Found an interesting article that I'd like to hear opinions on this from TIGs folks!

Charmingly utopian though this all sounds, I have two major problems with it:

1) It says "artists are paid for what they create, not for what others create". However, without copyright that's not actually true. What would happen is that you make something and then a large corporation with much better output channels than you would make almost all the money from your work. This sometimes happens already, but at least right now the fact it's illegal stops most companies from doing it.

2) The idea that all you need to do is credit the original artist doesn't play nicely with allowing remixes. If I combine elements from ten videos each of which has ten people credited I would in theory have to credit a hundred people. So what will happen in practice is that people will credit selectively, which means that lesser-known artists will often find their work used uncredited on major projects.

More reuse of content may indeed be the future, but I don't expect to see that happening via copyright being abandoned as a concept. It seems to me that a more promising route is the kind of compulsory license currently used (for example) for cover versions in the music industry. Roughly speaking, the way it works is that a license is automatic, requiring no negotiation or even contact with the original author. That doesn't mean no license fees, though - a standard royalty is payable to the copyright holder.

Further to Nix's point - the player's character in a platformer is usually under very precise control. Even something as simple as momentum can make the character feel hard to control. Indeed, many of the most popular platformers have characters who can do absurd things like change direction in mid air!

Box2D can support these sorts of things, but there isn't a single, canonical "how to" since you'll be needing to adjust default behaviour a lot.

Asking how to implement platformer physics is like asking "How do I cook a meal?" - the question is simply too broad to answer clearly.

Note that you need to be careful about "free image" sites in general because they might be wrong about the images being free. The images on http://opengameart.org/ are checked by the site staff before approval, so are probably OK, but few sites are so careful.

Silly things like the main function being wrong are easily ruled out by compiling HelloWorld.cc (or similar).

Step two is to compile a stupidly simple toy SFML project to check that the basics work.

Debugging an issue like this by working directly with the big project you actually want is asking for hours of wasted time!

I wouldn't wait more than a few days before sending a (friendly) followup email requesting a prompt reply. That said, you're not necessarily doomed - your contractor might just be on holiday for a week and has forgotten to tell you.

Also I've occasionally worked with people who have terrible email habits. For example, one guy literally stopped checking his email at once point because he had "too much stuff" in his inbox and didn't have time to go through it all!

[true]

I find this style fairly unreadable, but as far as I can see this:

Code:

if((m_lListIt->get()->getLeft() <= bumper->get()->getRight() && m_lListIt->get()->getRight() >= bumper->get()->getLeft()) && (m_lListIt->get()->getDown() <= bumper->get()->getUp() || m_lListIt->get()->getUp() >= bumper->get()->getDown()))

...says this (in pseudocode)...

Code:

if (m.left() < b.right && m.right > b.left && (m.down() < b.up() || m.up() > b.down))

...which checks that the character overlaps the buffer horizontally and then checks that either it's bottom edge is below the top of the buffer or its top edge is above the bottom of the buffer. However, that might not be a collision at all.

Consider, for example, the case where the character is a long way directly above the buffer. His top edge is way above the bottom of the buffer, so the expression will evaluate to true. This doesn't seem to be what you want, because no collision has occurred.

This is really fun - good luck finding a sponsor!

A few minor grumbles:

* Given the nature of the gameplay, info dialogs which pop up should have a minimum time on screen (say 1/2 second) and ignore (not remember) any clicks during that time. The current implementation caused me to miss several dialogs because I happened to click something as they appeared.
* When you first encounter powerups, some explanation of how to collect them would be good. (Or did I click through that?).
* You might want to consider biassing bullet fire directions away from the edges of the screen. Because Flash can't track the mouse outside its boundaries it's awkward and annoying trying to pick off bullets aimed by low altitude aliens into the corners of the screen. I keep accidentally pausing the game, as well as missing the majority of them, neither of which is fun!

But these are details, the game's already good. Nice work! 

I think multiply is simply src * dest?

Correct. You have to be careful though - because it's performed channel-wise the results can vary if you switch colour space.

Right now this doesn't quite beat my browser homepage (which is a huge grid of sites I visit, but has the advantage that I can reach it from any machine whether or not I own it).

That said, it could be useful if it also included some way to associate Windows actions with buttons. That would allow me to clean all the links off my desktop and replace them with just this button panel. And I could launch lots of things from a single button, so when I click "Unity dev" it launches the Unity reference in Firefox, emacs, the Unity editor and Photoshop.

Note that whilst Moo do some very cool things, if you just want business cards then going to your local printer will likely work out significantly cheaper.

You play as the Chicken Bandit, an outlaw riding a big chicken, who robs trains and must fight Robotic Cowboys.

Looking good!

I like the lassoo concept - lots of scope for coolness there.

In such a situation I always prefix: get, set or create.

Perfectly reasonable, but I see no reason why a lack of prefix should canonically mean "do" for a noun. It seems to me that the least ambiguous is to always prefix with something, but if no prefix is going to be used, at least recognise that it doesn't necessarily mean "do".

Why do people put "do" at the beginning of function names?

It's a function! It's tautological that it's doing something. There's brackets after it!

I think this can be legimate sometimes. The "do" part applies to the other words in the function (presumably a noun), not necessarily the function itself. If it were removed, there might be a more obvious implicit verb.

For example "doNextAction()" seems fine to me, because "nextAction()" could be "getNextAction()" or "selectNextAction()" or any number of other things.

To see more straightforwardly why the answer given is correct, imagine a system consisting only of a single ball and a flat surface.

With no energy loss and perfect simulation, releasing the ball at some height above the circle will cause it to bounce and then return to exactly the same height. This bounce repeats indefinitely.

As with your example, this is not the behaviour one would expect from a real ball.

The question you have to ask yourself is who the audience is for your game. As the above two comments imply, many audiences are not unduly concerned with state-of-the-art graphics.

The last two games which I paid money for were Vertex Dispenser and Frozen Synapse, neither of which has anything advanced going on in graphical terms.

GameTunnel's coming back? Excellent!

I would strongly recommend using someone else's engine for your first few games. This will give you a sense of which features are essential to you and which optional features you feel are worth the effort to implement.

(Apologies for not recommending a specific engine to start with, but I haven't used C++ for a serious project in a long time. I hear SFML is some good...)

I type control-x, then I type h. It says: "Mark set"

That's correct. It's the command for "select whole buffer".

Then I press tab. Nothing happens.

That's the default behaviour IIRC. If Nix is an emacs user he may have rebound tab to something clever, but normally indent region is C-M-\ (read as "control-meta-backslash, which is Ctrl-Alt-"\" on Windows).

That said, I don't really recommend using emacs for indentation (even though I love it as an editor) because some of its default indentation rules for common languages are quite idiosyncratic. At least not unless you're also going to use it for editing, in which case you might want to set up a suitable command to do this efficiently. (This being the emacs way. What it does out of the box is barely relevant to most users.)

Guys, FaceBook may not be to your tastes, but it is widely used as a gaming platform. If you don't want to use it, that's no excuse to rant all over someone else's game announcement thread.

Yesterday I finished my raycaster demo, just now I added a simple day/night and improved the code a little: demo.

Ooh, smooth! I assume it's a 2D raycast then rendering a column of pixels?