• Content Count

  • Joined

  • Last visited

  • Days Won


Jammy last won the day on July 22

Jammy had the most liked content!


About Jammy

  • Rank
    Advanced Member

Profile Information

  • Gender
  • Location
  • Interests
    Games, dev, pixi, web, mentoring

Contact Methods

  • Twitter

Recent Profile Visitors

2755 profile views
  1. Web games have always been traditionally free, or MMO based P2W/Microtransactions. Gamers seek either the free lunchtime filler, or the long term MMO in the browser, the need for a platform specifically designated for fleshed out HTML5 games is very low. It's probably only other developers who care how it was made. As you know though, the potential to develop the game for desktop and release it on Steam is still there. Just gamer's don't care how its made, as long as it's a good game. If anything it might deter them to know it was made on web technology rather than blistering fast native technology, but that's just anecdotal. If you really want to be a hardcore platform that makes money, then I think MMO/Persistent/Mobile game is the way to go.
  2. I believe Minecraft was made by one individual. Making a game as big as minecraft is about time and effort more than skill. If you spend 5 years making something it will be playable. you won't be making minecraft over night in HTML5. The easiest way is with Unity or Unreal since the engine is practically there. Also remember for 1 success of minecraft, probably 10,000 other indie games failed to make profit around that time and that number is growing every year.
  3. @louislourson this project is so cool! I love oxygen not included, can't believe the amount of effort thats gone into this
  4. Jammy

    Sprite compression

    @ivan.popelyshev @Exca Thanks guys really appreciate the information. Looks like I've got some references here to start looking at, Thanks!
  5. Jammy

    Sprite compression

    Hiya, I'm wondering about png and sprite compression in general, and how PIXI see's it. Note: these questions are not about the network or other web impacts, I'm interested in the point where PIXI gets the data and loads it then renders it. I'm wondering first, if a PNG is compressed e.g. down to 20% from its original size, would this have any bearing when it came to pixijs loading or rendering either, for example one being 100mb and the next being 20mb, but the dimensions and contents (from a visual perspective, not data) of the file remain the same. Secondly, I'm wondering if there are compression techniques that I'm not aware of for sprites, not simple spritesheet tricks but any compression algorithms or package type things. Any feedback/input much appreciated! Thanks! Jam.
  6. pixi-v5 zIndex + helps a lot thanks, So i have two types of sorting going on now, one on the layer level and one on the objects contents (collection of sprites inside a container). I've got bloom on a mixture of layers coz of this so I assume im probably going to see some perf hit at some point, but I think I can control that. Thanks!
  7. Hiya, thats just a logic choice I think. If(myTurn){allowDragging = true} else {allowDragging = false} If it's a turn based game probably store somewhere whos turn it is, or have an event u run when it changes turns. Addition: It sounds like in your original post you're worried about using globals. In my opinion, its your application and game therefore the entire scope is there for you to use; unless you're thinking about the game being integrated into other peoples projects at some point, in which case then it should definitely be considered.. You should though create a nice structure, for example game.state contains game state, game.events contains events etc. In my projects, although its a bit ropey, i have one global object called game, and everything else is a part of that.
  8. @HimeWorks onmouseup on the sprite they drop it on if(game.dragging==true) {hitsquare=this} hope that helps.
  9. Hey, I'm using pixi-layers and I have 1 layer which is for bloom, I add an object to that layer whenever I want bloom (I believe it was also better practice to add all objects for a specific filter to one layer and apply the filter to the layer, for performance). When I apply parentGroup to a bloom object, I believe pixi-layers swaps its layer/container and therefore it loses its filter. Is this right? Is it possible at all, to have a layer full of objects, that still sort on groups, without changing container? I guess it may be impossible. Do we have any idea the difference it may cause if I add the bloom filter to sprites individually (may be up to 100-200 sprites) Thanks, Jammy.
  10. Pretty fun actually, only one thing really. This piece unbalanced. Needs to happen after score 20 or something.
  11. I would recommend not using mousecontraints etc from matter, just the matter physics, if you want your game to have physics, add them to matterjs as required (as bodies or whatever) and calculate the constraints yourself.
  12. Does toWorld work? https://davidfig.github.io/pixi-viewport/jsdoc/Viewport.html#toWorld
  13. Hi, Potential workaround: Check if mouseclick is being fired for the one underneath, then clone the mouseover code into there? Edit: I have achieved this actually... When you click the sprite you want to drag, set its alpha to 0 and leave it where it is, create a replica of that sprite with a hitArea= rect(0,0,0,0) and use that to follow your mouse. Now when you move it over the other sprite, the pointerOver event should still fire. If you then drop the mouse-sprite, just destroy the original sprite, or move it to the new location and destroy the mouse sprite. Or if its a failed movement (i.e. they dropped it in a bad place) simple destroy the mouse sprite, and set alpha back to 1 on the original. Free-rollback! So whatever your application is, there may be potential for creating a ghost copy or setting the hitArea's to 0 temporarily to get the desired results. It's buggy at first but can be nailed eventually.
  14. Jammy


    Yup, this method works but is not flexible (e.g. adding new armours to the game later, etc) but if you want just the two sprites combined, for sure do it in photoshop or something. But again, keep in mind this is exponential, every time you add a new "layer" e.g. hat or armour or a new sprite sheet, that is a new amount of textures required goes up exponentially. Also this will have an impact on the network, for each PNG the user will have to download that spritesheet, rather than the raw sprites that are later stacked. Have to say though, this is going very deep into the rabbit hole, and it's always best to KISS (keep it simple silly). Get something working and experiment.
  15. Jammy


    You savage! Haha. In my opinion the method you have suggested will work only if there are not many combinations, for example if you have 100 bodies and 100 armors, thats 100x100 = 10,000 textures. You may also introduce boots, hats, etc which would then square this further. The same example with separate sprites would mean just 200 textures (for 1 direction/sprite). edit; Further to this, if you have used the container method, you could extend the container (class Player extends PIXI.Container) and then inside the class have "setDir" as a function, which updates both sprites.)