newt's Forum Posts

  • The runtime stretches the texture like the layout editor shows. Canvases have always used as much VRAM as their overall size is (I wouldn't recommend canvases bigger than the size of the window), so this doesn't affect that.

    Well the reason I ask is because that even if you do resize at runtime it process's from the original size. That actually works well if your using an array, or distortmap. The strange thing is the amount of pixels will remain the same, but scaled. But updating the collision mask takes much longer once its scaled. Dunno it seems like its using the new size for that.

    Any way, some awesome work guys.

  • > Please - report bugs to the tracker, and include a clear explanation of the problem, what you expect to happen, what you see happening, and steps to reproduce. I just ran that .cap and it looks like everything is working fine. What's not working?

    >

    The canvas is set at a specific size and looks fine in the cap, but when run, it appears as a small square less than a third of the size it should be, like in the pic. It's already been mentioned in the tracker too as far as I'm aware, a few weeks back when I first mentioned it, someone else said it was a bug and then reported it.

    The example in the cap is an over the top one with it made much larger, to show the difference.

    Resize the canvas in the pic editor to your desired size, then check how much vram you use.

    Resizing needs to happen at runtime otherwise you'll eat up ram like milk duds, and god help you when you update the collision mask.

    OK, that one's fixed in next build.

    Is it just fixed for the editor display, or is it going to actually change the canvas dimensions for runtime?

    Edit:

    [quote:1gzdivmn]- [FIX] Canvas: cropping or invisible if not scrolled to 0,0 on start of layout

    Can I has no scroll always?

  • Hmmm, sooo..., how much would you be willing to pay for said objects?

  • Did you have multiple instances of the same object selected?

  • I'm not sure if it's a bug, or more like incomplete, any way yeah you can only use one named timer on the timer behavior. It is doable if you add an new behavior for each named timer you need.

    .... oh Rich... where are youuuu?

  • You can add components to your array, z for example, but that has a limitation at some point.

    You can also add multiple arrays, but again that may be limiting if it takes up a lot of memory.

    Probably your best bet is a private variable for the stat.

  • I tried lerp a bit, and it seems to be working much better, Im not sure If I have it set up exactly right but It does seem to be getting there.

    for x, and y loops

    set u and v to lerp(0, maxtilesize, loopindex(x, and y)*0.1

    That gives me a decent showing, but Im unsure of using loopindex, as the percentage.

  • You cant get per pixel on anything less than 2 pixels, so yeah point would be better.

    If your ship isn't terribly oblong you could try distance(x1,y1,x2,y2). Something like:

    system compare distance(bullet.x, bullet.y, ship.x, ship.y) is less than ship.width

    bullet destroy

    or what ever

    Distance works off of the image point, so as long as it comes within that circumference it should work.

    Other wise the only other thing you could try is instant hit with step, but that would be best with a percentage type of system... wouldn't really work with a bullet hell shump.

  • [quote:3r1yrsfh]Player attacks (with corresponding movement)

    Delay until it finishes and slight afterdelay

    Damage floater

    Delay until it finishes and slight afterdelay

    Prompt for next turn

    I think I see what you mean. The function behavior probably is your best bet, but I would use it in combination with groups. You could then enable group player a with a function, at the same time disabling player b group.

    Timescale would be better for messages, pop up windows, imputing stats, etc.

    BTW the timer behavior does remember picked objects, but you cant make new timers. Its a bug I think, but you can add additional timer behaviors.

  • Actually..., wait whut?

    What about 0,0?

    The thing is I'm trying to set z heights over many rows/columns.

    So will it automatically set the first vertex then?

    mmmkay off to experiment, thanks.

  • Try Construct 3

    Develop games in your browser. Powerful, performant & highly capable.

    Try Now Construct 3 users don't see these ads

  • Very nice indeed. Basic things refined, the way it should be.

  • That's pretty good. It solves the issue with destroying particles in a container, plus your not using a ton of instances.

  • Here is a cap of what I have been able to figure out. It tiles, but not perfectly.

    http://dl.dropbox.com/u/666516/dtm.cap

    I think I've figured out about as much as I can with what little experience I have.

    I still would like to be able to get this to tile like the bg object.

    Dunno perhaps there's a bug with distort maps?

    Any way, if any one wants to give a go at figuring it out it would be much appreciated.

  • If your using behaviors that use timedelta, you could use timescale to stop those actions.

    There's a nice article on the wiki about it.