WackyToaster's Forum Posts

  • This has already been filed as an issue --> https://github.com/Scirra/Construct-3-bugs/issues/1396

    Where Ashley said that it´s an issue with chrome --> https://bugs.chromium.org/p/chromium/is ... ?id=823247

    Where it was fixed a couple of days ago actually. Based on https://www.chromium.org/blink/when-wil ... -or-canary this will take ~10 weeks before it´s shipped into the stable chrome version, so we need patience, or use https://www.google.com/chrome/browser/canary.html (I wouldn´t though)

    The problem is if you search for the issue on github (with "zoom") you will not find it unless you delete the "is:open" because Ashley closed this error. I think what would be great is some sort of bugtracker, possibly even inside of construct, that shows the current open bugs and if they are beeing worked on. So if you encounter an error, you could quickly check if it´s already been filed aswell as it´s status without even leaving construct. This will likely lead to some amount of overhead, but also might lead to reduce duplicate bugreports and frustration. I mean the bug has already been filed, fixed and will roll out soonish, so a user shouldn´t be left uninformed needing to make a "help pls" thread, and Ashley doesn´t need to be told for the 15th time that there is a bug.

    At minimum, Ashley shouldn´t close an issue unless it´s actually been 100% fixed and released, because else users won´t find it (that easy) when they search for it and just end up filing them again, posting suggestions, make threads about it, etc. I personally almost filed that zoom issue again recently, I don´t even know anymore how I found out it has actually been filed and is beeing worked on already.

  • Absolutely, basically just have it like the "Else" I think this has already been suggested on several occations, but the more the merrier

  • Try Construct 3

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

    Try Now Construct 3 users don't see these ads

  • I wouldn't worry too much about performance in your case. You have more or less a single moving physics object. Static objects don't have much of an impact. On PC you can easily use quite a bunch, on mobile you should limit it a bit. There is no real difference between webbrowser or node webkit (that is still basically just chrome). It's probably best to test it out, see how many physics objects you can drop until the framerate tanks.

  • Short answer: No

    Long answer: Also no, unless you consider replicating all events, sprites, variables, etc... 1:1 in C2 by hand viable (not to mention it might still end up not working as expected because of some differences)

  • I think that you cannot have an "Else" on trigger events (see the small green arrow next to it) directly, so using sub-events is the way to go.

  • Hmm, hard to tell without seeing everything. Maybe it´s the trigger once inside the or block (OR trigger once. In that case you don´t even need it actually), maybe the animation is set wrong, maybe try to play the animation from the beginning instead of current frame... etc etc. If you don´t mind you can share the project file.

  • Datamining happens quite often, even if you try to hide it well. I´m just gonna drop this here:

    [quote:34x4u2w7]

    https://youtu.be/UJVuLvGjPOc?t=8204

    Vinny: You guys put a lot of secrets in Isaac, a lot of codes, a lot of seeds. I remember how you didn't want people to discover them right away. My question to you guys is-

    Edmund: You already fucking fucked it up! They fucked it up.

    Tyrone: They swallowed a hundred dollar steak in one bite.

    E: Yep. It's... man. Tyrone's analogy is right. It's true. They couldn't resist it, they had to break it apart. And even though people were actively working on some of the biggest secrets that would have taken awhile to unfold, they said "NOOOOO! I WANT IT ALL NOW!"

    V: And then they spoiled it for the rest of us.

    E: It's disheartening. It sucks because we're gonna start working on an expansion soon, but we sure as hell aren't gonna fucking take all the time that it took - especially with Simon, who tried his best to really bury the stuff so it would deter people from doing this, but all they see is a challenge, so they're going to dive right in and do it anyway. I can say right now that I don't think the expansion is going to feature any buried secrets that anybody will care about.

    T: It was months of our life, between me and Ed and Simon. Ed had this really crazy idea, we were like "let's do it, it'll take months, some people won't find this for years!" Nope, one week.

    E: Nobody likes to wait anymore. They like to eat it now!

    T: I'm going to challenge the dataminers for the expansion, because I have something that I haven't even told Ed about that guarantees that they won't be able to mine shit. It'll take at least three days this time! (laughs)

    E: There was some stuff in the original Isaac that I thought wouldn't be uncovered for awhile. It ended up being uncovered legitimately, but faster than I thought it'd be. I thought I'd go even further [with Rebirth], but I think I've learned that there's no point in really doing that stuff, and the best way to hide a secret is to lock it in a very challenging area. I think if we end up doing anything like that in Rebirth, it'll just be behind a barrier of entry; having a certain amount of skill. Not something that I thought was going to be a really fun and community-based ARG, which would have been nice if it was allowed to take place, but you know... whatever.

    Sadly the era where you could actually hide secrets is sort of over. If people wanna know they will tear your game apart as they see fit. If you toss roadblocks into their way, they will just try anyway and find a way around. I mean, you might be able to that by using some heavy encryption, and having a server that does the decryption, making the game requiring online just for that. And even then, people can still figure out the decryption, maybe send "fake" data packages to your server to decrypt them so they can just decrypt anything they want.

    Long story short, it´s either a huge chunk of work for very little reward, or you just simply accept the fact that people will datamine your game.

  • When you check "Key is down" this event will fire every tick. So you are telling your sprite to play the animation from the current frame over and over, it never gets to the next frame. Try using "on key pressed" or add a "System > Trigger once". In that case you want to have your animation set to loop.

    Also since you use the same animation for all 4 directions you can create an "OR" block (right click on the event) and put all 4 direction keys inside, so you only need 1 event (unless you just do this for testing and want to replace the animations later down the road).

    Also also if you do intend to replace the animations later on so you have individual animations for up, down, left and right, keep in mind that when a player presses up & right, it will sort of attempt to play both animations so you have to make sure it does not do that.

  • I don´t have that much insight on tilemaps but from what I understand...

    1

    The drawcalls will be the same. Each sprites needs a drawcall, each tile also does. This also means it´s not recommended to have single objects span over multiple tiles, because each tile will be rendered individually. So an entire tree is better off beeing a sprite rather than 20 tiles in a tilemap. I think though there is more going on in regards to rendering. Also:

    [quote:144y045s]Tilemaps also have significant performance benefits over achieving the same results with other kinds of objects, such as arranging a grid of Sprites. The Tilemap object can optimise collision detection and rendering in a way that scales well even with extremely large Tilemap objects.

    I have no idea how impactful this is though, considering construct already uses collision cells and basically doesn´t render anything outside the viewport and such. On a sidenote, people have been building some cool stuff using 1x1 pixel tilemaps, where there are several thousands of 1 pixel sized tiles on the layout with no performance impacts. Maybe Ashley has better answers to that. It also has some downsides though, that´s why I personally never really liked working with tilemaps. For example, you cannot animate tiles and I generally found it a bit cumbersome to work with for various reasons. It also depends on where you want to play the final product. For mobile it´s probably a better idea to use tilemaps because you wanna use that limited power well, on a PC you probably can get away with just using sprites, I don´t think you´d run into a bottleneck anytime soon on any reasonably modern machine.

    2

    Hard to say. If the game is entirely based on tiles, it´s probably a good idea to use a tilemap since it offers tools (like a drawing tool) that are not available with sprites. You have to keep the downsides in mind though. If you want to use animations in all your tiles except a few, you probably should just use sprites. I don´t think you can calculate it.

    3

    No, I doubt that using 4 tilemaps with 1 tile each is the way to go. You defenitely want to use one tilemap with 4 tiles.

    Also always a good thing, the manual --> https://www.construct.net/at/make-games ... ce/tilemap

  • It´s a known issue with chromium. I assume you use the desktop app (and chrome). Simply go to editor.construct.net in chrome, open the menu and set the zoom to 100%. I think the shortcut for this is Ctrl + 0. After that everything should be back to normal.

  • I had a sort of similar issue today, I tried downloading a Sprite from inside the Project (via the image editor) but clicking the download button did nothing.

    What I realised is pretty simple, it attempts to open a popup, and if you have popups blocked this will simply not work. If you use the desktop app and you get that "a popup was blocked blabla" thing when trying to connect to google drive, your popups are blocked. Interestingly you won´t get any error if you try to save a sprite or download an apk. If you go to editor.construct.net it will show a small icon on the right of the URL bar when a popup was blocked, there you can set it to "always allow". Once done the download should work just as well in the desktop app.

    Hope this helps, cheers!

  • Well yes, reencoding might be the solution but I haven´t tried it. I just needed any video file for a small test so I didn´t care that much. The point is that an error imo should tell me where the problem is so that I, as an enduser, understand how I can solve it (unless it´s something obscure that I cannot solve as an enduser). I know programmers love their cryptic error messages but I´d rather understand what´s going on.

    [quote:15e0rfxv]Error: Something happened

  • https://i.imgur.com/8cXzmEM.png

    Tried to import a video and I got this error. Some other videos work fine, changing folder does nothing. What exactly is this error trying to tell me?

    I assume it might have something to do with the video format?

  • [quote:1qiwvofa]The simple fact is, nothing will be easier or more CPU efficient than just using a proper trace/raycast to find out the normal of the surface the player is walking on.

    Shouldn´t it be sort of possible bullet behavior? I mean yeah maybe not as CPU efficient but you can continously shoot invisible sprites down from the player while he is on ground, have them bounce off and get the grounds normal by comparing the initial angle (90) with the angle after the bounce. At least in theory, I haven´t tried it. You can destroy them directly after the bounce so even if you shoot every tick there shouldn´t be a huge amount of sprites. You probably can even get away with shooting it like every 3 ticks if you want to optimize.

    Anyway, I recently tinkered around with the SDK creating https://www.construct.net/at/make-games ... 7/utc-time It´s the first time I ever made a plugin and I´m rather noobish with JS. From that experience I can just say...

    [quote:1qiwvofa]- Having some larger sample/open source plugins for people to learn from.

    • Something to make setting up C3 plugins faster. Currently there is a lot of cutting and pasting names, many files etc. It's for the better, but I think we could make something that would speeds up the process.

    ... basically this.

    Especially the initial setup (UPDATING PLUGIN IDENTIFIERS) gave me some trouble and I found the documentation not sufficient/confusing. I could swear I changed everything but as soon as I changed the plugin ID it stopped working. I don´t remember exactly what ended up solving it. Same goes for the ACEs. In particular, I needed a "combo" dropdown thingy but the ACEs documentation doesn´t show how exactly this has to be defined. There´s just a lot of {[]} going on and it´s confusing if you only get told

    [quote:1qiwvofa]"combo" — a dropdown list. Items must be specified with the "items" property.

    I ended up downloading another addon and check it´s code to figure it out.

    So from my experience: More/better samples, better documentation and it would be awesome to have an easy method for the initial setup of ID, names and references.

  • A very good portion of QOL stuff, new features you probably won´t see in C2, new runtime incoming. A year is cheap enough to give it a go and see if you like it. If not, you won´t starve either way.