Interesting approach! So then I would also place all the objects initial images and their extra animation images in the same animation set? I wonder if this would bring perplexity later on.

How did people do such things before families were introduced?

I was also shocked to find this Signal thing, it can decouple code and I wonder why people dont use it regularly.

What is the difference between the *condition* "System: on signal X" and the *action* "System: wait for signal X"?

In other words, where should we use "wait for signal"? Can you give an example of "wait for signal" proper usage?

If i use it inside a set of actions and the signal never arrives, what happens with that block of code? Potential bugs magnet?

Many thanks, very informative post! :)

eski

Nice game and very playable, congrats! I see you have more than 500 downloads in two weeks, which means that users are spreading the word...

If you don't mind, what scaling options have you used for the objects to be spread across the device vertical axis? Scale outer + Anchors to layout?

Something other, it seems that the sounds are a bit delayed (about 300 ms after the touch). Is this by design or due to chromium audio delay? I thought there must be a solution for this issue but all the C3 games I've downloaded have the exact same audio lag.

So, what is the reason of changing engines? The switch export or the engine itself?

I think that c3 is customizable enough. The reason of its existance is that it handles the project architecture for the developer and leaves just the logic to be built.

What the developer has to excell in is creativity. One of the best 2019 games, Baba is You, was built with Fusion! It could be easily built with c3. The tools are there...

Now, would baba develop his next project with c3? I doubt it. C3 is marketed and developed as web-first (and that leaves the other exports with inefficiencies). He would want to choose a tool that offers convenience but secures the target platforms first. Broader audience is now the goal. So he will probably end up to unity+playmaker/bolt even for games that are better deliverable with c3...

Cocos family has a lot of projects. One of them is Cocos2d-js in which you write javascript cocos calls and it's translated (binded) to c++ cocos calls via Cocos2d-x. For C3 to use this route, I guess it should map its core actions to the 2d-js actions.

However, my point is not to promote cocos, but rather to encourage the team to consider whatever technologies it takes for the end result to scale according to the needs of the crowd. Html5 is just one thing. Flutter is probably the next/better Flash. UX tools with code-autogeneration is another big thing. Mono/.net is the new gamedev lingua franca. Things evolve. Construct is not the best html5 game engine (Defold is), but it is the best event-sheet software around. For me, Construct is a shiny event-sheet that happens to lack native exporters. Eitherwise I would abandon Unity in no time for Construct! Do you know how many "professional Unity developers" actually work with shit-tools like Playmaker? It's ridiculous. Construct is miles ahead. And it is consumed for making ...mini games for kongragate, while the industry lives its most wealthy times in history with mobiles to be 1/3rd of the revenue generator and a tablet console to sell better than ps4. The paradise of the indies is not in ...kongragate! I wonder why such a wonderfull product is spared on a web-first strategy. Web what? Desktop web is dead. Web is on mobile (90%). Web IS mobile (and mobile is changing... it's not Android neither iOS anymore... it's becoming "Flutter", and it happens to have a powerful webview!).

Just connecting dots...

Xbox One has "GamePass", which will be huge in the forthcoming future...

However, I also believe that Construct needs a second export option as an alternative to html5. For example, GDevelop is experimenting with Cocos2d-js for native builds. A smarter move for Construct would be to use cocos2d-x lite which is also what powers Cocos Creator (2d-js is getting EOL now). Cocos2d is the most mature engine in the mobile scene and exports to everything except consoles (even this may be doable with some additional tweaks).

I understand that the team has strong opinions on keeping the project on a clean bet (html5) in which it specializes. However, I (hope and) believe that Construct deserves to be bought for some millions by a big player and become popularized to the masses. The value of your product is not it's subscriptions or the user base - it's the technology concept that works wonderfully and has not yet been monetized by a big company.

It makes no sense for FSM-based visual scripting to be popular nowadays (Blueprint, Playmaker, Bolt etc), when eventsheets are so more readable and productive. But Fusion is almost abandoned and Construct/GDevelop are considered as toys due to the lack of a native exporter. I'm not saying that you should change your strategy. Just consider some auto-translation options that exist. Just my 2c...

I want to calculate and set some "optimal content width and height" global variables, lets say = x% of the window size, depending on the aspect ratio of the screen. I know how to do that, but I can't find a way to set these globals before my objects are created (since they use these vars with a "on created" event).

I can solve this with an intro layout where I set the vars and then "go to next layout". This way, however, the setup is broken everytime I press preview (as it runs only the current layout, by default).

The second option I thought of was to set them inside a js script in "runOnStartup" function. However I can't find how to get the screen width,height from there.

I guess there must be some event condition that becomes true before objects creation ("on layout start" happens after objects are created)...

Any ideas?