So… what have I been up to?

I’m leaving Imba Games after a year and a half. I decided it’s time to move on from developing purely in AS3. I feel like my career has come to an halt. I also fear labeling myself as an “AS3/Flash developer” because of the unfortunate fact that a “Flash developer” is an underrated job, especially in Israel. It’s time to move on and find a different job, even if it’ll not be related to game development.

Nevertheless, it doesn’t mean I abandon game development and AS3 entirely! I still develop under Indigon! Remember Physics Defense? I killed it. I had an idea in mind, which didn’t work as a game. The important thing is that I have recently started working on a new small game called Frogget-it! It’s a game developed for Android, with possible ports to iOS and web Flash portals. I can’t say anything about it now besides “IT’S FUN!”.

 

 

I decided to start posting some technical blog posts from time to time, this being the first.

While I do have respect for Adobe (Flash is where my money comes from after all) , and I am also in gratitude for a 2600$ worth prize (won a CS5 Master Collection in a lottery at the latests Global Game Jam event), they do know how to annoy me. I will begin with something I encountered yesterday:  AS3′s Math.atan2 function. A very common function, especially in game development. I had to use it in order to point graphical arrows to various locations. When I tested the result, all the angels were wrong. Obviously the first thing I did was checking if I inserted the right parameters. And I did.

The documentation in Flash Builder 4 shows the ‘x’ parameter before the ‘y’ one. Seeing the above tooltip, I automatically wrote the parameters as I was told to. I had used atan2 in the past, and in the cold depths of my head I knew that someone was wrong (used it countless of times in the past, before I started using FB4), but who would argue with formal documentation? After 30 minutes of trying to figure what the hell went wrong, I checked the AS3 API on the web, and saw the following:

Adobe = FAIL

The function’s parameters are actually in the opposite order (!). Needless to say that I reversed their order in my code and got the desired result.

—–

Unfortunately, the resolution for the second issue is not so simple as the above: It seems the classical MovieClip’s gotoAndStop() method isn’t 100% consistent when using frame labels as parameters (which I do quite often when working on Gladelings). gotoAndStop() receives 2 parameters:

1. An Object object which can be either a number (for a frame number) or a string (for a frame label name).
2. A scene name, in case your MovieClip has more than one scene and you want to switch between them. This parameter has a default value of null.

I know that sometimes using gotoAndStop() with frame labels names doesn’t change the currentFrame property of the MovieClip immediately, which causes all sort of problems (Adobe claims this is a feature, not a bug [...]). In order to fix this problem when it happens, I search for the desired frame label name in the MovieClip’s currentLabels, and then use gotoAndStop() with the found frame label number instead. Today, I noticed a new, very odd, behavior:

After returning from a gotoAndStop() call with a frame label name as parameter, the currentFrame doesn’t change. The thing that does change against all common logic is the currentScene property! Instead of jumping to the desired frame, Flash overrides the existing Scene object, and replaces it with what seems to be a cloned copy of the original scene (it has the same length and same labels, but you can see it’s an entirely new object in debug mode). As usual, I tried overriding the problem by calling gotoAndStop() with a frame number, but the same thing happened again – the scene was modified and the MovieClip stayed at frame 1.

I have absolutely no idea why this happens. I thought it’s a good idea to share this since I couldn’t find any reference to this behavior in the web. Maybe others encountered it and can elaborate.

—–

Next week I’ll return to update about my undergoing projects and spew some information about a future Star Relic 2.

2011 is going to be great for me. Many endings bring many more new beginnings.

Two weeks the game company I work at, released its first game, Gladelings. Gladelings is Facebook social game we have been working on for the last 9 months. I was mainly responsible for the AS3 programming of the various mini-games that the game includes, the sound work (originally I even did the dubbing!) and other programming tasks. We continue to work on the game for a big upcoming worldwide release in the upcoming weeks. Hopefully the game will succeed and I will be more relieved. Success means I get to work on my own projects with less pressure :P.

What'ca looking at?

Besides Gladelings, I have other big news. I finish my Computer Science degree at the end of January. This means two things:

1. SCREW YOU TEL-AVIV UNIVERSITY!
2. More time to develop my own games.

Since Star Relic, I have really never sat down to work seriously on a new project. There have been a few ideas and few start-ups here and then, but due to work, studies and personal issues, nothing materialized into a future game. Now this changes. There is Physics Defense. I started working on the game a few weeks ago (and even bought a domain – no idea how nobody bought it before), and if things go well, I hope to release it by the end of the year (just in time for the next Israeli Game Awards).

To end this post, I will mention that my university AI project, Fivel, now appears in the blog. The game was made by myself and with two friends and we received very good feedback for it. Originally, we made a different game called Petri but be cancelled it due to its simplicity.

It has been a while.

I’m covered by work on my company’s game, university projects and my own new game. The game I’m working on at Imba Games will be released next month, I’ll post about it for sure. It’s my last semester in university (!) and one of the courses requires to build a game and implement some AI into it. As the only game developer in the course, I started working on a game even before the project was announced. I experimented with an small engine for board games which allows to change rules easily, and after a few hours of drawing I received the early version of…

PETRI

>squeak<

The game is made along with two fellow students. Now that the game is ready our main goal is to create impressive AI and develop the game furthermore. We wonder whether we should write the game’s logic and AI in Java in order to use multi-threading, the obstacle of Flash getting dumb under deep recursions, and the fact that Java is faster than Flash by many ways. On the other hand, there’s the option to use chunking (a method to split recursion into smaller ‘chunks’) which will make this project more challenging and fun. Eventually this will be handed as a university project so you will only be able to find it here and not commercialized in the web.

As for my own project, I started really working on one only last week. I had many ideas since of what I want to develop since Star Relic, most of them are big and demanding project I have always wanted to produce. Last month I decided that I should focus on smaller scale projects due to the current lack of time and resources (three concurrent projects are enough, no?). And thus, I started working on the third game under the Indigon label:

More information in the future to come.

And for the end – good news: Star Relic is nominated this year for the first Israeli Game Awards under the independent games category. The winners will be announced on November 9th.