Whoops, `Promise.race` isn't quite what I wanted here. This meant that, if the image promise _fails_ before the movie _succeeds_, the outfit would crash even though it doesn't need to. (And this was happening too often, due to a bug in /api/assetImage!)
Now, we accept whichever _successful_ result loads first, or reject if they _both_ fail.
I tested this by having /api/assetImage always throw, and confirmed that it crashed the outfit before this change, and no longer does after this change!
Doing this for two reasons! One is that I want the movie layer component to be a bit thinner in general - I think we might even want to move the fallback image logic out, too.
The second is that I want the onError for something else soon!
Oops, my inbox was getting full of uncaught promise rejections of `loadImage`!
I'm pretty sure they're caused when multiple images in a movie fail to load (e.g. network problems), but we fail to cancel them. So, the first failure would be caught as a part of `Promise.all` in `loadMovieLibrary`, but then subsequent failures wouldn't be caught by anything, and would propagate up to the console and to Sentry as uncaught errors.
In this change, we make a number of improvements to cancellation. The most relevant change for this bug is that `loadMovieLibrary` will now automatically cancel all resource promises when it throws an error! But this improved robustness also enabled us to finally offer a simple `cancel()` method on movie library promises, which we now available ourselves of at call sites, too.
I'm seeing uncaught promise rejections in `loadImage`? It's hard to know exactly where it's actually coming from, those _should_ be caught?
My guess is that it's coming from canceled images, which are throwing errors even after loading? I don't totally understand how, because looking back, I don't think the `cancel` method was actually called???
Anyway, I fixed it so cancel actually _is_ called, and that we don't throw errors when the canceled image _correctly_ fails to load.
This should be more robust either way, but hopefully it also stops the flow of errors?
Right, when there are zero layers, we shouldn't say we're loading!
This is a consequence of the HACK below. If we didn't short-circuit the effect when length == 0, then we would go through and successfully load 0 layers.
Movies often have a lot of assets, which are more likely to be cache misses, and take script time to render! So the time until the user sees something is often huge.
Here, we start loading our PNG image at the same time. This is a filesize loading increase, but even in slow connections, it's generally worth it as a _sharp_ improvement in time until you get to see something!
One noteworthy UI weakness here is that we don't show _any_ loading indicator while the image is visible and the movie is still loading. This makes sense from a practical standpoint, but could be a problem when a movie takes a particularly long amount of time. I also want to be cognizant of whether the blink-of-content ever gets annoying! (We could make it fade out 🤔)
In my last change, I didn't try to change the APIs too much, and kept the concept of `crossOrigin` running through `getBestImageUrlForLayer`.
Now, I've moved the `safeImageUrl` call _outside_ `getBestImageUrlForLayer`, by putting it at the call site: We now call `safeImageUrl` from `loadImage` (which needs to know the `crossOrigin` flag anyway!), and at the `img` tag call site.
This simplifies all of the call sites a lot, I think!
I've noticed that our Fastly proxy adds a surprising amount of latency on cache misses (500-1000ms). And, while our overall hit ratio of 80% is pretty good, most misses happen at inopportune times, like loading items from search.
But now that the Neopets CDN supports HTTPS, we can safely switch back to theirs for *most* image loads. (Some features, like downloads and movies, still require CORS headers, which our proxy is still reponsible for adding.)
This forgoes some minor performance wins (like the Download button now requires separate network requests), and some potential filesize reduction opportunities (like Fastly's auto-gzip which we're today using for SVGs, and eventually using their Image Optimizer for assets), to decrease latency. We could still potentially do something more powerful for low-power connections someday… but for now, with the cache miss latency being *so* heavy, this seems like the clear win for almost certainly *all* users today.
Oops, okay, I guess I didn't test the new preview centering stuff with 1200x1200 images, like the Usul's Damask Markings.
Now, I apply a max size to the whole-ass container, and make the parent responsible for centering it.
So I finally started looking into the race condition that makes item previews sometimes fail to load, and as expected, it was that we were trying to load the movie before CreateJS had necessarily loaded. Usually the timing worked out, esp after a reload, but not under certain circumstances!
Anyway, I've been wanting for a while to just bundle them instead. That'll help us more eagerly load them when we need them, and not depend on external CDNs, and remove a bunch of loading state!
So yeah, I had to learn how the `easeljs` and `tweenjs` NPM packages did their bundling, and how to use `imports-loader` to let them just register straight onto `window`! But we got there and it's pretty nice tbh!
Woof, the "Swirl of Power Effect" item tanks my CPU waaay too much
(I bet it's those 7000x7000 PNGs lolol 😬)
Anyway, before thinking about optimizing specific issues, I'm just adding this emergency switch: if we detect FPS < 2 on any layer, we just pause the whole outfit, until the user decides to unpause.
Ah oops, because I forgot to set `hiResMode` here in the image preloader, we would preload the PNG, and _then_ load the SVG separately.
This doubled the effective image loading time in Hi-Res Mode!
Now, the image preloader respects hi-res mode, and will preload the SVG in the SVG case, and the PNG in the PNG case.
We're just having too many glitchy SVGs for my taste, esp since TNT seems to just be using PNGs for now?
This change defaults us to using PNGs for users by default, with the option to use SVGs as a new "hi-res mode" setting.
This is our first ever setting, wow!
I'm also envisioning that like, if we get Fastly Image Optimizer set up, this could be a way to tune the quality of the incoming images.
We could also consider a setting to turn off animations altogether—like, just download the PNG instead of the movie, whereas right now we download the movie on the assumption that you might play it at any time.
Oops, it was possible after saving an outfit to get into a state where we would show the `<OutfitThumbnailIfCached />` behind the outfit even after it was saved, and then removing items would look weird until auto-saving caught up.
We had used the `backdrop` property because we wanted smoother partial load-ins, but for now I'm just fixing this by switching it to `placeholder`, which already has the right loading-only behavior.
This was also the only call site for `backdrop`, so I've removed it!
The layer preloader already takes advantage of, and primes, the HTTP cache.
But we still do duplicate work, on every OutfitPreview render, to re-execute movie clip libraries, and create a movie clip to test for animations. The former is nontrivial cost, and the latter is often large cost. This can make even basic outfit changes slow, when there's no change to the movie clip layers and the player is paused!
Here, we add an LRU cache for movie clip libraries, and for the question of "is it animated?". This should speed up a number of places where we would reload the movie (including between toggling the item), and various changes that were triggering full movie clip rebuilds unnecessarily.
We _aren't_ solving here for the fact that toggling an animated item requires rebuilding the movie clip, which could conceivably be cached—but with some state management trickiness, because ideally it should be a separate clip for each context where it's being shown. Imo not yet worth the effort! (esp because I think users understand that toggling an animated item can be slow, whereas this was affecting _other_ actions way too much)
This was one more bit that needed fixing for "Flying in an Airplane": it wasn't just the official SVG that was incorrect, but also the official SWF. So our converted PNG was also incorrect!
Here, we now try to use the official Neopets PNG when the manifest provides it, instead of our own.
Oops, I never actually saw the practically invisible text in light mode! Let's make it actually dark in light mode item pages, and still dark in all wardrobe pages!
Here, we offer a second syntax for `<OutfitPreview />`: a hook that offers the same UI as `preview`, but _also_ shares the `appearance` data.
This makes it easier to have UI that depends on the outfit appearance, without having to commit to all the `useOutfitAppearance` stuff in the parent. Same easy syntax! :3
I've refactored the item page to use this for compatibility testing, instead of using the Apollo cache (which was also cute and same perf impact, but more overhead!)
Huh, I'm not sure why SVGs ever didn't have `crossOrigin: "anonymous"`? The old commit isn't really super helpful for understanding that. Maybe I just didn't notice the problem in that case?
Well, whatever. Let's see if this breaks something else! (I'm also wondering if we should just like, _always_ ask for things with crossOrigin set?)
To help the load time for outfits feel shorter, we now reuse the outfit thumbnail from the Your Outfits page as a placeholder!
This doesn't add any overhead when the thumbnail data _isn't_ in your session cache, e.g. if you navigate to the outfit directly. But if we have the thumbnail on hand already, we just show it, easy peasy!
Oops, when switching to @emotion/react, it turns out they no longer support that cute hack I was doing to append suffixes to class names!
Here, I change strategy and let `CSSTransition` apply the plain `exit` and `exit-active` classes to its children, and apply Emotion styles to the child to check for _also_ having those classes.
Looks like there was some kind of runtime conflict when running @emotion/css and @emotion/react at the same time in this app? Some styles would just get clobbered, making things look all weird.
Here, I've removed our @emotion/css dependency, and use the `<ClassNames>` utility element from `@emotion/react` instead. I'm not thrilled about the solution, but it seems okay for now...
...one other thing I tried was passing a `css` prop to Chakra elements, which seemed to work, but to clobber the element's own Emotion-based styles. I assumed that the Babel macro wouldn't help us, and wouldn't convert css props to className props for non-HTML elements... but I suppose I'm not sure!
Anyway, I don't love this syntax... but I'm happy for the site to be working again. I wonder if we can find something better.
We do animation detection during the preload now, but this wasn't always working correctly: some movies don't actually fully mount the children onto the stage until we start playing. This caused the play/pause button to be missing on the outfit page and the item page, but the animations would still play, depending on the user's saved play/pause state in localStorage.
Not sure why movie clip building is failing! But it happened outside our try-catch, so it left us in an infinite spinner state.
The repro item is the Spring Topiary Garden Background!
I guess something got more picky about the loading sequencing: the fade in animation was happening faster than the cached image could load. Now, we explicitly wait for the image to load (even though we know it's probably cached) before fading it in.
Oops, our movie layer promises don't have a .cancel() method, so calling it crashed our error handler. Now, when there's an error loading a layer and there are HTML5 layers visible, we'll correctly show the "Could not load preview. Try again?" message.
So I broke the Download button when we switched to impress-2020.openneo.net, and I forgot to update the Amazon S3 config.
But in addition to that, I'm making some code changes here, to make downloads faster: we now use exactly the same URL and crossOrigin configuration between the <img> tag on the page, and the image that the Download button requests, which ensures that it can use the cached copy instead of loading new stuff. (There were two main cases: 1. it always loaded the PNGs instead of the SVG, which doesn't matter for quality if we're rendering a 600x600 bitmap anyway, but is good caching, and 2. send `crossOrigin` on the <img> tag, which isn't necessary there, but is necessary for Download, and having them match means we can use the cached copy.)
Previously I tried to be clever and pre-optimize by putting all the layers onto one canvas… I think this probably helped by batching their paints, but it made fades less smooth by not taking advantage of native CSS transitions, and it made us dip into JS way more often than necessary.
Here, I take the simpler approach: just layers of <img> and <canvas> tags, with each animated layer on its own canvas, and letting the browser handle transitions and compositing, and separate `setInterval` timers to manage their framerates.
I have a suspicion that batching the paints could help performance more, but honestly, maybe that batching is already happening somehow, because things look pretty great on my big-screen stress test now; and so if it _is_ relevant, I want to wait and see after testing on low-power devices.
Honestly kinda surprised this worked on the first go! I was worried something about the process would make the sorta like, instant-cache expectation not work.
Still thinking it might be considerate to like, keep a LRU cache of MovieClip options, so that we don't double-execute these scripts when adding stuff… we even re-execute the ones already applied lol 😅 and that adds lots of script tags to the body!
But yeah I'm not gonna push on it yet until I see evidence that it actually causes performance issues in practice
reflecting further on the abstraction, I'm noticing that this isn't an Easel abstraction like I envisioned early, and that we're baking some Neopets stuff into it. And I think that's the right call, esp with the tricky MovieClip stuff coming up, where I think more barriers would hurt more than they help. So, a new name!
Did it revert? Or did I just never notice that it only worked on mount, not on new loading states?
Also, fixed a bug where we were injecting the script tag way too much, and triggering loading too much that way too!
still just for static stuff, but it's good to be working!
PosePicker got a bit broken, CSS scaling doesn't work quite right anymore, we might need to just up the internal resolution or something?
Fading the whole preview to a black overlay for the loading state was feeling aggressive, especially since loading delay wasn't working correctly!
In this change, I fix loading delay, and I add a nice subtle "corner" variant for outfit preview spinners :)
