There is some bit of information about computers which I was never told in school or at the university. This writing is trying to cover these bits in details for those interested out there.

Processor’s guts

In essence, processor consists of these main blocks:

• instruction decoder, which decodes the operation code (see below) and tells processor what it should do next
• ALU (arithmetical-logic unit) - the device, which performs all the operations a processor is capable of - arithmetic (addition, subtraction, multiplication) and logical (bitwise OR, AND, XOR, NOT; comparing numbers)
• registers are the places, where processor takes the input data for each operation and stores the results of each operations
• data bus - the place, which passes data to or from memory (RAM, peripheral devices and other types of memory)
• address bus passes memory addresses between processor and memory (to be read from or written to)

To best illustrate our examples, consider this much simplified processor model:

It has only three registers, ALU and data/address buses. I will explain all the examples below in reference to this processor.

Workflow

Processor works discretely, meaning it executes commands at certain points in time (as opposed to continuous work, where something works all the time). Those points in time are marked by clock signals (or simply, clocks). Processor is connected to (or has an internal one) clock signal generator, which regularly tells processor to operate. Operation is nothing but a change of the internal processor’ state.

So how does processor knows what to do next? On the first clock signal, CPU loads the command with the help of instruction decoder. On the next clock tick, processor starts executing an instruction (or a command) - it either copies data to / from registers or RAM, or involves ALU. On the next clock, processor increases the command counter by one and hence proceeds to the next instruction from the pool.

In reality, this mechanism is somewhat more complex and involves more steps to just even read the command from the pool. And the pool itself is a unit on a processor chip as well and acts as another register. But to explain the processor operation, I’ve simplified that.

Read more to where I explain how a program in Assembly language is transformed to ones and zeroes; show a sample processor instruction set and how a program in C could be compiled to that instruction set.

This post is based on a talk on IntelliJ productivity and contains some tips and tricks I found handy in my everyday work.

You may disagree with me or find these recommendations useless for yourself, but I and some of my teammates have found out these tips have made us more efficient when dealing with lots of code in a big project.

• say “no!” to tabs – that’s it – disable them. Forever. IntellJ has many ways to navigate the project, just check them out: Cmd+E, Cmd+Alt+← and Cmd+Alt+→

• stop using mouse – just as with navigation in open files, IntelliJ offers awesome code navigation: want to go to the method definition? Cmd+B, want to find a file or class or method? Shift, Shift or Ctrl+Shift+N or Ctrl+N, want to move around in Project Manager? Cmd+1 (and then Esc to switch between editor and Manager)

• use IntelliJ’s code intelligence – if you forgot what params method takes, use Cmp+P, if you want to find variable or class or method usage – use Alt+F7, use the power of “Refactor -> Extract Constant/Field” function

• use action finder popup – that one can save you lot of time when you do not remember where the function needed is located or what key shortcut to use – Cmd+Shift+A is your friend

WARNING: lots of video/gifs under the cut!

Did you ever think of how old games like tetris or snake (hail the Nokia fans!) work?

I always wanted to write a few relatively simple games just to practice my coding skills. Back when I was at the uni, I’ve made peg solitaire and three-in-a-row games in C/C++ with SFML. I even wrote a (pretty shitty) server-client chess game (be warned: this code was written in 2011 and only contains fixes to make it run on a new version of SFML, so it might look… smelly… on each and every line…). My latest achievements were the Entanglement game and a well-designed (at least from my perspective) arcade space shooter.

But I’ve never implemented more well-known games like tetris or snake. And they are interesting from the algorithmic perspective.

And recently I’ve came to the point when I really wanted to create something interesting. And that was the chance to try myself as a early-2000-ies developer!

As I wanted to accomplish my goal as soon as possible, I’ve decided to not pick up any libraries and just implement everything with HTML5 Canvas API. And this is what I ended up with:

In this post I’ll talk about Tetris. Under the cut you’ll find the interesting algorithmic solutions I’ve used for it.

Update

There is a revised version of this blog, showing how an algorithm implementation could be optimized.

The story behind this post

Recently I’ve received +10 karma on StackOverflow. I was curious for what question or answer and clicked to check this. It appeared to be a seven-year-old answer about a Floyd-Warshall algorithm. I was surprised with both my bad English (back those days…) and the very small value the answer had. So I’ve revised it and here it is – the brand-new version!

The definitions

Let us have a graph, described by matrix D, where D[i][j] is the length of the edge (i -> j) (from graph’s vertex with index i to the vertex with index j).

Matrix D has the size of N * N, where N is a total number of vertices in a graph because we can reach the maximum of paths by connecting each graph’s vertex to each other.

Also, we’ll need matrix R, where we will store the vertices of the shortest paths (R[i][j] contains the index of a vertex, where the shortest path from vertex i to vertex j lies).

Matrix R has the same size as D.

The Floyd-Warshall algorithm performs these steps:

1. initialize the matrix of all the paths between any two pairs of vertices in a graph with the edge’s end vertex (this is important since this value will be used for path reconstruction)

2. for each pair of connected vertices (read: for each edge (u -> v)), u and v, find the vertex, which forms shortest path between them: if the vertex k defines two valid edges (u -> k) and (k -> v) (if they are present in the graph), which are together shorter than path (u -> v), then assume the shortest path between u and v lies through k; set the shortest pivot point in matrix R for edge (u -> v) to be the corresponding pivot point for edge (u -> k)

But how do we read the matrix D?

Foreword

Some time ago I’ve published an article about bit O notation. I’ve mentioned its source but never published it. So here it is, the original article.

The article

Big O notation is the most widely used method which describes algorithm complexity - the execution time required or the space used in memory or in the disk by an algorithm. Often in exams or interviews, you will be asked some questions about algorithm complexity in the following form:

For an algorithm that uses a data structure of size n, what is the run-time complexity or space complexity of the algorithm? The answer to such questions often uses big O notations to describe the algorithm complexity, such as O(1), O(n), O(n^2) or O(log(n)).

At work, I’ve had a task to implement a Gantt chart diagram to show dependencies and order of some… let’s say, milestones. Given this feature is in a very unstable beta in Google Charts, I thought to myself: “Why don’t I implement it on my own?”. And tried to recall my D3 knowledge.

I’ve also found a minimalistic, but helpful example / screenshot of some Gantt chart implementation:

The challenges I’ve faced were:

1. order milestones on a timeline
2. scale milestones to fit in a viewport
3. create pretty connection lines
4. center text inside each milestone

And since D3 is a data-driven library, I’ve used map/reduce where possible.

Here’s how the result looked like:

The implementation details are under the cut.

Update August 2020

There are few updates to this original implementation in my new blog.

Entanglement?

Some time ago there was a game popular over there, called Entanglement:

There are a few implementations of this game under Android:

But there was no such game for iOS. And as I pursued my second M. Sc. degree, I have had a course “iOS development”, where we were learning to use Swift and iOS platform.

I decided to implement this game as my course project. In Swift. Originally the game was implemented in Swift 2 (there was no Swift 3 back those days).

And recently I decided to refactor the game a bit and update it to the most recent Swift version (which is Swift 3 at the moment of writing this post).

In this post I’ll describe some interesting decisions I made while creating this game.

libc.js

Recently I’ve written a post about functional programming techniques, coming into the world of front-end and the library I crafted as an experiment. That library, libc.js was highly inspired by Elm and Mithril. But it suffered two major features:

1. components were hardly able to be used in other components
2. the interaction between components was nearly impossible (or, at least, not very transparent)

What’s hidden beneath the next version of the library?

Have you ever asked anyone if assembly language might be useful nowadays? So, here’s the short answer: YES. When you know how your computer works (not a processor itself, but the whole thing - memory organization, math co-processor and others), you may optimize your code while writing it. In this short article, I shall try to show you some use cases of optimizations, which you may incorporate with the usage of low-level programming.

Recently I was reading through my old posts and found out there is a gap in the article about SSE - the post did not cover some of the implementation caveats. I decided to fulfill this and re-publish a new version.

In last couple of years the functional programming paradigm became very popular. A huge amount of libraries, tools, tutorials and blog posts appeared. Although the paradigm itself is rather old (lambda calculus was developed around 1930 and the Lisp language was introduced in 1950), its popularity blew up rapidly somewhere in 2014-2016 and that’s what is happening right now. Probably one of the most powerful influencers, giving FP (functional programming) that thrust is web development. Since Facebook introduced React, the community started incorporating many things from FP with React - including Redux and Immutable.js. But there are much more interesting things which were invented on this wave of FP popularity. One of them is Elm.

This is a story how I implemented invented yet another web framework wheel.