Put the things that need to be static on a separate layer and set that layers parallax to 0,0 - then that layer won't react to any scrolling.

Nearly always behaviors can do all the basics.

Just add a sine behavior to it, set it to value only and use that value for the brightness amount, like this:

...and then...

If you want it to not stay dark for half the cycle just set the Magnitude to 50 instead of 100 and then in the event make it Sprite.Sine.Value+50

Edit: dop2000 beat me to it, fair and square

Magistross thanks, man, you are a lifesaver! I had put a wait in there for a test, but guess it wasn't in the correct position. All works as it should now.

This was one of the biggest things for my project and it's great to see it working just as expected (actually better).

Neat little thing. Would be nice to have simple drum parts, like bass, snare and hihat.

Some really great art direction going on here with the limited colors and resolution. Looking forward to seeing the finished product.

Thanks again! Everything is fixed now! Since I'm planning to use NW.js I just added the nice NWjs.ReadFile function and it all works out!

First page was not working correctly, because I had added array saving to page changes, but not before saving the file - so wherever you would save just being on an untouched page the array would be empty (or represent an older version of that "page").

One small "bug" still left is that if we have an empty page upon adding that page its array is already there, so we get one [0,0] junk entry. But in real life the array would have more data, some of which would never be 0 so I can just check for that. Or I could add a function that only adds an array on the first click, if needed.

Edit: Some strange behavior, but perhaps I just don't understand sequence of events or such. Anyhow, I have a " On open dialog OK" event that triggers the loading of a JSON file into a Dictionary and from there into Arrays. It seems to work, but for some reason when in the same set of actions I call the function to redraw the sprites based on array contents it just does not fire at all.

If you are willing to have a look it's right at the bottom here, under the last comment:

https ://files. fm/u/3zpa9n6k#_ (Still need to remove space, it seems).

Any clue why the function does not work there? It works for all other things we do, like flip pages, delete or add them etc.

You have placed the backup and not the capx on that link, so perhaps there are more effects in your "real" example file, but for me everything looks and behaves correctly here (latest nvidia drivers). Duplicated what you have there to have hundreds of objects on screen and still no problems. A screenshot might help.

Apart from that I have noticed that the Editor seems to only reliably display shaders with high precision. So you may try finding the line in the shader (.fx file in "Construct 2\effects" folder) that says "precision mediump float;" or "precision lowp float;" and change that to "precision highp float;". In theory it would affect mobile performance, but you can save the original and replace it back for final export if necessary.

Also, just a heads up, from what I'm seeing the Scirra HLS effect still bugs out at certain hue values with certain colors:

Hi again!

Perhaps you can help me out a little more - I used Magistross suggestion to have dynamic arrays and made a little "flipbook" test - basically you have "pages" (arrays of coordinates) and you can add them and flip between them.

Here's a demo capx - http s : //files. fm/u/prtvuhm8#_ (Please remove spaces as I'm new here - it's just a file hosting service, one of the cleanest I have seen).

The problem part is on the bottom, where we try to save and load to files - saving seems to work ok - I'm using a Dictionary to sequentially add all the arrays to it using their index as key.

But loading seems to not load anything, although it sounds like it should. Probably I'm just using the data format wrong. Any suggestions?

Actually it helps to sleep on a problem...

Yesterday I was digging into the code and hitting walls everywhere, when today I realized the solution was right there the whole time!

Well, it's a stupid, lazy solution, but still: Just use a Temp Paster object that's bigger than your sprite, For each sprite paste the sprite into the Temp Paster, apply the shader effects to that and paste that into your main Paster object. Voila!

Reduced the resolution on purpose, so we can see it working. Works well enough for my needs. Thanks again for making such a great plugin.

R0J0hound

Thanks for the reply! I see. It's more than likely that in that case I'll just add more transparent padding to the sprites and try to use math to resize them properly to keep exact sizes needed...

But also does this mean I have your blessing to try and modify Paster should I decide to go down that path?

I can see that R0J0hound is probably not really interested in these older plugins, but... perhaps there is a small chance of having a look at this:

It seems like if you have Shaders that rely on extend-box-horizontal and extend-box-vertical to work they get messed up when pasting (things that happen on the very edge do not happen), because, I'm assuming, Paster doesn't add anything to the sprite size. An option to overcome this is not having your sprite fill the entirety of the image, even a 1px transparent "outline" helps, BUT... in my case that's not an option, need to have the correct sprite size.

Tried looking into the source code, it seems like there may have been something that deals with this, but it's commented out. My skills are insufficient to figure it out... if there's any chance of R0J0hound or someone else who understands plugins to look at this I would be eternally grateful.

To make it easier I have prepared a little demo capx (with the shader included): h t t p s ://files . fm/u/7epu5z3v#_

And example images here: h t t p : //imgur . com/a/1rlYp (Will have to remove spaces, sorry).

Hmm, hadn't thought of an array object the same way as the "usual" objects... That would require some small utility function to change their "index" variables when deleting or adding one in the middle, but that doesn't sound too hard.

Thanks for the idea, I'll look into it!