As I work with him, I’ve seen a few of the things he’s brewing for the session and some of them look cool as. So get along to it. After all, who has even heard of these Branden Hall and Mike Chambers characters that he’s up against?

Three weeks of taking it easy and I’m ill suited to the tasks that lay ahead over the next three days before I get back to work.

First off, I have to download some four hundred photos of weddings, lakes, food, cathedrals, statues, paintings, people I vaguely know, gondalas, canals, etc and work out where they were taken and why. At the same time, I’ll have to sort wheat from chaff and, because some will be passed on to the wedding party we were out there with, I will have to kick start good ol’ Photoshop and make a little better those pictures worth tampering with.

Nice to know then, that Windows 7 doesn’t even recognise my Canon EOS 350 camera. I know, it is at least 3 years old and thererfore should be resigned to the scrap heap of modern digital consumerism, but I’m perfectly happy with it and I’m a hoarder like that.

For those who suffer from the same lack of acknowledgement between said popular camera and said popular operating system (oh, that evil corporations would have better communication with one another), a handy tip, gleemed from Google:

Press Menu > select the last menu > set Communication to Print/PTP.

Then plug in the camera and the two items act as if they were childhood sweethearts reunited after many years.

http://bit.ly/bUd9XF

I had already started looking at this Native Client technology announced by Google and trying to work out what it could be used for – what the advantages were – but the possibility of playing Unity games without a plugin and directly through a browser (admittedly just a Chromium build at the moment) seems pretty cool.

Now, if all the other browsers were to take up Native Client then we’d have  a truly 3d game-friendly web and with Unity 3 right around the corner things would get exciting!

This makes it even easier for indie and hobbyist game developers to get into Unity and I think it is a good move by the Unity team. They are obviously getting enough progression from Indie to Pro to make this model viable. At the same time, I’m a bit disgruntled as I forked out for an Indie licence in the dark old days when these things cost money.

Unity 2.6 comes with some good features added: http://unity3d.com/unity/whats-new/unity-2.6. There aren’t many better IDEs around and all the features – particularly better performance, source control compatibility and the profiler – seem useful and should boost productivity. I think Adobe could take some pointers – especially for integrating the 3D features of flash player 10 into the IDE. How about, rather than gimmicky additions like the iPhone stuff concentrate on making the IDE a better experience to work with (faster, less buggy) and adding features that will be used day-in, day-out.

Unity 2.6 ftw.

Now, I might be the only Flash developer who doesn’t own either an iPhone or iPod Touch, but is this a big deal? A lot of developers are already making iPhone apps, either with Unity iPhone or objective-c. Although the iPhone market may be a profitable one it is also saturated and a million more Flash developers joining in isn’t going to make that any better.

Also, this isn’t Flash in the iPhone browser, which I think would have improved the iPhone and Flash, this just allows another way of making iPhone applications. There’s already Mono, Unity and objective-c, so Flash is going to open this platform up to some more Flash people and maybe help Adobe leak some of its market but I wonder if this might be another example of feature bloat from Adobe. I’d prefer to see this sort of thing as a plugin or separate app in the Adobe CS5 package not as part of the Flash IDE.

Flash on the Beach was, yet again, a most enjoyable event. The three days of networking with fellow Flash designers and developers and learning some new tips, tricks and techniques from the gurus who frequent these events left me with some ideas and directions that I’d very much like to pursue.

Of real note, for me, were sessions by Joa Ebert, Mario Klingemann and Ralph Hauwert. I picked up some good tips elsewhere but felt I already knew a lot of what I was hearing, whereas these three individuals continue to broaden the horizon of what flash is capable of.

Joa Ebert’s session was the first time I’ve seen a standing ovation at one of these events and the guy stands out as a true master of what he is doing. He knows flash and its compiler inside-out and has used that knowledge and a great deal of research into optimisation techniques to create TAAS a program that streamlines and optimises the bytecode of a flash movie. This allows use of the Alchemy memory codes that improve the speed of some operations in flash. These have only been available via Alchemy and Haxe up until now. Joa was getting some significant speed improvements thanks to the project but other features include automatic UML generation and compiling Java or C# in to flash. Check out Joa’s blog for more on his fantastic work.

Mario Klingemann gave a typical session proving that math experimentation and play can yield great results for experimental generative art work. It’s good stuff and he explains it well. Also, he doesn’t seem to fall into the sub-philosophical or spiritualist nonsense that surrounds a lot of art.

Ralph Hauwert gave a very interesting session on the state of 3D and real time graphics in Flash. With Joa’s TAAS and the improvements in speed brought about by the native 3D apis in Flash, examples from Ralph and the Away 3D team are starting to look half way decent. The triangle culling is much better, shading is becoming a realistic possibility and the number of polygons is reaching above a few thousand. Flash isn’t at the Unity3D standard and won’t get there until hardware rendering is supported but you can do more than in the past.

The other big news from Ralph was his departure from the Papervision team, this seems a big blow for that project and leaves open several questions as to what will happen to Papervision3D X.

Those sessions left me impressed as to the progress Flash and its community is making. Lets hope the upcoming Adobe MAX conference brings even more exciting news about the platform we work on and build for.

This is the fourth installment of the conference and it will be my third visit. The first time I went, I was an isolated flash developer and still fairly wet behind the ears. The experience proved both inspiring and very educational. I’d like to think my skills as a developer have improved immensely since that time but I still look forward to learning from the masters of the medium.

I haven’t decided my exact schedule yet but here are a few of my must sees:

• Keith Peters – Casual Game Architecture – his Actionscript Animation books have helped me out on more than one occasion and I find his sessions provide clear and straightforward explanations of fairly complicated subjects. Also great to get some further insight into casual games development.
• Joa Ebert – Leaving the Sandbox – Joa is one of those rare flash developers who really understand compilers and bytecode and, from his recent blog entries, it seems he’s been working on some tools that could really help make actionscript code faster and better.
• Mario Klingemann – Connecting the Dots – Mario never fails to astound with his brilliance at solving problems and the artistic output that he turns his skills toward. I always come away from his sessions with new ideas and a renewed enthusiasm.
• Craig Swann – Choose Your Own Adventure – I saw Craig’s session at the first Flash on the Beach and it opened my mind to the possibilities of combining Flash, and the kind of visual output it is capable of, with physical world input via cameras, microphones and electronic detectors (a task made easier by the Arduino kit).
• Seb Lee Delisle – Work/Play – I’ve enjoyed several of his past sessions and really like some of the work coming out of Plugin Media (see, for example, Big and Small) so look forward to what should be an entertaining session about putting fun and play back into the digital.

I’ll post my impressions of how the event went sometime after September 23rd.

I’m still looking for a decent idea to run with but I’d just like to say I’m very impressed with the wonderfl site.

I’ve known about it for a while but this is the first time I’ve sat down and actually put some code into it. The compiler seems to respond well and the list of libraries available is good and expanding fast. It seems to be a really good tool for spiking new code ideas. The attributes that make it good for spiking are:

• online – so collaboration is easier
• auto compiles every time code is changed
• auto saves
• allows you to fork and browse a lot of public code (including some from bonafide flash experts like Quasimondo)

So, if you haven’t had a look already I recommend a visit to wonderfl.

Some of the most interesting flash work out there has to be by those developers who can push the platform that little bit further; thankfully a lot of them are willing to share their techniques.

In this article, I thought I’d pull together some of the advice on optimisation I’ve picked up from blogs, conferences, comments and books.

My earliest exposure to data structures and optimisation came from Polygonal labs. The blog is immensely informative. Here are some of the best articles for understanding optimisation:

• Explains the int, uint and number data types and which to use and when
• Bitwise gems for fast math
• Fast sin/cosine approximation
• Object pools

Data structures and motor2 are both really good libraries and worth checking out, too.

Joa Ebert has an actionscript wiki with lots of useful optimisation techniques.

Nicolas Cannasse is the brilliant mind behind MTASC and Haxe and his blog is great for tips on how the flash virtual machines work, what actionscript gets turned into and indeed on optimising that code. His blog is worth keeping an eye on for articles such as:

• Optimising perlin noise
• Virtual memory API for Haxe

Haxe is also worth a look, as it gives you access to the op codes that Alchemy uses for performance gains and has some really nice features.

Mario Klingemann is a great source of ideas when it comes to boosting performance and applies them to some really interesting flash output. I found this article on optimising a seam carving technique very useful as are his conference presentations – slides from past events.

Ryan Christensen at Drawlogic does a good job of explaining number weirdness in flash and also has a recent list of optimisation links.

Optimisation will make your code close to unreadable. Do not optimise prematurely. The job is probably best done by a compiler or a pre/post-processor. Unfortunately, the flash compiler does very little to optimise code. Haxe can do this job and as3c is another option.

Thanks go to those developers who came up with these techniques. I’m sure there’s more to add and I will probably come back to this subject in a later post but, in the meantime, are there any tips that you’d like to share? If the answer is yes then please leave a comment.

I’m sure it is a good physics engine for C++ developers but it has been ported very directly and I think could do with rewriting to fit with the actionscript language. It doesn’t feel like you’re learning an API, more like you’re venturing into another language, a task not helped by some lacklustre documentation and a need for good examples. No doubt, some optimisation could also improve matters.

Having said that, the engine is obviously powerful and provided many more features and physical simulations than I could possibly need in a simple game.

It was only after I’d got a fair way into the project that I had a look at another port of Box2D, this time from polygonal labs. I’ve been impressed with a lot of their articles, demos and the data structures class library, so I thought I’d investigate the motor2 physics engine a little further.

I have to admit I’ve only just scratched the surface of what both of these physics engines are capable of and my intentions are to carry on playing around and experimenting with both to see what I can get out of them.

In the meantime, here are some quick examples (with source code) to compare the two libraries:

Motor2 version (requires flash player 10):

Download the source code.

Box2D version:

Download the source code.

In both cases, I have made a very simple weighing scale from a revolute joint and a rigid box. I’ve then added another 98 boxes (49 either side) that fall on to the scale.

The motor2 engine has been updated to use the Vector functionality included with flash player 10 and makes use of the neat optimisation techniques you’d expect from Polygonal labs. Comparing the two on my PC (Dual Core 2.13GHz, 1Gb RAM with Windows XP and running the Firefox 3 browser), the motor 2 demo seems to run at a steady 45 to 50 fps where as the box2d version runs between 25 and 40 fps.

The motor 2 API has also had some minor tweaks to fit a little snugger with actionscript 3. I feel this makes it a little easier for a developer who works predominantly in Flash to start using the library.

While these are some initial impressions I have from looking at the two class libraries, I hope to put motor 2 to further use and will post some more demos here, in the near future.