next »

[rogerlevy]

CockScript$+/-

[Moczan]

It really depends on the kind of game you're trying to make. NME wasn't really designed with GPU-powered rendering in mind, it's most useful as an abstraction layer for porting Flash-programming based games to multiple platforms. In that capacity it is excellent.

I do agree that NME isn't ready for large-scale development efforts. But for individuals who have experience developing for Flash, and a desire to see their work on other systems, it's ideal.

I can't agree that it wasn't designed with GPU-powered rendering. It didn't even support vector assets in SWF until one of the recent updates (and through external library). It was designed as a 'flash-like' abstraction over SDL, but seeing as Flash and non-Flash rendering is completely different in it, you can't even use same code base to deliver to all target. Comparing to AIR, this project has a very long way to go.

Just like I said before, the idea is great, but currently it's an undocumented, buggy mess. The dev's attitude is screwed (why fix the core functionality when I can add a fancy class and make my update blog post look pretty. They also insist on using same class/function/variable names as Flash's library, but they change the behaviour of some of them to something completely opposite which is confusing.

[Richard Kain]

It didn't even support vector assets in SWF until one of the recent updates (and through external library). It was designed as a 'flash-like' abstraction over SDL, but seeing as Flash and non-Flash rendering is completely different in it, you can't even use same code base to deliver to all target.

I've been taking more of a Flixel approach to Flash game development. As such, I've been using pixel assets instead of vector importing, and blitting instead of vector rendering. That's probably why I haven't run into nearly as many conflicts as you seem to.

I can certainly appreciate your concerns. But keep in mind that I already qualified NME as being appropriate for small projects by individuals. It's not a major development environment being produced by a corporation with funding. It's an open-source project being developed by one guy. Of course it's going to have bugs and be feature incomplete. Part of the reason it's been working for me is because I've dialed my expectations back to a reasonable level, and worked on projects that play to it's strengths, while minimizing it's weaknesses.

[Moczan]

Read my post again. I can't believe you even downloaded the NME, because actually the bitmap rendering is broken, I haven't touched the SWF support yet. Seriously, just use AIR, you will save yourself a lot of time and stress, NME doesn't even have any killer-features and they are both completely free (with better IDE support for AS3 than haXe).

It's just stupid to dial your expectation back to the level where you don't expect the core functionality to work. It's like saying "I'm using OpenGL, but I don't expect it to render anything other than pink circles".

[JMickle]

Trouble with air is the end-user having to install the runtime (like .net), unless they have fixed this?

[Fallsburg]

Trouble with air is the end-user having to install the runtime (like .net), unless they have fixed this?

Yeah, because who installs games. 

[Hima]

Trouble with air is the end-user having to install the runtime (like .net), unless they have fixed this?

I thought he meant using AIR to compile to native mobile target.

[JMickle]

well its more like, I tried to give my little android AIR game to a few friends, and they had to download the AIR runtime as well, why can't that be packaged with the game? or can it?

And also it's pretty crap to have to install AIR for a shortform downloadable thing. Of course, you could just make it online, but maybe there is a feature of AIR that you need?

I dunno, I just think its an annoying obstacle.

[st33d]

AIR has very few features that the Flash Player doesn't already support and no one has come out with any benchmarks showing that AIR is faster than Flash.

Many games don't require installing. AIR requires two installs, one for AIR and then one for the game.

AIR does allow packaging the installer with the result as of recently, but you have to sign up to Adobe's distribution program:

http://stackoverflow.com/questions/7831142/how-i-can-download-adobe-air-bundled-runtime-installer

Basically, there's not much point in using AIR unless you really have to (mobile, saving files without a file-manager). You're gaining nothing.

[Fallsburg]

One of the major advantages is the seamless file saving.  But that can be a pretty nice thing.
But the allowance of native extensions is probably the biggest thing that AIR allows (it's a new feature).  It allows for things like the integration of game pads without requiring something like joy2key.  Essentially, anything that you could care to do in a C++ extension you can do.


And you don't have to sign up for Adobe's distribution program.  I mean, the #1 answer in your link says as much.  AIR 3 (and above) allows you to have a captive version of AIR packaged with the app.  This is essentially the same thing as making a projector of a .swf since it means you are guaranteed to know which version of AIR (Flash) you will be running.

[jotapeh]

Yeah, look, Adobe's PR is bad enough without us spreading misinformation.

You CAN distribute a standalone (captive AIR) runtime without requiring AIR on the end system. No catches.

[st33d]

My bad, I was working under the assumption that AIR was a largely pointless and cumbersome exercise.

Turns out it's just largely pointless now.

[Fallsburg]

Again, I think that's a bit harsh.  But whatever, we obviously disagree a fair amount about Flash development.

I would say that if you are looking to create a downloadable Flash game, there is no reason not to use AIR.

[jotapeh]

If you want to qualify for distribution on, eg., the Mac App Store, captive AIR runtimes now give you that possibility. (and yes, the latest AIR updates guaranteed this.)

[JMickle]

You CAN distribute a standalone (captive AIR) runtime without requiring AIR on the end system. No catches.

Ah! This is good.

I would like to see some benchmarks regarding the speed of AIR, though.