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!
When you navigated directly to ItemPage, the new `safeImageUrl` function would crash during the loading state, because it was trying to safe-ify `undefined`.
Now, I've just made `safeImageUrl` more resilient to that particular kind of unexpected input, by passing through null-y values without change.
When we decided to start out with /api/assetProxy, we didn't know how much the load would be in practice, so we just went ahead and tried it! Turns out, it was too high, and Vercel shut down our deployment 😅
Now, we've off-loaded this to a Fastly CDN proxy, which should run even faster and more efficiently, without adding pressure to Vercel servers and pushing our usage numbers! And I suspect we're gonna stay comfortably in Fastly's free tier :) but we'll see!
(Though, as always, if Neopets can finally upgrade their own stuff to HTTPS, we'll get to tear down this whole proxy altogether!)
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.
I noticed that, if you're _reading_ the beta callout it's obviously a feedback link, but it's easy to glaze over "Tell us what you think". Here, I've added the word "feedback" to make it stand out on scanning the page, while adding "Got ideas?" to keep it feeling colloquial.
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.)
Oops, I shipped with the images.neopets.com TODO undone! Also, the white background was intersecting with the close X for the feedback form.
In this change, we move the xwee image into our bundle instead of depending on images.neopets.com, and we edit it to have a transparent background, which looks nicer for dark mode. (And we do a srcset!)
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.
Not 100% sure on the copy, but I like that it's a bit clearer about the value prop. I tried to work in customization for SEO, but it feels too clunky in a sentence, might need to put it elsewhere in the copy!
These changes are most relevant for playing around in the dev server, modeing against an empty database. But they'll also help in real-world modeling scenarios! e.g. modeling a new species/color combo is now a bit nicer, we don't show a blank entry in the color picker
I think it's great that we hide the button when it's not relevant, but that makes it hard to know that it exists. Here, we do some cute tricks to blink up the "Paused" button when it first appears, even if the user doesn't have the controls visible right now
We did this a while back too, but I guess something changed in Apollo: I guess it used to return identical item objects from the cache on its own, and now it returns brand new item objects. So we gotta do the object caching hacks ourselves!
This speeds up add/remove item state updates from 500ms to 100ms on my Mac, because we stop re-rendering all the Item components and their complex Chakra children.
This is especially worth doing now, because animations make long updates much more noticeable! (It interrupts the animation 😅)
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
This is really very cute, but too many items it turns out are lod despite not actually being animated 🙃 it's helpful for looking for test cases tho, so I'm keeping it, but support only!
I also ended up really liking the icon-badge+tooltip design as a way to summarize lil things, so I'm trying Own/Want short badges in the same style.
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?
notable layout change is that the text content will now try to center itself, and we push the buttons off to the right. we also needed to tweak the layout code a bit to get the buttons to feel centered with the top two lines, bc centering against the full block just feels wrong, they want to be top-y in terms of positioning, but still feel centered-y in terms of visual balance
Honestly I don't think this is the long-term home for the Modeling link, I think it'll become a homepage-only link as we add more modules there. But I wanted to get it out of the way!
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 :)
Here, we add loading skeletons to lots of individual elements, instead of doing a whole item placeholder skeleton. That helps when coming from pages where we have some data, like name and thumbnail, but things is isNc are still missing.
It looks nice, but also particularly means we can handle the loading for the preview separately, get that started faster and iterate better on it in dev!
Initially the spinner was only used in OutfitPreview, where the background was always pretty dark. Now that we use it in more general contexts, we need a light/dark distinction!
Also went and standardized out the `size` props
Here, we extract a lean WardrobePageLayout component, so that we can bundle it into the main app as a loading state for WardrobePage.
This means that clicking Start from the homepage will, instead of flashing the screen to white while WardrobePage loads, show the correctly-sized black/white page layout instead.
I think the Chakra upgrades made these overrides stop working? added !important so that they happen again!
The regression meant the homepage looked worse, always having the selects fade in :/
I think I got all up in my head about direct queries for this one, because of a previous implementation I had in mind, and I forgot that I could just query species and color from the cache by reference without breaking out of the API provided to the cache function!
I also learned in here that I _can_ look up things from the root by doing `readField("allSpecies", {__ref: "ROOT_QUERY"})`, which I struggled to figure out my previous time. I couldn't figure out how to read an uncached field with arguments (I couldn't quite figure out how to build a proper FieldNode, and passing the string form seemed to provide `null` to the `species` cache field reader), but it's probably doable!
(the permissioning happens on the backend in the prev change! but yeah we send the auth token in the headers, so the backend knows who you are and whether to show you private data)
(also it is just owned items not in any list!)
Essentially I want the center-y visual balance around the image, and the name and _one_ line of tags. If there are extra tags, I want that to go down on its own, rather than bringing down the image+buttons to center against them.
The single-line-of-tags case is the most common, and I think it makes things feel more consistent for all the items to stick to that basic layout, while trying to keep that layout feeling center-y
I was getting annoyed by how, when you're using search, trying on an item will remove conflicting stuff, and then if you decide you don't like what you tried the old stuff _doesn't come back_
As of this change, it does! When you start a new search, we save the outfit state, and then whenever you change the items we ask "hey can these old ones safely be re-worn again?" and re-wear them if so.
On the search panel, not all items have the remove button, and it's confusing to have the other buttons be in inconsistent positions!
Move the Remove button to the left side of the list, so that everything else is positioned the same regardless
Okay, we handle the new pages correctly! Still some weird bugs when you send requests near each other? Probably wise to migrate to Apollo's new way of doing this
This is just an implementation thing, but I realized we can just insert the Zone data into the initial Apollo cache, instead of doing weird field definitions
I _do_ still want the @client tags in the queries though, to tell them not to make server requests at all
This was a subtle little thing for a while! If you switch species/color, such that an item doesn't fit the pet anymore, we used to just hide it. Now, we show it in a list, so that you can understand what went wrong, and have the option to remove it.
This was bothering me, I'm surprised and pleased by how easy it seems it was to fix? :)
The strategy is just, look for groups that are provably redundant, and filter them out.
I hope it's correct! It's definitely cozier. Kyrii Mage items are good tests, they have a lot of interesting zones!
I figured that we'd want simpler UI in the ItemsPanel when possible… but now that we've got it pretty simple and comfy, I think the consistency is better
These are nice! :)
The `hideSimpleZones` option I'm not sure about yet, but I figure that:
1. For a new user just doing simple outfits, I feel like the double data on the items page just looks silly, so I want to streamline for that
2. But I _do_ want to let the user think about zone complexity when things _are_ multi-zone.
I did also consider just hiding the zone badge for the header you're under, but I figured the consistency of having the item and its badges look the same in all the places in the list was more important.
This is in preparation for hiding bio zone restrictions but showing item zone restrictions!
I also refactor the build-cached-data script substantially, to run GraphQL against the server instead of a custom query.
okay so the PetAppearance restrictions are stored on the asset, because that's how they're defined on Neopets.com too
but I think that's a confusing API, so here I define `PetAppearance.restrictedZones`, which just maps over the layers and aggregates the zones server-side, same as we would have done on the client
I think that's much easier to understand than having layer contain a field, but having to know that item restrictions _don't_ work that way, you know?
Huh, maybe this is a Firefox bug or something? but the container wasn't applying partial opacity correctly to its children, it was only doing 0 or 1, I think maybe because the children weren't static? I refactor here to make the DOM structure a bit more natural, and fade ins work again 🤷♀️
I'm not sure when this regressed, but changing the outfit was clearing out the whole preview and showing an empty loading state, instead of the intended behavior of showing a loading spinner over the old preview. This affected both the home page and the wardrobe page.
Yeah, so, huh. Fixed! I hope I didn't goof anything else with these effect trigger changes though 😅
The reason I'm doing this now is not just that it's annoying, but as a pair with the color change fix from just now, I want those color changes feeling buttery smooth!
Previously, when changing a pet's color, we would refresh the items panel and send a new network request for the item appearances, even though they're all the same. This is because item appearance data is queried by species/color, for ease of specification.
But! Item appearances are //cached// by body ID. So, if this is a standard color, it's not hard to look in the cache for the standard color's body ID!
Now, most color changes are faster and don't flicker the item panel anymore. We do still refresh the panel and send the requests for color changes that _do_ matter though, like standard <-> mutant!
Oops, opening the dropdown would auto-focus the left arrow button, because it's now the first focusable element! Make explicit that we want the dropdown instead.
Easier to move between appearances quickly! I'm adding this in anticipation of a use case of rapidly labeling Unknown appearances—I want it to be easy to label one and go to the next!
For most users, I want to hide the pose picker if there's not actually anything to pick from.
But I want Support users to be able to open it and use Support mode, to inspect and label Unknown poses!
In this change, I add new UI to show a question mark pose picker button, and a note explaining the empty picker; but only show them for Support users.
I also made a new `useSupport` hook, to replace `useSupportSecret`, now that I have a use case for "is support user?" that doesn't require the secret. Will migrate the rest!
Previously, toggling between the two PosePicker modes could dramatically disrupt the layout, because Popover didn't know to re-measure itself.
In this change, we add a hacky workaround to simulate a window resize event in moments when we know the content is changing. (The realistic ideal solution would still have these manual triggers, but would use an official API in Chakra to notify the Popper instance directly.)
Still just read-only stuff, but now you can look at all the different poses we have for a species/color!
Soon I'll make the pose/glitched stuff editable :3
Some sizable refactors here to add the ability to specify appearance ID as well as pose… most of the app still doesn't use it, it's mostly just lil extra logic to make it win if it's available!
(The rationale for making it an override, rather than always tracking appearance ID, is that it gets really inconvenient in practice to //wait// on looking up the appearance ID in order to start loading various queries. Species/color/pose is a more intuitive key, and works better and faster when the canonical appearance is what you want!)
similar to the layer zoning tools I just rolled out!
not thrilled about the outfit state hacks here bc of how we cache restrict on the appearance rather than the item, but oh well! this escape hatch is pretty easy and solid, and it's a cleanup for another day
Also did a code split here, now that this file is getting larger, to only load this for support users. I don't actually care about restricting console stuff to support users (I'd honestly rather not), but saving the bytes is worth it I think, since support mode is pretty easy to enter when we need to
This is in support of a caching issue in a hack tool coming next! Without this, the change to ItemAppearance restricted zones would make other ItemAppearance fields go missing (bc our hack tool didn't also specify them), so the query would re-execute over the network to find the missing fields we overwrote with nothingness—which would undo the local hack change.
A dev tool to preview what an item would look like in a different zone!
To use, open the browser console, and type `DTIHackLayerZone(layerId, zoneIdOrName)`, but replace `layerId` with the layer's DTI ID, and `zoneIdOrName` with either the zone ID or its name in quotes.
Previously, we were using a custom-y `id` field to help Apollo cross-reference `petAppearance` queries with the results from bulk `petAppearances` queries. Now, instead, we deprecate `petStateId`, and start using `id` to have the same stable value!
This is in anticipation of pet appearance support tools: a stable ID will make it easier to edit them, esp changing their pose (which would otherwise have changed the ID!)
Previously, we would load all `petAppearances` in `PosePicker`, and use cache keys to instantly find it again as a single `petAppearance` in `OutfitPreview` after switching poses.
In this change, we instead have `PosePicker` explicitly load all 6 poses as separate `petAppearance` queries. This simplifies cache sharing between the two components' queries: Apollo can do it automatically, because they were queried the same way in the first place.
I'm doing this in preparation for changing the `id` field of `PetAppearance`, to become `petStateId`. This will help me build pet appearance support tools, by giving the appearances stable identifiers that won't be affected by editing which pose an appearance is!
Dice reported this, thank you!
My mistake here was that `loadImage` _does_ reject when the image fails to load… but it ends up throwing `undefined`, since I forgot to pass the error along from `onerror` to `reject`!
So we would cancel stuff, but then store `undefined` as our error in state, which our component interprets as no-error.
I tested this by using Firefox DevTools request blocking!
On the homepage, I want to keep the ability to enter invalid species/color pairs, so that you can say "Alien, Aisha" instead of having to pick the Aisha first.
But on the wardrobe page, we were rejecting invalid state changes anyway, so I decided to remove invalid color options from the list. And I added an ability to still switch to any species, and potentially resetting to a basic color automatically to match.
Oops, our loading state logic was eating the error case! I'm not sure exactly where the gap was happening, but I've rewritten the states to be a bit more foolproof, since that first condition was confusing I think.
Ahaha I fucked up a bit! I was indexing into the array of cached zones, instead of looking up by ID. This meant that all zone names were wrong, and some search results weren't loading bc there was no zone data!
I made a fix here, and also added some fallback values, so that if there's an issue in the future we can at least fall back more gracefully than the infinite-spinner case we had here.
In this change, we cache the zones table as part of the JS build process. This keeps the database as our source of truth, while aggressively caching the data at deploy time.
See the new README for some rationale!
I tested this by pulling up dev Honeycomb, and observing that we no longer run db queries to `zones` in the new traces for the wardrobe page. (It's a good thing we did it this way, because I noticed some code in the server that was still loading the zone anyway, and fixed it here!)
An even better resolution mode for "Dyeworks Pink: Peaceful Tree Garland", which doesn't leave any big chunks or holes, and instead takes all the leaves and gives them chunky outlines 😅
The "Dyeworks Pink: Peaceful Tree Garland" was a tricky case, with animated falling leaves…
we decided that having transparency in the main pet area, and some incorrect transparent holes in the trees, was a better conflict resolution for this one
Probably would be good to manually upload a Totally Good version, but like, this flag is probably good to have
Hey wow this was not so hard, just set some global styles, removed some hardcoded colors, and walked through the remaining hardcoded colors to pick a dark mode variant :) neat!!
oops another thing I missed in the refactor! I wasn't providing the selected pose to the hook, so it wasn't styling any of the poses as selected. Now it does again!
Oops, I removed some fields from the pose object, so the aria-label for the options were coming out as "undefined and undefined"!
Now they come out correct!
I tried to get the alignment juuust right in Chrome and Firefox, but Dice's computer was still including a 1px sliver of border.
In this change, I'm adding 1px of extra padding between the border and the Flash area. I figure that, if the screenshots are incorrectly aligned by 1px, that's no big deal and I'm okay accepting that slightly-incorrect upload rather than having folks not be able to use the tool!
This happened in an earlier change too, I forget where, but there's some compiler bug that causes comments placed in this kind of spot to yield this error in production, but not in dev:
https://reactjs.org/docs/error-decoder.html/?invariant=152&args
Oh well, moved the comment!
On mobile, it was pretty annoying that you had to show the controls by tapping the preview area to simulate a hover—because it could also click the underlying elements, and do bad stuff!
Here, I add a click capture to prevent the clicks from going down if the controls aren't visible. And I add a toggle, so that you can dismiss the controls, like how YouTube feels :)
Huh, weird how we seem to need preventDefault for buttons, but stopPropagation for links? idk, in any case, the Remove button was doing that thing where it clicks Remove but also bubbles up to the container and clicks it too :p
This was a surprisingly big win! Item is heavier than it looks, because it has like 6 Chakra components, which aren't expensive but aren't _cheap_ in a re-rendered list that needs to be fast, you know?
And it's even more important on search, where there's a lot of items on the page. (we should virtualize it too but that's a thing for another day)
I noticed that item wear/unwear is slow on mobile, because we re-render the whole app tree, and my laptop handles that super fine, but my few-years-old fun takes ~300ms, which is very noticeable.
There's some hacks we could do to get faster feedback, but first I'm diving into the render tree to find the unnecessary renders and stop 'em! That should help build perf across the board, rather than in just one spot, and hopefully be less of a weird sore spot :)
The button section was capable of flex-shrinking, and having those wrap pretty much always looks bad imo. Here, we get more precise about the areas and their flex rules, including only letting the name area grow/shrink. (If the screen is too small for the name to wrap further, the panel area container gains a horizontal scrollbar, which feels like a really good compromise imo)
I think it's more helpful to auto-switch: if you're using the species/color picker for pet compatibility, that's probably what you want! And it's all buffered behind the Save Changes button anyway, so nbd to play a bit fast and loose!
One fix was that `position: absolute` stuff was appearing over the drawer, including the item search clear button, oof!
Another fix was for a weird bug(?) in Apollo Client, see comments
Note that there's a bug when switching back to the null case… when I look in the Apollo dev tools, it's definitely getting set in the cache correctly at the right time… but the query isn't updating for some reason? I'm hoping it's an Apollo bug that will fix itself someday with an upgrade!
Optimistic UI seems to just be like, not working… I'm seeing some Google results suggesting maybe just get to v3, which is a bit of upgrade work but might be worth it
