2. How do you mean?

http://dl.dropbox.com/u/10316081/example.png

It's somewhat in line with what I said with my first post. This picture consists of three objects, the circle has the XOR effect and the 2 squares are over lapping. From what I've read, most of the canvas effects require at least two objects. What I'm asking is, would it be possible to make the XOR effect (For example) put a hole in the red block, but ignore the green one and allow you to see it. Or am I completely screwing up what these particular effects are for?

I guess that's the only option for the moment. Mind if I ask a few questions though?

1. Would layer effects be possible in HTML5?

2. Would it be possible to restrict what object effects react with?

3. Is the current ensemble of effects all that could ever be made available without the HTML5 standards changing?

Right now I'm trying to see if I can completely redo the original construct 0.x ghost shooter demo. I've run into a problem though with lights. My plan was to use a semi-transparent black sheet to simulate the darkness, then use one of the effects to eat out part of it when an object gives off light. So far however any time I've tried to use an effect if would just react with everything meaning the light would eat through the darkness, the ghosts, the player, and the floor. Any advice?

I can't figure out how to stop my platforming character from hopping up a steep slope. I've tried bounding box and per pixel and neither helps. I want him to jump on the slope then slide down.

My suggestions for jumping

Disable jumping when on the slope

Or

Give character horizontal momentum when jumping off a slope; enough to subvert their ability to jump up hill.

Haven't experimented with it much

But I think the new "additional velocity" command for platform behavior might be able to help you out on the sliding.

I've always thought if people regularly code their own movements instead of using behaviours (e.g. a custom platform movement instead of platform behaviour) then the behaviours should be improved so that everyone uses them. They're not meant to be a toy for newbies, they're actually meant to provide functionality in a useful and accessible way. They should also be customisable enough that you can implement custom functionality to make your game unique and interesting, be that springs or reversing gravity or whatever.

If you have any ideas on how behaviours can be improved towards this - especially in C2 - do tell us your ideas! I was under the impression most people use the behaviours, even for full games - if that's not the case, I'd be interested to learn why.

Hey Ashley, would you say the idea behind a C2 Behavior is to make it easier to modify than a C1? The biggest problem I've always found with C1's behaviors is how excluding they were to other code. Leading up to a point where you like most of a behavior but its coding rendered an idea in operable*; requiring the designer to either try to replicate the behavior from scratch or sloppily code around it.

Also just my suggestion, but the ability to synchronize values from multiple behaviors might be an interesting suggestion toward making them "play nice" especially if you could modify said behavior so one wouldn't have control over an object. So if two values are synchronized then they both share one value making any change to one a change to the other.

For example. You could synchronize the X and Y velocities of a platform and physics behavior, and then modify the platform movement to not be able to directly affect player movement, but do it through the physics object. But that's just me talking.

*Example: someone wanted a game using a time-slowing mechanic. The player himself would not be affected by the slow. How do you code around the platform behavior to accommodate this? (Issue with behaviors and timescale)

*Example 2: someone wanted a platform game where some platform characters would ghost through certain platforms and others wouldn't. I actually did manage to code this, but I had to completely abuse how the behavior worked to accomplish this and it looked kinda sloppy. (Issues with behaviors and collision)

Hm, this feature that is critical for beginners is only available as a paid asset. I'm not too keen on spending money to try something out.

Apparently they're going to come out with a trial version and possibly a cheaper licence soon. On the flip side of that, the $100 price tag is apparently a "SPECIAL INTRODUCTORY PRICE for a limited time only"

It's a little bit more complicated than that, because:

User 1 buys normal license

User 2 buys normal license

1 month later user 1 upgrades for total cost of new license - amount paid

1 day BEFORE user 2's license expires, he upgrades.

Both paid the same amount, but license expiries would be different.

Although, as you point out it wouldn't be hard to work out, we're just considering options at the moment.

Would it make sense to do something like this?

Indie 2 Year to Commercial 2 Year

Original Purchase 1 year old

?149-?39*(1/2)=?129.50

License time resets.

Also, will other currencies will just be direct translations (What ever the exchange rate is)?

I've already cast my vote for animations and have no intention of taking it back. But a method of online and offline saving for games would eventually be nice.

With online, I'd imagine you could even make a basic CMS.

We need some way of differentiating from Construct 0.x and Construct 2 - I think just "Construct" would get confusing as hell at this point

Construct 0.x = Construct Zero?

Construct 2 = Construct Duo?

The first bit of "3D" in Construct 2.

Live

Cap

A what

I believe it's a Chrome OS laptop.

Do we have alpha support yet?

Um yeah image editor is a biggie.

Some kind of data storage plug, I'm guessing array is the most versatile.

But that brings up an issue. How do you incorporate external storage types into a web app?

Also the old system expressions, distance(), lerp(), etc, but the UI for those should probably come first.

Actually, we've had alpha support. Use a program other than MS Paint by setting the default tool for PNG's to something else (Like Paint.net)