pirx's Forum Posts

  • 1drv.ms/u/s!AkmrWgxeuxlKhIdtK-WtIlVlBvH2pQ

    Here you go. But yeah, could easily be an effect.

    Thanks - but then again, blur. Even a darkened sprite copy effect like this would be convenient to have, though. But still, what I meant was a soft shadow with blur, which I think is the most problematic part.

  • make a copy of your ui behind your UI, paint it black or grey, make it 50% opacity, put it a few pixels down and a few pixels right, badabing-badaboom

    Yes, the key word though is "blur". 😉 (and no, none of the currently available blur effects give a good gaussian blur, including combining horizontal and vertical)

  • Agreed, one of the features that make you feel 100 kg lighter. Try implementing a full-game save of all states in something like Core Data in Swift and we'll be talking days of development vs 30 seconds 😬

  • There appears to never have been a good solution for this. I can't imagine I'm the only one regularly looking to drop a simple shadow with offset/blur/intensity on a UI element or sprite in one game or another.

    Shadow Caster doesn't count 1. it's way overkill for simple shadows 2. no good blur effect available.

    Drawing sprites with shadows already on them also doesn't count because it's super inflexible and doesn't work with scaling, rotation etc.

    I remember Ashley (?) explaining a few years back that shadows, or maybe this was specifically blurs, are problematic but can't remember the reason. Heavy performance-wise?

    On the other hand it's such a common feature in all kinds of software; so is there really no good solution to have soft shadows or at least a good blur effect?

  • Try Construct 3

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

    Try Now Construct 3 users don't see these ads
  • Thanks! Well there's other reasons, the score was just an example.

    E.g. the game has an away timer that depends on it knowing the last saved time at all times, things like that. Yes this can also (should) be done with localStorage but it was more of a theoretical question.

    Thought maybe there was some rule of thumb like 'make sure not to save stuff too often because it kills performance/disk/battery'...

  • Hi, was wondering this today, working on an idle/clicker game.

    Ideally I'd save game every time the player increases their score, which could happen every couple of seconds. Is there a particular reason not to do that around 20 times a minute?

    Would that have an impact on performance? How about 20 times a second, for example?

    Not talking about extreme cases of saving thousands of things on screen or something, let's assume a pretty ordinary save with some optimization done (excluding static objects etc).

  • I wish you all the best and would be very interested to see the result, but not sure how much experience you have with developing software in general.

    Just a heads up, although it's technically possible, doing this with multiplayer and/or own backend using Firebase or something would be a lot of pain even for an experienced developer (Construct or not). Multiplayer is evil.

  • I would also very much like the ability to replace objects in events with a different object type.

    Yes, I understand that some events or actions may be incompatible, but many events exist in all objects. So the editor should allow me to replace "Sprite Set position" to "TiledBackground Set position" for example. Or "Text set text" to "SpriteFont set text".

    However this would be very welcome. Couldn't count all the situations when I wanted to replace something and got confused as to why it doesn't let me. Ah 'mkayyy.. this is a tiled background not a sprite/9-patch/whatever. .. and the only way out is redoing a bunch of things from scratch. Would be great if some basic types were compatible for replacement.

  • lionz

    Thanks, great suggestion. I actually also dislike the way with the family because it feels clunky, and would much rather work with the two objects directly. However, unless I'm doing something wrong it seems that the following logic just doesn't work:

    On Drag Ended -> If Sprite1 is overlapping Sprite1 -> do something.

    The "do something" action just doesn't trigger, whereas it does without any problems when it's Sprite1 overlapping a Family.

  • Hi.

    I'd like to detect a collision (actually an overlap) between an object and another instance of that object.

    Found on multiple threads in the forums that it's best to put the object into a family and then detect the overlap between object and the family.

    That works, but I then need to compare instance variable between these two participants of the event. How?

    E.g. The object has a Strength instance variable. How do I get this value from the other object (referred to as Family)?

    Thanks!

  • Thank you Ashley; so as I understand the way to go is to use

    - full screen

    - parallax 0

    - anchor behaviour

    to make sure my UI adjusts to the container. Is that more or less right?

  • Hi,

    Using C2/3 for years on and off now; never had a real need for this. Now I need to develop a couple of web based small tools for my website, hence the question:

    Is there currently a way to make game/app frame *actually* responsive? Not talking letterbox scale etc. where fixed aspect screen is zoomed/cropped but rather positions of objects actually reacting to screen size; along with adapting to right positions when the window is being scaled and all that.

    Can it be done?

  • Thanks! These plugins look interesting -- sounds like it's possible to manage Playfab with a humanly interface :)

    Also for anyone interested, I found a service called Sashido.io that enables a super easy Parse server deployment (no advanced conf. needed). I guess you need to use JS code in C3 to interface with the API but other than that it's pretty neat.

  • You do not have permission to view this post

  • For those of us who develop apps or app-sy games this is the main limiation of Construct:

    it is not well suited for developing user interfaces.

    Whereas Construct excels at game creation, game logic etc., a lot of game staples such as popup boxes, scrollable menus, inventories etc. usually require clever "making it work" via various things that are not normally designed for this purpose.

    For example, most of what I create involves some sort of a scrollable panel, be it a level list or a menu, and it's always a pain in the butt to set it up; pinning stuff together, clamping, easing, including invisible objects and whatnot. I understand that in classic programming this sort of stuff may involve even more intricate tinkering but considering how easy and straightforward it is, for example, to have a platformer sprite, it's a bit of a cognitive dissonance.

    Also basic non-game-y elements such as text, textboxes, checkboxes etc. behave their own way since they use the standard HTMl5 controls (e.g. they float over the canvas so you can't cover them with anything). A dream come true would be if all these were rewritten as "native" C3 objects but I assume that would be a whole new undertaking that would also take some focus away from what's most important.

    Still I surely wouldn't be averse to some more Ui related features/facilities/tools (the frickin scrollviews!).