I recently got my digital hands on a review copy of Packt Publishing’s new book Cinder – Begin Creative Coding by Krisjanis Rijnieks, the first of what will hopefully be many books centered around Cinder.

The book begins with a brief description of what creative coding is and eases into how Cinder fits into the equation. Before jumping into any actual code examples, the author does his due diligence and walks through how to get Cinder downloaded and set up on both OSX and Windows, showing how easy it is with the release version of Cinder. To get your first glimpse into seeing Cinder in action, the book guides you through some of the included samples. It’s an exercise that every first time Cinder user should go through anyway and continues to be a good resource when playing with a feature for the first time.

This is the first tutorial in a series designed to help Flash developers transfer their skills to the C++ creative coding framework Cinder. I’ll provide links to sample code at the end of the post. There’s a lot to learn, so lets get started.

Timing events to happen in a repeating loop is a common technique used in games and applications. It’s commonly used in games when for instance, you want a missile to attack you’re hero every 2 seconds or in an application when you want to remind the user where to click if they haven’t touched anything in a few minutes. I recently used a timed event loop for an installation that required an image transition every 30 seconds.

I’m about to start putting together a series of short tutorials dedicated to helping Flash devs get up to speed with using Cinder. Since there seems to be a wave of Flash developers making this move, I wanted to share some useful bits of code that are a result of research that I’ve done myself and put together into something that someone else might find useful. Cinder can be an intimidating environment to get going with, especially if you’ve never touched C++ at all before (as was my experience). My hope is to try and take some of the intimidation out of the process and maybe even save some of you some time so that you can go ahead and start using those flash skills to make some cool shit in Cinder too.

If anyone has requests for things you’d like to know how to do with Cinder, let me know. I hope to put one of these out every couple of weeks. The first one should be out shortly.

This post is really for myself so that I have somewhere that I can look this up again. Maybe it’ll be helpful to someone else in to the future too.

When working in flash (which I find myself doing less and less now a days), one of my favorite ways to work with assets is through the use of swc files. That’s what they’re there for anyways right? Well what’s not so obvious is how to use swc files to load assets at runtime without using the import command. There are times that it’s useful to refer to an asset at runtime by name that you may or may not use, so you don’t want to explicitly go through the hassle of writing the lines to embed it. This often happens with sounds or items that are easier to loop through to use, such as when using a map of the US. It’s much easier to write a loop vs writing 51 embed statements (yes, I know that there are a million different ways of doing this without writing all of that).

When you do something like going through a loop and saying getDefinitionByName(“state_”+i) to grab a library item that resides in the swc asset file, there are no compiletime errors because Flash Builder can find that swc in the library. But once it runs, the debugger will yell at you. To fix this, you need to embed the swc file at compile time. In Flash Builder, this is done in the Actionscript Compiler settings and all you have to do is add this line (replacing it with the path to your asset swc file obviously)

-include-libraries /assets/assets.swc

Now when you run your swf, it’ll work like a charm and you’ll have those assets available without having to import them in the head of your class file. That being said, be sure to only do this when it makes sense. This embeds the swc file into your swf. So if there are assets in the library that are outdated and set to export at compile time, they’ll be included adding extra senseless bloat. As I write this, I’m using this method for a prototype where bloat doesn’t really matter, but just be mindful of that.

I submitted my first ever JS1K competition entry. Really, my first javascript or any kind of programing contest entry. And I’ll admit that it was quite a challenge too. It was like a putting together a jigsaw puzzle that you have extra pieces to. For my entry,  I decided to take something that I worked on previously and rework it into something that fit the theme and into the 1024 byte limit. The theme for this year was love, so what I did was reworked the spider webs explorations that I had previously worked on into a heart shape and simplified it a bit. I also stripped out anything that wasn’t pure javascript. In the process, I was able to reduce my demo from 25K (not including the processing and toxiclibs js libraries) down to 1022 bytes.

For a recent motion graphics project that I was working on, I needed to use a sucking-in vacuum effect where some logos were being sucked into a TV. Not being a motion graphics expert and jumping into after effects for the first time in a while, I first went searching  some forums to see if anyone else asked the question and had gotten answers. Indeed people have asked how to get this affect, but not too many useful answers were out there. Fortunately, I was able to put together something I was happy with and wanted to show anyone that’s interested how the effect was achieved.

Here’s an example of what it will look like:

Have you ever completed a flash project and started testing it online only to realize that the swf is looking for assets either in a different directory than the one you specified? Or that the folder that you have all your media stored is not in a relative directory like it was when working on it locally?

The solution that I mustered up in the past was to create a flashvar called something like “mediafolder” and put that in the embed code. From there, I would go back into my actionscript code and make sure all external media had the passed “mediafolder” variable appended to the beginning of the url for each external asset and then republish my swf.

This always worked well, so I usually plan for this, but my latest project didn’t have that built in. When it came to testing and we realized that the media folder was going to be in a different location than the base swf file and it was supposed to launch soon, it was time to see if there was another alternative to the usual solution. That’s when I discovered the “base” attribute in the swf object embed code.

The “base” attribute is actually an href attribute that specifies a base URL for all relative URLs on a page. Traditionally, you should be able to put it in the head of an html page and have all links on a page link relative to the string specified in the base attribute. For this it’s just specified for the swf, which picks it up when passed as a param and picks up any external asset from that folder by default instead of the folder that the actual swf is in.

Here’s a sample of the swfobject embed code:

<script type="text/javascript"> 
	var flashvars = {};
	var params = {"base": "../project/subfolder/"};
	var attributes = {id: "Main", name: "Main"};
	swfobject.embedSWF("MainLoader.swf", "myAlternativeContent", "920", "550", "10.0.0", "expressInstall.swf", flashvars, params, attributes);
</script>

With the example above, any asset that the main swf is trying to load relative to itself can now be placed in “../project/subfolder” without having to change any actionscript and republishing your swf. This is a tip that has saved me hours of work and would have saved me even more time had I discovered it in the past.

I feel like I’ve run into this issue so many times and either fixed it by accident or came up with some other solution in order to do the same thing. This time I cam to a conclusion that I understand and hopefully other people can use.

The Problem:

You create an flash app for the web which loads in data from a variety of other external sites, such as thumbnails from an aggregated feed using a 3rd party API and it works swimmingly while you’re testing locally. As soon as you put it on a remote server, you get a Flash Player security error, perhaps something like:

SecurityError: Error #2122: Security sandbox violation: LoaderInfo.content:http://staging.site.com/project/Main.swf cannot access http://www.website.com/image.jpg. A policy file is required, but the checkPolicyFile flag was not set when this media was loaded.
    at flash.display::LoaderInfo/get content()
    at com.gregkepler.testProject.imageloader::ImageLoader/onImageLoad()

Depending on if you’re ready for this error or not, you may not get a popup and can fail silently.

Why This Happens:

This is happening because you don’t have your policy files set up on your server and your swf is “illegally” trying to access data from another domain. This is actually a good thing since it stops potentially dangerous files from doing bad shit to your project/server/network/computer. But you want to ALLOW access to these different domains since that’s a big part of your project, right? The solutions are actually very easy to implement, just sometimes a pain in the ass to figure out.

The Solution(s):

Depending on the project and how wide-spread and prominent it will be on your website and if other projects will need the same sort of access on your server, you may want to allow the entire domain access to the external domains or just one individual project.

For an entire domain

All you have to do is place a crossdomain.xml file is the root of the server where the project will reside. The crossdomain file will look similar to this:

<cross-domain-policy>  
<allow-access-from domain="*" />
</cross-domain-policy>

Replace the * with a specific domain if you don’t want to allow access to all sites. This works because the flash player automatically looks for a crossdomain.xml file when it is asked to load any file in a different domain. It checks to see if it should allow the file based on which domains are specified in the file. If it doesn’t see this file or a domain isn’t listed, or * isn’t specified, it will throw an error and won’t allow the file to be loaded into your app.

For a sub-domain

Sometimes the above solution isn’t appropriate, especially for a website that can potentially house many different projects that have different rules about what is can and should allow. This solution involves a few extra steps, but is still pretty easy.


Step 1:
You still need to add a crossdomain.xml file to the root of your server, but instead of allowing all domains to be accessed, you want to tell it that it should allow crossdomain policy files to be set elsewhere on the server. This is what the crossdomain.xml file would look like instead:

<cross-domain-policy>
<site-control permitted-cross-domain-policies="all" />
</cross-domain-policy>

Step 2:
Create a crossdomain file somewhere else on the server that makes sense, such as where your project exists. This will look a lot more like the original file. In fact it’s exactly the same, unless you want to change the domains that you want to allow for this specific project.

<cross-domain-policy>
<allow-access-from domain="*" />
</cross-domain-policy>

Make note as to where you are putting this file, since you will be specifying this actual xml file as the one to look at for cross domain access (ie: http://staging.site.com/project/crossdomain.xml).


Step 3:
The next 2 steps involve actually adding some as3 code and republishing.
Specify the policy file that should be loaded instead of the one in the root. This should be done before trying to actually load anything. I normally place in the document class as one of the first things that should be done.

Security.loadPolicyFile('http://staging.site.com/project/crossdomain.xml');

Don’t forget to import flash.system.Security.


Step 4:
Go to where you are trying to load the external data and create a LoaderContext instance.

var context: LoaderContext = new LoaderContext();
context.checkPolicyFile = true;

As noted from adobe’s website, what checkPolicyFile does is specifies whether a cross-domain policy file should be loaded from the loaded object’s server before beginning to load the object itself.

Then you will pass that context as an argument when you go to actually load the file, as seen below.

imgLoader = new Loader();
iimgLoader.contentLoaderInfo.addEventListener(Event.COMPLETE, onImageLoad, false, 0, true);
imgLoader.load(new URLRequest(thumbUrl), context);

This article got me on the right foot: http://stunnedgrowth.livejournal.com/4540.html
More about LoaderContext: LoaderContext