y
+		) {
+			return true
+		}
+	}
+	return false
+}
+
+struct Pipe {
+mut:
+	x f64 = 80
+	y f64 = 250
+	width f64 = 40
+	height f64 = 30
+	speed f64 = 3
+}
+
+fn (mut p Pipe) update() {
+	p.x -= p.speed
+}
+
+fn (p Pipe) is_out() bool {
+	return p.x + p.width < 0
+}
+
+struct App {
+mut:
+	gg    &gg.Context
+	background gg.Image
+	bird gg.Image
+	pipetop gg.Image
+	pipebottom gg.Image
+
+	pipes []Pipe
+	birds []Bird
+	score int
+	max_score int
+	width f64 = win_width
+	height f64 = win_height
+	spawn_interval f64 = 90
+	interval f64
+
+	nv neuroevolution.Generations
+	gen []neuroevolution.Network
+	alives int
+	generation int
+
+	background_speed f64 = 0.5
+	background_x f64
+}
+
+fn (mut app App) start() {
+	app.interval = 0
+	app.score = 0
+	app.pipes = []
+	app.birds = []
+	app.gen = app.nv.generate()
+
+	for _ in 0 .. app.gen.len {
+		app.birds << Bird{}
+	}
+	app.generation++
+	app.alives = app.birds.len
+}
+
+fn (app &App) is_it_end() bool {
+	for i in 0 .. app.birds.len {
+		if app.birds[i].alive {
+			return false
+		}
+	}
+
+	return true
+}
+
+fn (mut app App) update() {
+	app.background_x += app.background_speed
+	mut next_holl := f64(0)
+
+	if app.birds.len > 0 {
+		for i := 0; i < app.pipes.len; i += 2 {
+			if app.pipes[i].x + app.pipes[i].width > app.birds[0].x {
+				next_holl = app.pipes[i].height / app.height
+				break
+			}
+		}
+	}
+
+	for mut j, bird in app.birds {
+		if bird.alive {
+			inputs := [
+				bird.y / app.height,
+				next_holl,
+			]
+			res := app.gen[j].compute(inputs)
+			if res[0] > 0.5 {
+				bird.flap()
+			}
+
+			bird.update()
+
+			if bird.is_dead(app.height, app.pipes) {
+				bird.alive = false
+				app.alives--
+				app.nv.network_score(app.gen[j], app.score)
+				if app.is_it_end() {
+					app.start()
+				}
+			}
+
+		}
+	}
+
+	for k := 0; k < app.pipes.len; k++ {
+		app.pipes[k].update()
+		if app.pipes[k].is_out() {
+			app.pipes.delete(k)
+			k--
+		}
+	}
+
+	if app.interval == 0 {
+		delta_bord := f64(50)
+		pipe_holl := f64(120)
+		holl_position := math.round(rand.f64() * (app.height - delta_bord * 2.0 - pipe_holl)) + delta_bord
+		app.pipes << Pipe{
+			x: app.width
+			y: 0
+			height: holl_position
+		}
+
+		app.pipes << Pipe{
+			x: app.width
+			y: holl_position + pipe_holl
+			height: app.height
+		}
+	}
+
+	app.interval++
+
+	if app.interval == app.spawn_interval {
+		app.interval = 0
+	}
+
+	app.score++
+	app.max_score = if app.score > app.max_score {
+		app.score
+	} else {
+		app.max_score
+	}
+
+}
+
+fn main() {
+	mut app := &App{
+		gg: 0
+	}
+	app.gg = gg.new_context({
+		bg_color: gx.white
+		width: win_width
+		height: win_height
+		use_ortho: true // This is needed for 2D drawing
+		create_window: true
+		window_title: 'flappylearning-v'
+		frame_fn: frame
+		user_data: app
+		init_fn: init_images
+		font_path: os.resource_abs_path('../assets/fonts/RobotoMono-Regular.ttf')
+	})
+	app.nv = neuroevolution.Generations{
+		population: 50
+		network: [2, 2, 1]
+	}
+	app.start()
+	go app.run()
+	app.gg.run()
+}
+
+fn (mut app App) run() {
+	for {
+		app.update()
+		time.sleep_ms(timer_period)
+	}
+}
+
+fn init_images(mut app App) {
+	app.background = app.gg.create_image(os.resource_abs_path('./img/background.png'))
+	app.bird = app.gg.create_image(os.resource_abs_path('./img/bird.png'))
+	app.pipetop = app.gg.create_image(os.resource_abs_path('./img/pipetop.png'))
+	app.pipebottom = app.gg.create_image(os.resource_abs_path('./img/pipebottom.png'))
+}
+
+fn frame(app &App) {
+	app.gg.begin()
+	app.draw()
+	app.gg.end()
+}
+
+fn (app &App) display() {
+	for i := 0; i < int(math.ceil(app.width / app.background.width) + 1.0); i++ {
+		background_x := i * app.background.width - math.floor(int(app.background_x) % int(app.background.width))
+		app.gg.draw_image(f32(background_x), 0, app.background.width, app.background.height, app.background)
+	}
+
+	for i, pipe in app.pipes {
+		if i % 2 == 0 {
+			app.gg.draw_image(f32(pipe.x), f32(pipe.y + pipe.height - app.pipetop.height), app.pipetop.width, app.pipetop.height, app.pipetop)
+		} else {
+			app.gg.draw_image(f32(pipe.x), f32(pipe.y), app.pipebottom.width, app.pipebottom.height, app.pipebottom)
+		}
+	}
+
+	for bird in app.birds {
+		if bird.alive {
+			app.gg.draw_image(f32(bird.x), f32(bird.y), app.bird.width, app.bird.height, app.bird)
+		}
+	}
+	app.gg.draw_text_def(10 ,25, 'Score: $app.score')
+	app.gg.draw_text_def(10 ,50, 'Max Score: $app.max_score')
+	app.gg.draw_text_def(10 ,75, 'Generation: $app.generation')
+	app.gg.draw_text_def(10 ,100, 'Alive: $app.alives / $app.nv.population')
+}
+
+fn (app &App) draw() {
+	app.display()
+}
diff --git a/examples/flappylearning/img/background.png b/examples/flappylearning/img/background.png
new file mode 100644
index 0000000000..22b307092c
Binary files /dev/null and b/examples/flappylearning/img/background.png differ
diff --git a/examples/flappylearning/img/bird.png b/examples/flappylearning/img/bird.png
new file mode 100644
index 0000000000..ca3a15990b
Binary files /dev/null and b/examples/flappylearning/img/bird.png differ
diff --git a/examples/flappylearning/img/flappy.png b/examples/flappylearning/img/flappy.png
new file mode 100644
index 0000000000..c7085aa8ad
Binary files /dev/null and b/examples/flappylearning/img/flappy.png differ
diff --git a/examples/flappylearning/img/pipebottom.png b/examples/flappylearning/img/pipebottom.png
new file mode 100644
index 0000000000..99d27b2679
Binary files /dev/null and b/examples/flappylearning/img/pipebottom.png differ
diff --git a/examples/flappylearning/img/pipetop.png b/examples/flappylearning/img/pipetop.png
new file mode 100644
index 0000000000..1fcd52b3c4
Binary files /dev/null and b/examples/flappylearning/img/pipetop.png differ
diff --git a/examples/flappylearning/modules/neuroevolution/neuronevolution.v b/examples/flappylearning/modules/neuroevolution/neuronevolution.v
new file mode 100644
index 0000000000..a9422d1a3c
--- /dev/null
+++ b/examples/flappylearning/modules/neuroevolution/neuronevolution.v
@@ -0,0 +1,329 @@
+
+module neuroevolution
+
+import rand
+import math
+
+fn random_clamped() f64 {
+	return rand.f64() * 2 - 1
+}
+
+pub fn activation(a f64) f64 {
+	ap := (-a) / 1
+	return (1 / (1 + math.exp(ap)))
+}
+
+fn round(a int, b f64) int {
+	return int(math.round(f64(a) * b))
+}
+
+struct Neuron {
+mut:
+	value f64
+	weights []f64
+}
+
+fn (mut n Neuron) populate(nb int) {
+	for _ in 0 .. nb {
+		n.weights << random_clamped()
+	}
+}
+
+struct Layer {
+	id int
+mut:
+	neurons []Neuron
+}
+
+fn (mut l Layer) populate(nb_neurons int, nb_inputs int) {
+	for _ in 0 .. nb_neurons {
+		mut n := Neuron{}
+		n.populate(nb_inputs)
+		l.neurons << n
+	}
+}
+
+struct Network {
+mut:
+	layers []Layer
+}
+
+fn (mut n Network) populate(network []int) {
+
+	assert network.len >= 2
+	input := network[0]
+	hiddens := network.slice(1, network.len - 1)
+	output := network[network.len - 1]
+
+	mut index := 0
+	mut previous_neurons := 0
+	mut input_layer := Layer{
+		id: index
+	}
+	input_layer.populate(input, previous_neurons)
+	n.layers << input_layer
+
+	previous_neurons = input
+	index++
+	for hidden in hiddens {
+		mut hidden_layer := Layer{
+			id: index
+		}
+		hidden_layer.populate(hidden, previous_neurons)
+		previous_neurons = hidden
+		n.layers << hidden_layer
+		index++
+	}
+
+	mut output_layer := Layer{
+		id: index
+	}
+	output_layer.populate(output, previous_neurons)
+	n.layers << output_layer
+}
+
+fn (n Network) get_save() Save {
+
+	mut save := Save{}
+	for layer in n.layers {
+		save.neurons << layer.neurons.len
+		for neuron in layer.neurons {
+			for weight in neuron.weights {
+				save.weights << weight
+			}
+		}
+	}
+	return save
+}
+
+fn (mut n Network) set_save(save Save) {
+
+	mut previous_neurons := 0
+	mut index := 0
+	mut index_weights := 0
+
+	n.layers = []
+	for save_neuron in save.neurons {
+		mut layer := Layer{
+			id: index
+		}
+		layer.populate(save_neuron, previous_neurons)
+		for mut neuron in layer.neurons {
+			for i in 0 .. neuron.weights.len {
+				neuron.weights[i] = save.weights[index_weights]
+				index_weights++
+			}
+		}
+		previous_neurons = save_neuron
+		index++
+		n.layers << layer
+	}
+}
+
+pub fn (mut n Network) compute(inputs []f64) []f64 {
+	assert n.layers.len > 0
+	assert inputs.len == n.layers[0].neurons.len
+
+	for i, input in inputs {
+		n.layers[0].neurons[i].value = input
+	}
+
+	mut prev_layer := n.layers[0]
+
+	for i in 1 .. n.layers.len {
+		for j, neuron in n.layers[i].neurons {
+			mut sum := f64(0)
+			for k, prev_layer_neuron in prev_layer.neurons {
+				sum += prev_layer_neuron.value * neuron.weights[k]
+			}
+			n.layers[i].neurons[j].value = activation(sum)
+		}
+		prev_layer = n.layers[i]
+	}
+
+	mut outputs := []f64{}
+	mut last_layer := n.layers[n.layers.len - 1]
+	for neuron in last_layer.neurons {
+		outputs << neuron.value
+	}
+
+	return outputs
+}
+
+struct Save {
+mut:
+	neurons []int
+	weights []f64
+}
+
+fn (s Save) clone() Save {
+	mut save := Save{}
+	save.neurons << s.neurons
+	save.weights << s.weights
+	return save
+}
+
+struct Genome {
+	score int
+	network Save
+}
+
+struct Generation {
+mut:
+	genomes []Genome
+}
+
+fn (mut g Generation) add_genome(genome Genome) {
+
+	mut i := 0
+
+	for gg in g.genomes {
+		if genome.score > gg.score {
+			break
+		}
+
+		i++
+	}
+
+	g.genomes.insert(i, genome)	
+}
+
+fn (g1 Genome) breed(g2 Genome, nb_child int) []Save {
+	mut datas := []Save{}
+
+	for _ in 0 .. nb_child {
+
+		mut data := g1.network.clone()
+
+		for i, weight in g2.network.weights {
+			if rand.f64() <= 0.5 {
+				data.weights[i] = weight
+			}
+		}
+
+		for i, _ in data.weights {
+			if rand.f64() <= 0.1 {
+				data.weights[i] += (rand.f64() * 2 - 1) * 0.5
+			}
+		}
+
+		datas << data
+	}
+
+	return datas
+}
+
+fn (g Generation) next(population int) []Save {
+
+	mut nexts := []Save{}
+
+	if population == 0 {
+		return nexts
+	}
+
+	keep := round(population, 0.2)
+
+	for i in 0 .. keep {
+		if nexts.len < population {
+			nexts << g.genomes[i].network.clone()
+		}
+	}
+
+	random := round(population, 0.2)
+
+	for _ in 0 ..  random {
+
+		if nexts.len < population {
+			mut n := g.genomes[0].network.clone()
+			for k, _ in n.weights {
+				n.weights[k] = random_clamped()
+			}
+			nexts << n
+		}
+	}
+
+	mut max := 0
+	out: for {
+		for i in 0 .. max {
+			mut childs := g.genomes[i].breed(g.genomes[max], 1)
+			for c in childs {
+				nexts << c
+				if nexts.len >= population {
+					break out
+				}
+			}
+		}
+		max++
+		if max >= g.genomes.len - 1 {
+			max = 0
+		}
+	}
+
+	return nexts
+}
+
+pub struct Generations {
+pub:
+	population int
+	network []int
+mut:
+	generations []Generation
+}
+
+fn (mut gs Generations) first() []Save {
+	mut out := []Save{}
+	for _ in 0 .. gs.population {
+		mut nn := Network{}
+		nn.populate(gs.network)
+		out << nn.get_save()
+	}
+
+	gs.generations << Generation{}
+	return out
+}
+
+fn (mut gs Generations) next() []Save {
+	assert gs.generations.len > 0
+	gen := gs.generations[gs.generations.len - 1].next(gs.population)
+	gs.generations << Generation{}
+	return gen
+}
+
+fn (mut gs Generations) add_genome(genome Genome) {
+	assert gs.generations.len > 0
+	gs.generations[gs.generations.len - 1].add_genome(genome)
+}
+
+fn (mut gs Generations) restart() {
+	gs.generations = []
+}
+
+pub fn (mut gs Generations) generate() []Network {
+
+	saves := if gs.generations.len == 0 {
+		gs.first()
+	} else {
+		gs.next()
+	}
+
+	mut nns := []Network{}
+	for save in saves {
+		mut nn := Network{}
+		nn.set_save(save)
+		nns << nn
+	}
+
+	if gs.generations.len >= 2 {
+		gs.generations.delete(0)
+	}
+
+	return nns
+}
+
+pub fn (mut gs Generations) network_score(network Network, score int) {
+	gs.add_genome(Genome{
+		score: score
+		network: network.get_save()
+	})
+}
+