Roccinio

instead of doing a for_each then triggering once via condition, you do for_each and have an instance variable you check

If the variable is true to do what you want it to do and then at the end of that you set the variable to another value

Most likely you'll just use 1 and 0 for TRUE and FALSE.

This way it goes through, and triggers only if it's TRUE, then immediately set to FALSE after triggering. You set the variable to TRUE next time you want it to trigger for that instance.

Basically it's the manual way of "Trigger Once", since when it's true, you set it to false the first time it runs so it doesn't run again next tick.

The closed bug forum isn't really a good place to get support, but at least you searched the forums :)

What exactly would you use the triggers for? Most of the time triggers on the platformer behavior get used for changing the animation and other events that need to trigger when the player lands or jumps. Since you don't move around and get pushed defaultly in 8dir, other than that I am unsure what you need

---

For 8dir depending on what you want to do, you can do it based on it's VectorX and VectorY values if it's something triggered on movement directions

This way you can change the animation when the values meet a certain requirements on the vector itself. This way you can do diagonal movement just by comparing the difference in vector totals.

If the 8Dir rotates, you wouldn't really need to change the animation based on what direction it's going so you would have to change the animation based on whatever angle the playerbox is at and then change the animation based on that.

I was having some issues with modifying a huge project of mine and noticed some key things that would improve troubleshooting:

1) Timestamps on Undo/Redos

This would be useful if you knew the time it last worked.

2) Differences

Highlighting the lines that have been changed, and the ability for the software to focus on that specific change would also be a great boon as you can tell what has been changed.

Currently without that, it's a crapshoot because if you need to undo a lot, you don't know what you're undoing.

There are also more things they could use as well, but those two are the only ones I can remember atm.

I've only done it once when my computer was lagging. Other than that Steam refuses to let me open multiple instances.

Got a trick for that?

1 - If this was done via the animation system, you would have to create additional frames for each possible state. Say that you had a platformer and you wanted to have this sort of system where you flash white after being hit (similar systems occur in lots of games). You would need to make white jumping frames, white running frames, white attacking frames, etc etc. If it's possible to do something programatically and not put additional strain on your art team, it should be done. You would just have to make the hurt animation like that, or at least part of it. It's not that many assets and if it is then you're trying to do too many.

Most of the time you're only get hurt in like 2-3 positions at max.

Although if there was a method to apply a mask on 1 layer only that would be pretty cool.

Deko

Glad to see someone is working towards this again.

I know that Velojet has stopped working on the project, but that doesn't mean you should hesitate trying to make a plugin that uses PhotonCloud. I think that someone taking up the mantle would be good in order to get the people who make PhotonCloud to want to support Construct 2 development.

There should be multiple ways of doing something, and currently Simple Online Multiplayer is a bit lacking for C2 at the moment.

I do suggest that once you have it more complete to make your own thread about it rather than use this old one. This thread has a lot of information that could confuse newcomers that would use your plugin instead.

Another possibility is to use google's support for expansion files. Ashley's suggestion of the app downloading the assets and relying on caching would be the simplest solution.With this wouldn't you need to host your own assets, or would Gplay host them?

Well what exactly are you trying to count? How many tiles in the tilemap? How many non-blank tiles? How many tiles use the tile type?

Clarification on what you want and how you would use it would help as they can't implement a feature without knowing what it needs to do.

So if a switch from Canvas2d to WebGL would be made, this fellow user would be "argh everythings fine".

As i said, just let it optional. So the users can switch betwen those two.

Well while I do like the option to allow people to do more if they want to explore developing with WebGL, Scirra's mission statement is to have C2 to be supported on older computers.

On these computers the end user doesn't even know what WebGL is, and would not know how to enable it IF they can. While most computers nowadays support it out of the box and the older ones drivers can be updated to support it, it's only really transparent on Node-Webkit. Browser based games they would have to enable flags, or update their browser version.

Lots of those requirements are very easy to do on PC, but mobile devices with HTML5 support do not have it that easy. Many of them run worse WITH WebGL enabled. Scirra does not like introducing a feature that requires too much WebGL without really having a go at Canvas2D implimentation first.

They could've made the tilemap plugin webgl only, but they worked at it and their implementation of it is wonderful. Perhaps in the coming years they could not put as much importance on canvas2D, but legacy support is very important at this moment in time. Plus if they can make it work for Canvas2D, it will work smoothly for WebGL and that's important too.

Out of curiosity: how it was made in CC? It worked perfectly there.

.CC used DX9 as it's foundation so it had more access to the video card. WebGL is essentially that, however Canvas2D is a bit weaker as it was designed to be simpler.

For animation I heard you can simulate that via "load from URL" in your events, although since it's part of the existing TiledBG I don't see why they could introduce filling from part of an image based on how the Sprite object splits and loads different sections of an image. At the very most it could create multiple images instead of one stitched image, if the area-fill command messes up with a image section.

Offsets and rotation, other than moving the object itself, would probably either be very tricky for Canvas2D, or WebGL only as it's more intensive on graphical manipulation. If they were WebGL only though, they would still have to not crash/make Canvas2D run any slower as well.

Instance variables should let you do that so when you check them you know they should be allowed to interact with each other.

Another thing is using families, but that also would require multiple objects. However you can have all the properties those objects need applies to the families and thus are able to add new objects without needing to re-do all the variables/behaviors.

Most of the time when collisions are being triggered globally, you're checking for a very generic collision event that includes the objects colliding.

You can change the preview browser/mode if you click on the project title and then select "Preview Browser" > Node WebkitYeah, that feature was added after this thread was created.

No - 'trigger once' ignores instances completely, because it's a system condition.Ah, okay then.

Although wouldn't it still track the instances of the previous picked conditions?