— yeah you're right, but at that time, in 2012, C2 didn't have bidirectional for loops, so the inner for loop was automagically skipped in the last iteration of the outer one, like in the wikipedia algorithm.

But because for loops are now bidirectional, since I don't know when, it does a for "i" from 10 to 9 instead of just skipping.

Good catch, I'll try to reupload a correction soon.

—

I edited the link, but don't expect something that helpful, it's just a basic example of text loading and displaying

99Instances2Go

hmm works well for me o.o

https://app.box.com/s/hqn9vrttsuoesx2oolgk

I think you are probably confusing a few things.

What I'm more or less sure about (although I haven't used C2 for a long while) is that the start/end of layout triggers in an expected way.

// start game

on start of layout

every tick

....

on end of layout

// change to other layout

on start of layout

every tick

....

on end of layout

etc

now, for your problem with countdown, it doesn't look like a bug really. When you use a digital stopwatch and start it, the first thing that will happen is then millisecond going down, so instantly 90seconds becom 89seconds and some miliseconds

if your algorithm is something like:

countDown = 90

every tick

-> countDown -= dt

-> set Text to floor(countDown)

you'll probably not see 90, and that's ok

now for the hp problem, I can be that you lose the value of the hp after changing layout, so it's logical that the value saved would be 0

I don't have enough details to be really conclusive, but if it's something like

if hp <= 0:

hp += 50

go to layout GameOver

// in layout GameOver

on start of layout

save hp to local storage

and your hp saved is 0, it might be that either hp isn't a global variable (local variable of the same name?), or, that something else is touching it before you're saving it

also, it's weird that you save the hp in the other layout, it looks more consistent to save the variable at the same place you update. But missing too much context to be really sure.

Fervir and great newt

dunno, I don't think I'll be using construct3 that much, not really using construct2 that much either, these days my tool of choice for game is unity.

so I'm not overly motivated in doing a port. Maybe if one day I'm bored and curious about the changes they made in the plugin API.

In any case, as long as the API isn't stable (like, you know, they planned changes for the runtime, so things like trigger might not work exactly the same... I dunno... ), I probably won't do anything.

Also, well, I won't critic how scirra manage their products, but since they still lock some important feature in the free version (families, creating JSON files, and, it seems, using their awesome brand new stuff like array and dictionary editor,...) I'm not sure I'll be able to properly test the plugin, like if you have a family of JSON plugin object, will it still work? I know that for this specific plugin, using families could be a bit weird, but that doesn't mean I shouldn't test this case.

(it's really weird that even in beta you can't access those, but meh... I'm busy not making games anyway \o\)

Well anyway, the short answer is, don't count on it... but maybe.

grigrizljac

yup, it's because I'm using CircularJSON so you can import/export complex datastructure.

There's still no way to change the special reference character (~) for another one or escape it, sorry.

A circle isn't a polygon. You can approximate a circle using a regular polygon with a lot of sides

I won't provide a capx, 'cause I'm lazy, but the pseudocode would look like:

sides  = 32
radius = 100
pos_x  = 200
pos_y  = 300

angle_step = 360 / sides
for i from 0 to sides:
     angle = angle_step * i
     x = pos_x + cos(angle) * radius
     y = pos_y + sin(angle) * radius
     addVertex(x,y)
[/code:3dq0ztn5]

something like that.

gumshoe2029 I'm supporting it for the time being as far as bug fixes go, but as far as adding new feature I'm less inclined to do it to avoid increasing the bug fix support load (don't have that much time) and also, since it's already not that user friendly, I prefer keeping it simple.

if you have any issue, report it on github, it's easier for me to keep track of it over there.

totoyan

share a capx with a case that reproduce the bug, so I know if it comes form the plugin, or a misuse of it

if it's a bug from the plugin, I should be able to quickly patch it =)

After more than 2 years I finally released a new version :D....

Now everything is hosted on github including the documentation

I'll edit my first post to reflect that.

tmntppn

yeah, the problem is that if I start adding indexOf, I'll have to add push, pop and shift, unshift etc... the plugin will quickly turn into an array plugin.

And also people might want something like that for object (keyOf, etc)...

Well... I'll think about it but since I'm not as motivated as when I started it, because I'm not using c2 anymore, the more feature I'll implement, the higher the risk of bug there will be, and I'd like not to have to spend too much time in maintaining it (:

hlebegue tmntppn

yeah I knew about this bug, but years ago, I started a new interesting feature and I lost interest in C2.... :D....

so I never released a new version (at that time I was so n00b I wasn't using version control so my fix was mixed with my new feature in progress...)

Anyway, yeah that's basically the fix. I think I'll try to release the fix with that new feature some day.... Work is eating up any will to do anything personal so well... (: