Subscribe to GameFromScratch on YouTube Support GameFromScratch on Patreon
2. December 2012

 

Having too much fun and want to make life a bit duller?

 

If so, I have the perfect site for you!  There is a newly launched game development blog at Dull as Dishwater. It’s extremely young at this point in time, with exactly two posts, one of which is the introduction post. 

 

image

 

Why then am I bothering to point it out?  That would be because the premise behind the blog is absolutely brilliant ( and boring! ).  He intends to cover the nuts and bolts “boring” parts of game development.  Game development blogs tend to focus heavily on the “exciting” parts of game development… graphics, audio and pushing the boundaries of the meaning of the word exciting, algorithms.

 

DullAsDishwater on the other hand is going to focus on the boring minutiae that go in to every game.  Currently the only technical post covers dealing with mouse cursor sensitivity across display resolutions.  Not the most exciting thing, but one of those things you will have to deal with.

 

So here’s hoping he continues to keep up bored with new updates in the future.

Totally Off Topic


14. October 2012

 

I don’t think there are many people out there that would argue that JavaScript isn’t an important language.  It certainly is, and with the recent death of Flash and Silverlight in the browser and the hype behind HTML5, to say nothing of Node, it is actually getting more popular

 

I also don’t think there are many people out there that would argue that JavaScript is a heavily broken language too.  As systems written in JavaScript become larger and more complex, the language flaws become more and more pronounced.  Just as C++ ultimately came about from C’s (arguable by some) inability to deal with complexity, a number of companies and groups are looking at possible successors to JavaScript.  Generally all of these languages are backward compatible with JavaScript in some form.  Let’s take a look at the contenders for the short term future of JavaScript.

 

 

ECMAScript 6 AKA Harmony AKA ES.next

 

When talking about the future of JavaScript, the most logical place to start is … well, the future of JavaScript, the next version.  The spec is a work in progress and you can read it here, the most recent revision as of writing (September 27th, 2012).  It would seem obvious that the next version of JavaScript would be ECMAScript 6, but don’t bet the farm just yet.  After all, ECMAScript 6 is effectively a scaled back version of ECMAScript 4, while ECMAScript 5 (the JavaScript version in common usage today), is essentially a fork of ECMAScript 3.  Firefox and Chrome already support bits of ECMAScript 6, so it’s future is quite bright, but by no means are things certain.  Plus there is no way of knowing how long until ECMAScript 6 is supported enough to actually make it useful.  For an idea of where ECMAScript support in various browsers is, check out this handy table.

 

This has opened the door to a number of alternatives, many are languages that compile down to JavaScript making them useful today. Others (like Dart) compile down to JavaScript as a fallback, but perform better in their native environment.

 

 

The 900lb gorillas.

 

Two of the companies are very interested in the future of JavaScript are Google and Microsoft.  Each has created a language they hope will be the future of JavaScript, Dart and TypeScript respectively.

 

This interview between Anders Hejlsberg ( Microsoft, creator of Delphi, C# and now TypeScript ) and Lars Bak ( Google, creator of V8 JavaScript engine and Dart ) is very good summary.  It allows each developer to describe their opinions of the existing problems with JavaScript ( on which they largely agree ) and describe the strengths of each of their companies offerings.

 

Anders Heljsberg & Lars Bak on TypeScript/Dart and the problems with JavaScript today

 

 

Google Dart

 

http://code.google.com/p/dart/ or http://www.dartlang.org/

image

 

In Their Words:

Why Dart?

At Google we've written our share of web apps, and we've tried in many ways to make improvements to that development process, short of introducing a new language. Now we think it's time to take that leap. We designed Dart to be easy to write development tools for, well-suited to modern app development, and capable of high-performance implementations.

 

Dynamic or Static Typed?

Dynamic

 

Key language features

Compiles down to JavaScript, or to the Dart VM ( which is currently only available in Chrome ).  Code running in the Dart VM will perform better.

Language is basically JavaScript with prototypes removed and classes added.  They’ve added lexical closures, sane this management, optional static types, named parameters, cleaner event and DOM programming and more.

The language also comes with an IDE, the Dart Editor, which is a mash-up of Eclipse, Chromium and the DartVM.  For those like me that prefer WebStorm/IntelliJ over Eclipse, they also have a plugin available.  There used to be a cloud based editor, but it appears to have been retired.

 

My 2 cents

Dart technically solves most of the problems with JavaScript and is compatible with existing JavaScript.  That said, I checked out Dart shortly after it was released, then again 6 months later and as usual, Google’s developer support stunk.  They just don’t put the extra effort in to making the experience easy on developers ( at the time, the Windows build didn’t work, you need to compile everything yourself, etc… this lack of polish is just as common with other Google developer tools such as GWT, NativeClient even Android), but the package seems to be improved quite a bit.  Just a warning though, if you aren’t experienced with Google developer support, be prepared for a ton of frustrations if you run in to any problems.

Probably the biggest flaw with Dart is going to be the lack of adoption.  Dart compiles to JavaScript as a fall back, but the VM is obviously the preferred platform.  Google announced their intention of submitting Dart to a standards body, but do you see Microsoft or Mozilla adopting Dart?  Yeah, me neither.  That said, Chrome is getting a pretty big install base.

My biggest problems are the above mentioned Google support, and their tendency to kill off projects at the drop of a hat lately. 

 

 

Microsoft TypeScript

 

http://www.typescriptlang.org/

 

image

In Their Words:

Why TypeScript?

TypeScript is a language for application-scale JavaScript development. TypeScript is a typed superset of JavaScript that compiles to plain JavaScript. Any browser. Any host. Any OS. Open Source.

Dynamic or Static Typed?

Static

 

Key language features

TypeScript is compiled down to JavaScript.

Language allows you to add static types to JavaScript, allowing you to catch type errors during the compilation process ( although the compilation will still generate a JS file ).  Allows you to mix and match TypeScript and JavaScript in existing files.  Adds new constructs to the language such as classes, interfaces and modules.

IDE support is available in the form of a Visual Studio 2012 plugin, as well as plugins for Sublime Text, Vi and Emacs.  They also have Node support out of the box and you can download TypeScript using NPM.

 

My 2 cents

I will admit, I am a bit of a Anders fanboy, so that has gotten me excited about this language.  Language design is all about pragmatic choices, and Anders just seems to pick the right compromises.

At the end of the day, TypeScript is less of a gamble than Dart, and this has it’s advantages and disadvantages.  As Lars said in that video above, TypeScript is “the safe bet” and that is a pretty apt description.  It is a layer on top of JavaScript and nothing more, in many ways it is quite similar to CoffeeScript.  In the end though, TypeScript is probably the one that will best integrate with existing Javascript and given the massive variety of existing code, that is probably a pretty big advantage.  At the end of the day though, TypeScript is ends up being JavaScript, so don’t expect any performance improvements.

The tooling support is shockingly not Microsoft.  Would you have ever expected Vi and Emacs support from day 1?  What about Node support?  There is a noted lack of documentation at this point, which is shocking for a Microsoft product.  They do however have a nifty Playground/Tutorial environment that allows you to type TypeScript and see the generated JavaScript.  Sadly for now, comments and whitespaces are mangled as part of the process, although this is a temporary flaw.

 

 

CoffeeScript

http://coffeescript.org/

 

In Their Words:

CoffeeScript is a little language that compiles into JavaScript. Underneath all those awkward braces and semicolons, JavaScript has alwaysimage had a gorgeous object model at its heart. CoffeeScript is an attempt to expose the good parts of JavaScript in a simple way.

The golden rule of CoffeeScript is: "It's just JavaScript". The code compiles one-to-one into the equivalent JS, and there is no interpretation at runtime

 

Dynamic or Static Typed?

Dynamic

 

Key Language Features

Coffeescript is a whitespace ( as opposed to curly brace ) based language that compiles down to JavaScript.  It is certainly the elder statements of the group and has been around for a number of years.

Coffeescript provides a number of features to JavaScript including a class system, lexical scoping, splats (…) for variable arguments, array slicing and more.  Additionally it provides a layer of syntactical sugar over some of the JavaScript warts such as is instead of === or unless as the inverse of if.

Coffeescript also offers impressive tooling support.  Like TypeScript, it is available as a Node utility, making installation via NPM trivial.  Most IDEs support Coffeescript ( IntelliJ, Eclipse, WebStorm, Netbeans, Sublime Text, etc ) out of the box.

 

My 2 cents

I only gave CoffeeScript the smallest of evaluations and that was a few years back.  You see, I have an irrational hatred of whitespace based languages.  What I found most irksome ( and TypeScript fixed this ), is that it’s a language on-top of JavaScript, not within Javascript.  So you had to program things the Coffeescript way, then compile it to JavaScript.  Coffeescript felt like a completely different language than JavaScript, while neither Dart nor TypeScript gave me that impression.  Mostly though, I never got over my hatred of whitespace based languages.

All that said, CoffeeScript solved many of the problems that Dart and TypeScript are setting out to solve, and it solved them years ago.  It’s by far the most mature and as a result had the most time to iron out the kinks.

 

 

Other Options

 

Dart, TypeScript and CoffeeScript are by no means the only options when it comes to more programmer friendly JavaScript programming languages, but they are the three that are closest to JavaScript itself.  The following are a selection of other languages that will compile to JavaScript.

 

 

GWT – Google Web Toolkit

 

https://developers.google.com/web-toolkit/

Google Web Toolkit (GWT) is a development toolkit for building and optimizing complex browser-based applications. GWT is used by many products atimage Google, including Google AdWords and Orkut. It's open source, completely free, and used by thousands of developers around the world.

Basically GTW is a Java –> JavaScript compiler.  That is over generalizing, but basically you write your code using a subset of libraries in the Java language, and it is compiled down to JavaScript.  Or a bit of mix and match, with Java running on the server and JavaScript on the client.  The programmer never really works in JavaScript when working with GWT.

 

As is pretty typical with Google projects, IDE support is mostly Eclipse based.  There are a few published books available for GWT.  GWT has been around for quite a while and integrates nicely with Google App Engine.  That said, if you aren’t a Java developer, GWT really isn’t for you.

 

 

 

Haxe

 

http://haxe.org/

Haxe is a powerful modern language with many compelling features. It is aimed at giving developers a tool to create websites & applications using aimage single unified programming language. Whether you use Haxe for its cross-platform features, or focus on a single platform, there are many reasons to adopt it.

Haxe is a language that allows you to compile down to JavaScript, Flash, NekoVM, PHP, C++, C# and Java.  It provides a standard language and library that compiles down to the mentioned languages, while filling in the missing pieces of each language, such as packages and typing for JavaScript, reflection for C++, etc.  The Haxe language itself is derived from ActionScript 3, which itself is a ECMAScript language. Haxe has pretty good IDE support. Haxe also has a few published books which is always handy.

 

I’ve not yet used Haxe, mostly because I have no prior ActionScript experience and because I never really had the time to look into it further. That said, HaxeNME, which is a gaming library targeting web, mobile and desktop certainly has piqued my interests.

 

 

 

Other other options…

 

This doesn’t even come close to listing all of the languages that compile to JavaScript, for a more complete list see this.  This does however cover the major players, at least as I see them.  If you think I made a major omission or mistake in this list, please let me know!

 

JavaScript is becoming more important and JavaScript programs are becoming more complex.  The language does have a number of flaws, but fortunately there are a number of options to help you cope, at least until ECMAScript 6 arrives.

General Programming


27. August 2012

 

As the title suggests, Blitz Basic Plus is currently free right now, marked down from the regular price of 60$.  bbasiclogo

 

Blitz Basic, in their own words, is:

BlitzPlus provides developers with a complete 2D programming solution for the PC. Utilising a BASIC/C hybrid language for use with its compiler, BlitzPlus is capable of creating any type of 2D program, with minimal development-time and with minimal hassle. Just run the included IDE (or use one of your own), type your code and your Windows executable file will be created, ready to be run on any computer running Windows 95/NT 4 or later.

 

While the Blitz Basic language is:

 

BlitzPlus' language is a BASIC/C hybrid. As well featuring the classic BASIC constructs such as goto and gosub, it also features C influences such as functions and types. On top of this, BlitzPlus' command set features over 500 commands. This command set is highly streamlined meaning complete programs require very few lines of code indeed.

 

Apparently the original Worms was developed in BlitzBasic, as was the Eschalon RPG series.  I personally have never tried Blitz and don’t really intend to as I have never really been a fan of BASIC, but it is certainly worth checking out, especially for the price of 0$.

 

You can purchase it here.  I have no idea how long the offer is on for.

News


3. July 2012

A poster, JakobProgsch, over on the TigSource forums has made a series of OpenGL samplesopengllogo available on GitHub. Each example is contained in a single file, doesn’t depend on any non-canonical libraries or hide code behind a framework.  As a results the examples are a bit longer, but easier to follow.

 

The following topics are covered ( description taken from each file header ).

 

Sample 1

Skeleton code that all the other examples are based on

 

Sample 2

This example shows the setup and usage of shaders and a vbo+vao

 

Sample 3

This example shows the setup and usage of shaders and a vbo+vao
In this example the attrbute locations are set via BindAttribLocation
instead of layout qualifiers in the shaders

 

Sample 4

same as the "Shader and VBO" example, only with an indexed vbo.

 

Sample 5

apply a texture to the fullscreen quad of "Indexed VBO"

 

Sample 6

set up a perspective projection and render a rotating cube

 

Sample 7

render the cube from the perspective example to a texture and

apply fxaa antialiasing to it.

 

Sample 8

create 8 instances of the cube from the perspective example

with an additional offset buffer and AttribDivisor

 

Sample 9

create 8 instances of the cube from the perspective example

the difference to the instancing1 example is that we are

using a texture buffer for the per instance data instead of a

vertex buffer with divisor.

 

Sample 10

create 8 instances of the cube from the perspective example

the per instance data is passed with a uniform buffer object

 

Sample 11

Uses a geometry shader to expand points to billboard quads.

The billboards are then blended while drawing to create a galaxy

made of particles.

 

Sample 12

This example uses the geometry shader again for particle drawing.

The particles are animated on the cpu and uploaded every frame by

mapping vbos. Multiple vbos are used to triple buffer the particle

data.

 

Sample 13

This example simulates the same particle system as the buffer mapping

example. Instead of updating particles on the cpu and uploading

the update is done on the gpu with transform feedback.

 

Sample 14

This example renders a "voxel landscape/cave" from the view of a

moveable camera. Occlusion queries and conditional rendering are used

to cull occluded parts of the world and timer queries are used

to measure the performance.

 

Hope he keeps them coming, good OpenGL resources are scarce.

 

Edit: 7/5/2012 Fixed links and added 14th sample, which by the way, is a bit of a monster.

Programming


23. March 2012

 

 

Have you ever read a technical book and thought to yourself “Man, the author of this book must be a gigantic jerk!”?  I can honestly say, I never have until now!

 

 

This review of sorts covers two different Javascript books I’ve read recently; two very different books with two very different goals.  As I have been working with Node and more recently Appcelerator, I decided I needed to polish up my Javascript skills, as things are much different these days than the DOM manipulation I traditionally used Javascript for.  I started out by searching for “Javascript closure” on Safari Books Online and it brought me to Javascript: The Good Parts, it was short and I was suffering insomnia so I read it that evening.  The book left me with so many questions that I ended up reading/contrasting it against Javascript: The definitive guide.  Aside from both books being about Javascript and both having colons in their titles, these two books couldn’t be further apart!

 

 

 

Let’s start with the Javascript: The Good Parts review.

 

 

First off, it’s worth mentioning the author, Douglas Crockford.  If you’ve never heard of him, he is the guy originally responsible for drafting the JSON standard and Yahoo’s head JavascriptJavascriptTheGoodParts guy, responsible for projects like the yUI library.  Simply put, this man knows his stuff when it comes to the topic.  Sadly, he knows it too and doesn’t let you forget it.  Many times in the book he pushes his opinion as fact, and often on very odd things ( like never use ++ operators ).  He often makes sweeping statements that really don’t work in a black and white world.  That said, if you can get over the personality ( or you aren’t sensitive to such things ), this is a wonderful little book.

 

 

I do mean little too, as this book weighs in at a mere 176 pages.  Amazingly enough there is a ton packed in to those 176 pages and even more amazingly, a great deal of it could be pruned and would make this a superior book ( such as the regular expressions chapter ).  A good example of what could be pruned out completely are the absolutely bizarre charts this book is littered with.  Here is one for example:

 

 

whitespace

 

 

That chart was meant to explain Javascripts whitespace rules… did it help you, even slightly?  You could have more concisely and effectively conveyed all of that information in a sentence or two.  I found that over and over, the author being overly clever or complicated leaves the reader baffled on what should be a pretty simple concept.  He does the exact same thing with code examples too, consider this code sample regarding inheritance:

 

var Cat = function (name) { this.name = name; this.saying = 'meow'; }; // Replace Cat.prototype with a new instance of Mammal Cat.prototype = new Mammal( ); // Augment the new prototype with // purr and get_name methods. Cat.prototype.purr = function (n) { var i, s = ''; for (i = 0; i < n; i += 1) { if (s) { s += '-'; } s += 'r'; } return s; }; Cat.prototype.get_name = function ( ) { return this.says( ) + ' ' + this.name + ' ' + this.says( ); }; var myCat = new Cat('Henrietta'); var says = myCat.says( ); // 'meow' var purr = myCat.purr(5); // 'r-r-r-r-r' var name = myCat.get_name( ); // 'meow Henrietta meow'

 

 

Notice how half of it is taken up by a purr() method which is overly convoluted and has nothing to do with the lesson at hand.  It simply causes the reader to have to dedicate a few more cycles to deciphering an example ( which is completely irrelevant to the topic at hand ), while allowing the author to feel just a little bit more clever. 

 

These kinds of things happen over and over and it makes reading the book more difficult than it should be, for no good reason.  The other major problem with trying to decipher the code samples in this book are the errors, my god the errors!  See every technical book is going to have a certain number of errors, but for a 176 page book, this one has far more then it’s share.  Just take a look at the errata.  Nothing makes learning a new concept more difficult than following code that doesn’t work!  When you are new to a subject, you assume the code you are working from is correct and it’s your newbishness that is the source of the problem.  That often simply isn’t true with this book, when you read it, if you encounter problems be sure to check the errata before you bash your head off the wall with frustration.

 

 

Next up is determining who this book’s audience is, which is a problem the author obviously struggled with.  Consider the following two paragraphs taken from the preface:

 

This is a book about the JavaScript programming language. It is intended for            programmers who, by happenstance or curiosity, are venturing into JavaScript for the   first time. It is also intended for programmers who have been working with JavaScript at a novice level and are now ready for a more sophisticated relationship with the language. JavaScript is a surprisingly powerful language. Its unconventionality presents some challenges, but being a small language, it is easily mastered.

 

Followed a paragraph later by:

 

This is not a book for beginners. Someday I hope to write a JavaScript: The First Parts book, but this is not that book. This is not a book about Ajax or web programming. The focus is exclusively on JavaScript, which is just one of the languages the web developer must master.

 

Let me put this in the simplest terms possible This is not a book for beginners!  If you are “venturing into Javascript for the first time”, well frankly you are absolutely screwed if this is the book you chose!

 

 

So I’ve dumped on this book pretty good, it’s short, the author comes across arrogant and I seem to hate at least half of it, so why the hell am I bothering to mention it to you?  That’s because the remaining half of this book is downright excellent.  It basically is the Javascript equivalent of Effective C++, which is one of the best technical books ever written.

 

 

Javascript as a language is an absolute minefield waiting to explode killing ( or at least mentally scaring ) the developer.  It lacks the data hiding facilities that basically every other language in existence has.  More so, it has extremely dangerous but seemingly benign constructs like new and == that are just waiting there to attack you.  Finally, the scoping rules are almost as byzantine as you’d expect from C++! It is far too easy to make a variable global, or mute out an existing variable leading to undesired behavior ( bugs ).  This book helps you to navigate that minefield, probably better than any other book in existence.

 

 

More to the point, most people simply use Javascript wrong ( now there I go sounding arrogant! Winking smile ).  But simple fact is, years of browser based DOM manipulation code written to mimic traditional OO code has lead to many people forming extremely bad habits, myself included.  This book does a very good job of showing you the intricacies of the language, as well as effective work arounds for perceived faults.  Some of these “faults” I disagree with, but I certainly did learn from the exercises anyways.

 

 

So in the end, do I recommend this book?  That’s a tricky one.  I really don’t like the way it is written, I am not keen on the way he teaches ( the choice of illustrations and code examples ), the excessive errors certainly don’t help.  That said, I learned a ton reading this book, in a very short period of time.  Besides, it’s less than 20 bucks!  So in the end, yes I do recommend you read it, much like Effective C++, it will make you a better Javascript programmer.  But for learning the language or as a first book, certainly not.  Fortunately in the regard, we have another book!

 

 

 

Javascript: The Definitive Guide review.

 

 

If Javascript has a bible ( like K&R C for the C Programming Language ), it’s this book.  The Definitive Guide is a much more traditional book, much longer at 1032 pages and with a much more traditional price tag.  This part of the review is going to be much shorter, as essentiallyJavascriptTheDefinitiveGuide there is a lot less to say.  The authors writing style in much more approachable than Javascript:The Good Parts and the code examples are far easier to digest.

 

 

One up or downside to this book, depending on your needs, is the amount of browser/DOM related content.  While Good parts effectively had no content pertaining to using Javascript in the browser ( an area extremely well documented in books and online ), the same is not true for this book.  Fully half of this book is on “client side” Javascript programming.  Then again, even at 50%, this is still a 1000+ page book, so you still got another 500 pages of non-client side Javascript.  Just remember, if you are using Javascript outside of the browser, if it starts with document. it probably isn’t supported.

 

 

The book is basically broken down into 4 sections.  The first is Core Javascript, which covers the language itself, regardless to how it is used.  This section is where you cover concepts like reserved words, objects, arrays etc.  The second is Client Side Javascript, which basically covers using Javascript in the browser, covering topics such as the DOM, jQuery, HTML5, etc.  The final two sections are reference sections for both the core and client side libraries.  These sections are strictly for reference.  Truth is though, with the sad state of the internet on this subject, a handy book reference is quite a useful thing.

 

 

For learning the language and your goto reference book, it’s really hard to beat Javascript: The Definitive Guide.  Taken in tandem, these two books are all you would need to succeed at Javascript development.  The Definitive Guide to get you started, and as a reference as you go along.  Then read Best Parts, to make you a better developer… if you are willing to put up with some annoying attitude that is!

Programming


AppGameKit Studio

See More Tutorials on DevGa.me!

Month List