Lyapunov Fractal

2017-12-30, post № 189

Java, mathematics, programming, #chaos, #chaos theory, #fractal geometry, #fractal viewer, #Java 1.8

Lyapunov fractals are mathematical objects living in the plane, graphing regions of stability or chaos regarding a certain logistical map following an ever-changing population growth, alternating between two values. I implemented a Lyapunov fractal viewer in Java 1.8 which lets you freely move around the plane, set sequences and iteration depths to explore the fractal. Source code can be seen below or downloaded, as can the Java executable (Lyapunov.java, Lyapunov.jar).

Zircon Zity

Articles on the topic of Lyapunov fractals are a post by Earl Glynn^[1] from the year 2000 in which they talk about their Lyapunov fractal generator written in Pascal — a rewrite of a software written in 1992. Also worth reading is A. Dewdney’s article about Mario Markus’ work (Scientific American, September 1991).

My first encounter with this fractal was whilst browsing Wikipedia and stumbling across its Wikipedia entry. Since it was a fractal and looked fairly intricate and thus explorable, I set out to write my own renderer — I chose to implement it in Java, as it is a compiled language with a nice set of GUI libraries. Since Wikipedia listed an algorithm for generating Lyapunov fractal images, I thought an implementation would be fairly straight-forward. However, als always, the devil is in the detail and I quickly noticed that while the short Wikipedia entry looks convincing and well-written at first glance, lacks to answer deeper questions regarding the very topic it is about.

Eerie Tendrils

The following is a description of the fractal generation algorithm. It is a modified version of the algorithm detailed by Wikipedia which addresses certain ambiguities the original had (more on those later).

Fractal generation algorithm

Take as input a complex point $(a,b)\in\mathbb{R}^2$ and a sequence $S^*$ consisting of the letters $\text{A}$ and $\text{B}$ ; for example $S^*=\text{AAAAAABBBBBB}$ .
Construct a function $S\colon\mathbb{N}_0\to\{\text{A},\text{B}\},\quad n\mapsto S^*_{(n\mod |S^*|)}$ which returns the zero indexed 𝑛-th sequence entry and wraps around to the sequence’s start once the sequence’s end is hit.
Construct a function
$r\colon\mathbb{N}_0\to\{a,b\},\quad n\mapsto\begin{cases}a&\text{if }S(n)=\text{A}\\b&\text{if }S(n)=\text{B}\end{cases}$
which selects either 𝑎 or 𝑏.
Let $x_0=0.5$ and define $x_n=r(n-1)\cdot x_{n-1}\cdot (1-x_{n-1})$ .
Calculate the Lyapunov exponent 𝜆 as follows.
$\lambda = \lim\limits_{N \to \infty} \frac{1}{N} \cdot \sum\limits_{n=1}^{N} \log_2{|r(n)\cdot(1-2\cdot x_n)|}$
Since calculating the limit as 𝑁 goes to infinity is rather difficult in practice, one can simply use a sufficiently large 𝑁 and pretend to have reached infinity.
Output a color value according to 𝜆. I chose green for 𝜆 < 𝟢 and blue for 𝜆 ≥ 𝟢, using arbitrarily defined value ranges to map 𝜆 to an actual color.

A Spec of Fractal

The following is a Java snippet which implements the algorithm described above.

// choose a or b according to Seq and n
static double r(String Seq, int n, double a, double b) {
    if (Seq.charAt(n%Seq.length()) == 'A') return a; else return b;
}
 
// calculate a pixel color (0x00rrggbb) for given parameters
static int LyapunovPixel(double a, double b, String Seq, int N) {
    // array of all x_n; x_0, the starting value, initialize all x_n values
    double[] X = new double[N+1]; X[0] = .5;
    for (int n = 1; n <= N; n++) X[n] = r(Seq,n-1,a,b)*X[n-1]*(1-X[n-1]);

    // calculate the Lyapunov exponent (to a certain precision dictated by N)
    double lmb = 0;
    for (int n = 1; n <= N; n++)
        lmb += Math.log(Math.abs(r(Seq,n,a,b)*(1-2*X[n])))/Math.log(2);
    lmb /= N;

    // infinity was hit, use a black pixel
    if (Double.isInfinite(lmb)) return 0x000000;

    // color pixel according to Lyapunov exponent
    double MIN = -1, MAX = 2;
    if (lmb < 0) return ((int)(lmb/MIN*255))<<8;
    else         return ((int)(lmb/MAX*255))<<0;
}

Lyapunov Spike

Coming back to Wikipedia’s algorithm, there were a few things I found irritating at best when attempting my implementation and thus addressed in the algorithm description seen above. A closer look at potentially misleading or ambiguos statements follows, together with my reasoning to resolve them.

It is not clear whether the sequence is zero or one indexed, though it has to be zero indexed as $x_0=0.5$ , $x_{n+1}=r_n\cdot\dots$ evaluates to $x_1=r_0\cdot\dots$ , implying the definition of $r_0$ and thus $S_0$ .
It is not clear which logarithm is meant; the natural logarithm, a generic $\log_b$ or the dyadic logarithm — the latter is actually used. To find the actual logarithm base, I dug deeper and used Wikipedia’s external links to find a post by Earl Glynn^[2], answering this question.
It is not clear what is meant by the second half of the sentence beneath the Lyapunov exponent. It reads “… dropping the first summand as $r_0(1-2x_0)=r_n\cdot 0=0$ for $x_0=0.5$ .” As the sum starts with 𝑛 = 𝟣 and one would not dare to calculate $\log_2{|r(0)\cdot(1-2\cdot x_0)|}=\log_2{0}$ for $x_0=0.5$ , this sentence’s existence bewilders me.
It is not clear how exactly the colors are derived, only that they have something to do with the Lyapunov exponent. I simply chose arbitrary values that looked convincing for mapping 𝜆 to a pixel color.

Dark Swirl

One of the most frustrating bugs I came across was an unexplainable axis flip. My code generated the fractal just fine except for the fact that every image was flipped along the diagonal crossing the origin with a 𝟦𝟧° angle to the horizontal. It was as though the coordinates $(a,b)$ were swapped somewhere in my code and I simply could not figure out where.
Finally, after hours of looking at the same code over and over again, a closer look at Earl Glynn’s post^[3] brought an end to my misery. Just below the three welcoming fractal renderings, a screenshot of their software is shown — complete with a blue and red line indicating the coordinate system’s orientation. 𝑎 and 𝑏 are — contrary to all coordinate systems involving parameters named after the first two letters in the latin alphabet — indeed flipped. Wikipedia’s images must have simply ran with this decision, without noting it.
Because of this flip, when one wants to render images specified in the reversed sequence format, they simply have to swap all letters (for example $\text{BBBABA}$ becomes $\text{AAABAB}$ ).

As Glynn themself says, “I would have reversed the ‘a’ and ‘b’ labels to be consistent with normal ‘x’ and ‘y’ axes conventions, but conform here with the same convention as used by Markus.” (Referring to Mario Markus, co-author of Lyapunov Exponents of the Logistic Map with Periodic Forcing.)

Slurping Cell

After having eliminated all vagueness regarding the fractal generation, writing the outer Java viewer hull was a fairly easy task. As a template I took my Mandelbrot Set III viewer (the reason why most variable names reference complex numbers) complete with multithreading, allowing a pixely fractal exploration until one stops their exploration, letting the thread catch up and display a higher resolution image. The same is done for rendering 𝟦K images and saving them to files in the background — 𝟦K renderings are saved as .png files and named according to their parameters. A list of program controls follows.

Controls

Mouse dragging lets one pan the complex plane.
Mouse scrolling lets one zoom the complex plane.
Pressing N evokes a dialogue box where one can specify an iteration depth 𝑁.
Pressing S evokes a dialogue box where one can enter a sequence $S^*$ .
Pressing R resets all fractal parameters to the default.
Pressing P initiates a 𝟦K fractal rendering. The 𝟦K fractal rendering thread can be killed by exiting the program prematurely, thus losing (!) the rendering.

Source code: Lyapunov.java

Christmas MMXVII

2017-12-25, post № 188

art, haiku, poetry, #fractal, #Mandelbrot set

Barren trees, white snow.
Cold and lasting winter nights.
Quiet fire crackling.

christmas-mmxvii_3840x2160_Zre0.9094277019548073_Zim-0.23235433971456212_Zom1.4334111979667834E-4_exp10.0.png

christmas-mmxvii_3840x2160_Zre0.9096493373321258_Zim-0.23226989571562165_Zom1.1947838420050083E-9_exp10.0.png

christmas-mmxvii_3840x2160_Zre0.7835723334259043_Zim0.34259288039158464_Zom6.855961324128004E-5_exp10.0.png

Python Matrix Module

2017-12-16, post № 187

mathematics, programming, Python, #linear algebra, #matrices

Matrices are an important part of linear algebra. By arranging scalars in a rectangular manner, one can elegantly encode vector transformations like scaling, rotating, shearing and squashing, solve systems of linear equations and represent general vector space homomorphisms.
However, as powerful as matrices are, when actually applying the theoretical tools one has to calculate specific values. Doing so by hand can be done, yet gets cumbersome quite quickly when dealing with any matrices which contain more than a few rows and columns.

So, even though there are a lot of other implementations already present, I set out to write a Python matrix module containing a matrix class capable of basic matrix arithmetic (matrix multiplication, transposition, …) together with a set of functions which perform some higher-level matrix manipulation like applying Gaussian elimination, calculating the reduced row echelon form, determinant, inversion and rank of a given matrix.

Module source code can be seen below and downloaded. When saved as matrix.py in the current working directory, one can import the module as follows.

>>> import matrix
>>> A = matrix.Matrix([[13,  1, 20, 18],
...                    [ 9, 24,  0,  9],
...                    [14, 22,  5, 18],
...                    [19,  9, 15, 14]])
>>> print A**-1
        -149/1268  -67/634   83/1268 171/1268
          51/1268 239/1902 -105/1268 -33/1268
Matrix(    73/634 803/4755  -113/634 -87/3170 )
          13/1268  -75/634  197/1268 -83/1268

Matrices are defined over a field, typically $\mathbb{F}=\mathbb{R}$ ^[1] in theoretical use, though for my implementation I chose not to use a double data structure^[2], as it lacked the conceptual precision in numbers like a third^[3]. As one cannot truly represent a large portion of the reals anyways, I chose to use $\mathbb{F}=\mathbb{Q}$ , which also is a field though can be — to a certain scalar size and precision — accurately represented using fractional data types (Python’s built-in Fraction is used here).

To simplify working with matrices, the implemented matrix class supports operator overloading such that the following expressions — A[i,j], A[i,j]=l, A*B, A*l, A+B, -A, A/l, A+B, A-B, A**-1, A**"t", ~A, A==B, A!=B — all behave in a coherent and intuitive way for matrices A, B, scalars l and indices i, j.

When working with matrices, there are certain rules that must be obeyed, like proper size when adding or multiplying, invertibility when inverting and others. To minimize potential bug track down problems, I tried to include a variety of detailed exceptions (ValueErrors) explaining the program’s failure at that point.

Apart from basic matrix arithmetic, a large part of my module centers around Gaussian elimination and the functions that follow from it. At their heart lies the implementation of GaussianElimination, a function which calculates the reduced row echelon form rref(A) of a matrix together with the transformation matrix T such that T*A = rref(A), a list of all matrix pivot coordinates, the number of row transpositions used to achieve row echelon form and a product of all scalars used to achieve reduced row echelon form.
From this function, rref(A) simply returns the first, rrefT(A) the second parameter. Functions rcef(A) (reduced column echelon form) and rcefS(A) (A*S = rcef(A)) follow from repeated transposition.
Determinant calculation uses characteristic determinant properties (multilinear, alternating and the unit hypercube has hypervolume one).

$\det \begin{pmatrix} a_{1 1}&a_{1 2}&\dots&a_{1 n}\\ \vdots&\vdots&\ddots&\vdots&\\ \lambda \cdot a_{i 1}&\lambda \cdot a_{i 2}&\dots&\lambda \cdot a_{i n}\\ \vdots&\vdots&\ddots&\vdots&\\ a_{n 1}&a_{n 2}&\dots&a_{n n} \end{pmatrix} = \lambda \cdot \det \begin{pmatrix} a_{1 1}&a_{1 2}&\dots&a_{1 n}\\ \vdots&\vdots&\ddots&\vdots&\\ a_{i 1}&a_{i 2}&\dots&a_{i n}\\ \vdots&\vdots&\ddots&\vdots&\\ a_{n 1}&a_{n 2}&\dots&a_{n n} \end{pmatrix}$
$\det \begin{pmatrix} a_{1 1}&a_{1 2}&\dots&a_{1 n}\\ \vdots&\vdots&\ddots&\vdots&\\ a_{i 1}&a_{i 2}&\dots&a_{i n}\\ \vdots&\vdots&\ddots&\vdots&\\ a_{j 1}&a_{j 2}&\dots&a_{j n}\\ \vdots&\vdots&\ddots&\vdots&\\ a_{n 1}&a_{n 2}&\dots&a_{n n} \end{pmatrix} = -\det \begin{pmatrix} a_{1 1}&a_{1 2}&\dots&a_{1 n}\\ \vdots&\vdots&\ddots&\vdots&\\ a_{j 1}&a_{j 2}&\dots&a_{j n}\\ \vdots&\vdots&\ddots&\vdots&\\ a_{i 1}&a_{i 2}&\dots&a_{i n}\\ \vdots&\vdots&\ddots&\vdots&\\ a_{n 1}&a_{n 2}&\dots&a_{n n} \end{pmatrix}$
$\det \begin{pmatrix}1&a_{1 2}&a_{1 3}&\dots&a_{1 n}\\0&1&a_{2 3}&\dots&a_{2 n}\\0&0&1&\dots&a_{3 n}\\\vdots&\vdots&\vdots&\ddots&\vdots\\0&0&0&\dots&1\end{pmatrix} = \det \bold{1}_n = 1$

Using these properties, the determinant is equal to the product of the total product of all factors used in transforming the matrix into reduced row echelon form and the permutation parity (minus one to the power of the number of transpositions used).
Questions regarding invertibility and rank can also conveniently be implemented using the above described information.

All in all, this Python module implements basic matrix functionality paired with a bit more involved matrix transformations to build a usable environment for manipulating matrices in Python.

Source code: python-matrix-module.py

Extra assets: python-matrix-module_old-version.py

Animating the Quantum Drunkard’s Walk

2017-12-02, post № 186

ImageMagick, PIL, programming, Python, #animation, #automation, #cellular, #PPCG, #quantum mechanics

A recent PPCG challenge titled The Quantum Drunkard’s Walk was about a tiny drunken person for which quantum mechanics apply and who — being drunk — will randomly walk in one of the four cardinal directions each step they take.
As they experience the strange world of quanta, they can simultaneously take every path at once and their paths influence each other. The challenge then asked to output all possible positions the quantum drunkard could occupy, all paths they could take in ASCII representation.

The question also states this problem’s equivalence to a cellular automaton, when one removes the story boilerplate.
Cells in this cellular automaton are square and can occupy one of three states: empty, awake or sleeping. Each iteration, all cells change according to three rules.

An empty cell wakes up iff there is exactly one awake cell amongst its cardinal neighbours, else it stays empty.
An awake cell goes to sleep.
A sleeping cell continues to sleep.

Being code golf, the aim was to come up with a minimally sized source code; my attempt required 𝟤𝟣𝟦 bytes and prints a nested array containing one-length strings (characters), as this output method is cheaper than concatenating all characters to one string.

python quantum.py^[1] -rmi 200

However, one user took the challenge idea a bit further and created an animated gif showing the walk’s, or cellular automaton’s, progression over time with an increasing number of iterations. My Python program shown in this post does exactly that, generating an animated gif showing the automaton’s progression. I even implemented rainbow support, possibly improving the automaton’s visual appearance.
Python source code can be downloaded and seen below.

I use the Python Imaging Library to produce all frames and use a shell call to let ImageMagick convert all frames to an animated gif. Animation parameters are taken via shell arguments, a full list of features follows (also available via the -h flag).

--iterations N: Number of iterations (initial frame not counted)

--colorempty C: Empty cell color (#rrggbb)
--colorawake C: Awake cell color (#rrggbb)
--colorsleeping C: Sleeping cell color (#rrggbb)
--rainbow: Use rainbow colors (overrides color settings)
--scale N: Cell square pixel size

--convert: Execute ImageMagick’s convert command
--delay N: Delay between frames in output image.
--loop N: Gif loops (0 means for indefinite looping)
--keepfiles: Do not delete files when converting

animating-the-quantum-drunkards-walk_ten-iterations.gif

python quantum.py -s 25 -md 50

Source code: animating-the-quantum-drunkards-walk.py

Extra assets: animating-the-quantum-drunkards-walk_backup-0.py, animating-the-quantum-drunkards-walk_backup-1.py

Fire Photographies

2017-11-18, post № 185

art, haiku, poetry,

A tiny spark flies;
Ignites a crumbled paper.
Thus fire is born.

Extra assets: fire-photographies-26.jpg, fire-photographies-35.jpg, fire-photographies-44.jpg

Generic C / Python Polyglot

2017-11-04, post № 184

C, programming, Python, #recreational

A polyglot — coming from the Greek words πολύς (many) and γλώττα (tongue) — is a piece of source code which can run in multiple languages, often performing language-dependent tasks.
Even though such a source code’s feature may not seem particularily useful or possible with certain language combinations, trying to bend not one but multiple language’s syntactic rules and overall behavior is still interesting.

An example of a rather simple polyglot would be a Python 2 / Python 3 polyglot — if one counts those as two separate languages. Because of their enormous similarities, one can pick out differences and use those to determine which language runs the source code.

if 1/2>0:print("Python 3")
else:print("Python 2")

Utilizing Python’s version difference regarding integer division and real division is one of the most common ways to tell them apart, as it can also be used to only control a specific part of a program instead of having to write two nearly identical programs (increases the program’s style and cuts on bytes — an important consideration if one golfs a polyglot).

However, polyglots combining languages that are not as similar as Python 2 and Python 3 require more effort. The following is a general Python 2 / C polyglot, meaning that nearly all C and Python 2 programs can be mixed using this template (there are a few rules both languages need to abide which will come apparent later).

#define _\
"""
main(){printf("C");}
#define _"""
#define/*
print"Python 2"
#*/_f

In the above code, main(){printf("C");} can be nearly any C code and print"Python 2" can be nearly any Python 2 code.
Language determination is exclusively done via syntax. A C compiler^[1] sees a #define statement and a line continuation \, another two #define statements with a block comment in between and actual compilable C source code (view first emphasis).Python, on the other hand, treats all octothorps, #, as comments, ignoring the line continuation, and triple-quoted strings, """...""", as strings rather than statements and thus only sees the Python code (view second emphasis).

My first ever polyglot used this exact syntactical language differentiation and solved a task titled Life and Death of Trees (link to my answer).

Halloween MMXVII

2017-10-31, post № 183

art, poetry, #dark, #iambic pentameter, #poem, #Shakespearean sonnet, #sonnet

Gentle whispers wake my worldly being,
for long have I been laying here asleep.
My eye I open, though without seeing;
a world possessed by darkness — far and deep.

I want to try and move but must I not.
An every thought consumed by this abyss.
I try unraveling this psychic knot,
though does this fight seem to be meaningless.

Alas, before my mind is swallowed whole,
a square of light appears in front of me.
A cat inside the light; it wants my soul,
strips me of my heart so I will not be.

Without its core, my outer shell lies here,

the inner power used to fuel all fear.

— Jonathan Frech, 30th of October 2017
Manuscript.

BMP Implementation in C

2017-10-21, post № 182

C, programming, #bitmap, #file, #file format

C is one cool and important language. CPython and Unix are based on it, the Mars Curiosity rover is run by it and even the GCC C compiler itself is written in C. However, as C is some years old by now, it lacks a lot of higher-level features most modern languages possess, being more down to the silicon, as the cool kids say. Concepts like pointer manipulation, bit fiddling and its string implementation — just to name a few — are at times cumbersome, insecure and error-prone; nevertheless is there a certain appeal to writing in C.

Being only one abstraction level away from Assembly — which itself is only one abstraction level above raw byte code — and having access to file manipulation down to the individual bit, I set out to write a Microsoft Bitmap (.bmp) implementation in pure C. As Microsoft’s standard for this image file format is quite feature-rich, I decided to focus on the bare minimum — a bitmap with 𝟤𝟦-bit color depth (three colors, one byte per), one color plane, no compression, no palette and 𝟥𝟢𝟢 DPI.
My Bitmap implementation supports both reading and writing .bmp files, as well as generating some test images — including a Mandelbrot set fractal renderer, of course. Implementation source code can be downloaded (bmp.c) or seen below.

A Mandelbrot set fractal rendering.

Implementing a file format requires knowing its specification. Although it is not the best article I have ever seen, this Wikipedia article gave me some insights. The missing pieces were reverse engineered using Adobe Photoshop CC and the HxD hex editor.
The following is a snippet of the implementation’s savebmp function (full source code listed below). It illustrates the Bitmap file’s byte layout only showing the file header, omitting a lengthy data part concatenated to the header. S, K, W, H and B are all byte arrays of length four (little-endian format) which contain the file’s total size, the bitmap data offset (which is constant, since the header is always exactly 𝟧𝟦 bytes large), the image’s dimensions (horizontal and vertical) and the bitmap data’s section’s size, respectively.

/*  bitmap file header  */
0x42, 0x4D,             // BM
S[0], S[1], S[2], S[3], // file size
0x00, 0x00, 0x00, 0x00, // unused
K[0], K[1], K[2], K[3], // bitmap data offset
/*      DIB header      */
0x28, 0x00, 0x00, 0x00, // DIB size
W[0], W[1], W[2], W[3], // pixel width
H[0], H[1], H[2], H[3], // pixel height
0x01, 0x00,             // one color plane
0x18, 0x00,             // 24 bit color depth
0x00, 0x00, 0x00, 0x00, // no compression
B[0], B[1], B[2], B[3], // bitmap data size
0x23, 0x2E, 0x00, 0x00, // 300 DPI (horizontal)
0x23, 0x2E, 0x00, 0x00, // 300 DPI (vertical)
0x00, 0x00, 0x00, 0x00, // no palette
0x00, 0x00, 0x00, 0x00  // color importance
/*  data bytes follow   */

Key bytes to note are the first two identifying the file type (the ASCII-encoded letters BM) and the DPI bytes, 0x23, 0x2E, which indicate 0x00002E23 = 11811 pixels per meter in both the horizontal and vertical direction. Converting from pixels per meter to dots per inch results in 11811 / (1 meter / 1 inch) = 11811 * 127 / 5000 = 300 DPI (roughly).
Most values are represented using four bytes in little-endian format. Translating an 𝟥𝟤-bit integer into four little-endian formatted bytes can be achieved as follows.

/* unsigned 32-bit integer */
unsigned int n = 0b10100100010000100000100000010000;
/*                 < m sig><sm sig><sl sig>< l sig> */

/* byte (unsigned char) array of size four */
unsigned char N[4] = {
    (n & 0xff000000) >>  0, // most significant byte
    (n & 0x00ff0000) >>  8, // second most significant byte
    (n & 0x0000ff00) >> 16, // second least significant byte
    (n & 0x000000ff) >> 24  // least significant byte
};

Other than rendering a fractal, I also implemented three nested loops which output an image containing every possible color exactly once ((2**8)**3 = 16777216 pixels in total).

All sixteen million colors in one image.

An image’s data type is implemented as a struct image which contains three variables — width and height, two integers specifying the image’s dimensions, and *px, a pointer to an one-dimensional integer array of size width*height which holds the entire image data.
Defined functions are listed ahead.

image * readbmp(char []); Reads an image specified by a file name. If reading fails, a NULL pointer is returned.
void savebmp(image *, char []); Saves given image to a file with specified name.

image * newimage(int, int); Returns a pointer to an image struct with specified dimensions (image will be filled with nothing but black pixels).
void freeimage(image *); Frees an image struct’s memory.
int getpx(image *, int, int); Returns the pixel color at specified coordinates.
void setpx(image *, int, int, int); Sets the pixel color at specified coordinates.
void fill(image *, int); Fills a given image with a given color (all pixels are set to specified color).

int color(byte, byte, byte); Returns a 𝟥𝟤-bit integer representing a color specified by three bytes (byte is defined through typedef unsigned char byte;(.
int hsl(double, double, double); Returns a 𝟥𝟤-bit integer representing a color specified by three doubles in the HSL color format.

Images shown in this post were converted to .png files as WordPress does not allow .bmp file uploads; the raw pixel data should, however, be identical.^[1]

Source code: bmp-implementation-in-c_bmp.c

Extra assets: bmp-implementation-in-c_post-draft.rtf, bmp-implementation-in-c_wordpress-blog-screenshot-1.png, bmp-implementation-in-c_wordpress-blog-screenshot-2.png, bmp-implementation-in-c_wordpress-blog-screenshot-3.png, bmp-implementation-in-c_wordpress-blog-screenshot.psd

Posts: 285-278, 277-270, 269-262, 261-254, 253-246, 245-238, 237-230, 229-222, 221-214, 213-206, 205-198, 197-190, 189-182, 181-174, 173-166, 165-158, 157-150, 149-142, 141-134, 133-126, 125-118, 117-110, 109-102, 101-94, 93-86, 85-78, 77-70, 69-62, 61-54, 53-46, 45-38, 37-30, 29-22, 21-14, 13-6, 5-1

Jonathan Frech's blog; built 2024/04/13 20:55:09 CEST