I'm gonna also use this for a task to try to warm up *all* the
manifests in the database! But to start, just a simple one, to prepare
the alt styles page quickly on first run. (This doesn't really matter
in production now that I've already visited the page once, but it helps
when resetting things in dev, and I think more it's about establishing
the pattern!)
The Neopets Media Archive is a service that mirrors `images.neopets.com`
over time! Right now we're starting by just loading manifests, and
using them to replace the hacks we used for determining the Alt Style
PNG and SVG URLs; but with time, I want to load *all* customization
media files, to have our own secondary file source that isn't dependent
on Neopets to always be up.
Impress 2020 already caches manifest files, but this strategy is
different in two ways:
1. We're using the filesystem rather than a database column. (That is,
manifest data is kinda duplicated in the system right now!) This is
because I intend to go in a more file-y way long-term anyway, to
load more than just the manifests.
2. Impress 2020 guesses at the manifest URLs by pattern, and reloads
them on a regular basis. Instead, we use the modeling system: when
TNT changes the URL of a manifest by appending a new `?v=` query
string to it, this system will consider it a new URL, and will load
the new copy accordingly.
Fun fact, I actually have been prototyping some of this stuff in a side
project I'd named `impress-media-server`! It's a little Sinatra app
that indeed *does* save all the files needed for customization, and can
generate lightweight lil preview iframes and images pretty easily. I
had initially been planning this as a separate service, but after
thinking over the arch a bit, I think it'll go smoother to just give
the main app all the same access and awareness—and I wrote it all in
Ruby and plain HTML/JS/CSS, so it should be pretty easy to port over
bit-by-bit!
Anyway, only Alt Styles use this for now, but my motivation is to be
able to use more-correct asset URL logic to be able to finally swap
over wardrobe-2020's item search to impress.openneo.net's item search
API endpoint—which will get "Items You Own" searches working again, and
whittle down one of the last big things Impress 2020 can do that the
main app can't. Let's see how it goes!
Preparing to finally move wardrobe-2020's item search to use the main
app's API endpoints instead!
One blocker I forgot about here: Impress 2020 has actual support for
knowing an item's true appearance, like by reading the manifest and
stuff, that we haven't really ported over. I feel like maybe I should
pause and work on the changes to manifest-archiving that I'd been
planning anyway? I'll think about it.
I changed the type of this tag without realizing the JS references it
by both class and `div`!
I think at the time this was a perf suggestion for jQuery, because the
best way to query by class name was to query by tag first then filter?
It's possible our jQuery still does this, but I don't imagine it's very
relevant today, so I'll just remove that for better guarding against
similar bugs in the future instead.
I've moved the support secret into the encrypted credentials file, and
moved the origin into a top-level custom config value in the
environment files, with different defaults per environment but still
the ability to override it. (I don't use this, but it feels polite to
not actually *demand* that people use port 4000, y'know?)
Okay, so I still don't know why rendering is just so slow (though
migrating away from item translations did help!), but I can at least
cache entire closet lists as a basic measure.
That way, the first user to see the latest version of a closet list
will still need just as much time to load it… but *only* the ones that
have changed since last time (rather than always the full page), and
then subsequent users get to reuse it too!
Should help a lot for high-traffic lists, which incidentally are likely
to be the big ones belonging to highly active traders!
One big change we needed to make was to extract the `user-owns` and
`user-wants` classes (which we use for trade matches for *the user
viewing the list right now*) out of the cached HTML, and apply them
after with Javascript instead. I always dislike moving stuff to JS, but
the wins here seem. truly very very good, all things considered!
From an era when we didn't have that! Now we do!
(My motivation is that I'm trying to add new JS to this page and errors
in stickUp are crashing the page early, womp womp!)
This one is important, I didn't notice that this is a way of setting
attributes that won't be written to both tables! `name` will only be
written to the translation table (which crashes the save), and the
other fields would only be written to the main table. Fixed! (I don't
like the super-dynamic this code was written before, anyway.)
Missed this at first - now that the `name` field is just a normal field
and is always English, it's now an error to provide the locale to it as
a parameter, like we used to for the translated version of the field!
Like with Species, Color, and Zone, we're moving the translation data
directly onto the model, and just using English. This will simplify
some of our queries a lot (way fewer joins!), and it's what Neopets
does now anyway, and I have a secret hope that removing the complexity
along the codepath for `item.name` might help speed up large item lists
if we're lucky?? 🤞
Anyway, this is the first step, performing the migration to copy the
data onto the `items` table, making sure to keep them in sync for the
2020 app for now!
I think this was to explain why `order` wasn't part of this query, and
we probably used to sort in the controller? But now the item search
module takes care of all that, this is just confusing to say now imo!
Impress 2020 has had this for a while, I've wanted it for reference on
occasion, let's bring it in!
Very similar logic, and Ruby & Rails's date affordances are super
helpful for simplifying how to express it!
The homepage used to point to old projects that don't work anymore
anyway! This is the only project that stuck, so just redirect here!
We also remove the openneo.net link from the footer, because there's
nothing useful to say there anymore!
It hasn't been updated in a long time, let's just be rid of it!
It's possible I'll replace it with another blog sometime if we get the
chance to do more development work, it could be a useful way to improve
communication—but not yet!
I think I cleared this from the outfits/new template a while ago, but
never cleaned up this file, because I was too anxious that I was
correctly identifying all its call sites. But now I'm more confident!
At least, they seem unused to me on a quick audit! The scriptaculous
stuff has long been replaced by jQuery UI equivalents. (Wow, so many
generations of libraries! lol)
Mostly this is just me testing out what it would look like to
modularize the app more… I've noticed that some concerns, like
fundraising, are just not relevant to most of the app, and being able
to lock them away inside subfolders feels like it'll help tidy up
long folder lists.
Notably, I haven't touched the models case yet, because I worry that
might be a bit more complex, whereas everything else seems pretty
well-isolated? We'll try it out!
Tbh I'm not sure `special_color` is actually used anywhere? It used to
be how we decide what to show in the previewer on the item page, but
that's been replaced with the 2020 logic, so idk…
But in any case, I noticed that the description doesn't match the
pattern we have, so here's the fix!
I looked at this and was like. "ok literally what is
`nonstandard_colors` trying to do"
reading it again now, I'm realizing the idea is that it probably runs
two queries: one to get nonstandard colors, then depends on
ActiveRecord to implicitly convert the relation to an array and then to
IDs for the second query? Instead of doing a join??
Idk, it's unused, so trash it!
This used to be the behavior, and the site has plenty of graceful
fallbacks for it, I just forgot this one when doing Rails upgrades!
Note that the impress-2020 stuff is *not* as graceful about this, so
the wardrobe page won't show the pet until the color is in the DB. Ah
well, still an improvement!
I thought this refactor of this change was working, but actually it was
just failing to build the JS lmao. Here's a version with correct syntax!
😅
Is there a syntax for this kind of thing that I'm just forgetting? Idk,
oh well!
Okay right, the wardrobe-2020 app treats `state` as a bit of an
override thing, and `pose` is the main canonical field for how a pet
looks. We were missing a few pieces here:
1. After loading a pet, we weren't including the `pose` field in the
initial query string for the wardrobe URL, but we _were_ including
the `state` field, so the outfit would get set up with a conflicting
pet state ID vs pose.
2. When saving an outfit, we weren't taking the `state` field into
account at all. This could cause the saved outfit to not quite match
how it actually looked in-app, because the default pet state for
that species/color/pose trio could be different; and regardless, the
outfit state would come back with `appearanceId` set to `null`,
which wouldn't match the local outfit state, which would trigger an
infinite loop.
Here, we complete the round-trip of the `state` field, from pet loading
to outfit saving to the outfit data that comes back after saving!
Now that DTI 2020 has been deployed without references to the
translations tables, we can stop keeping them in sync!
Next step is to drop the tables and be done with them altogether! (I
have a backup of the public data for this too, as does this repo!)
This happens on the Baby Kougra, where for most poses half of the
assets have a manifest that includes an SVG but no PNG. Skip 'em!
I considered adding a glitch tag for this, but idk I think we can do
that once we're aware of an actual case where this causes visible
issues.
On the small-screen layout, the popover goes down and covers the item
list, which isn't a big deal in context; so I want it to be basically
as tall as it can be without being unwieldy, to give more info.
But on the large-screen layout, it doesn't take long at all for the
popover to start intersecting the pet preview, because it *has* to go
up and cover the pet preview. So, I'm much more reserved about how much
vertical space I'm willing to give it!
(I also considered sending the popover off to the *right* of the button,
to cover the item list, but it felt *way* too weird imo! Especially in
the expressions case.)
The intent of this glitch message was that, when UC or Invisible pets
hide an item because of a zones-restrict thing, it would still show up
in the items panel as fitting a certain zone, whereas it should have
been in the "Incompatible" section and having none of its zones applied.
But the previously implementation would like, show this message even for
items that _were_ correctly marked as Incompatible? And that the server
returned no layers for, because it doesn't fit this body type to begin
with? (e.g. put a Grarrl hat on a Grarrl, then switch to Acara, and the
Grarrl hat is marked Incompatible—but would also show this confusing
message; or similarly with switching to Alt Styles)
So, when the server just returned no layers for this item to begin with,
don't show this message!
Now that we're tracking tab state ourselves, it's pretty easy to just
pass the `initialFocusRef` to the right place instead of to both!
This helps switching between the tabs feel a lot smoother, because we
don't have to re-render and fade-in all the poses again.
I'm keeping it in the same place for now rather than trying to fold it
into Styles, because I think that's net-less-confusing (since Styles
work pretty differently, e.g. different color requirements), and
certainly less work either way lol!
For Alt Style outfits, it's useful to call special attention to the Alt
Style feature as the likely cause of incompatibilities.
(Incompatibilities previously were most often caused by choosing a
species-specific item, then switching to another species. We generally
make it hard to enter this state, by hiding incompatible items during
search.)
Now that the pose picker button can have text content too, things were
getting pretty cramped horizontally on smaller screens! Now we use the
`sm` size when it's small-time.
I'm not planning to port full Alt Style support over to the 2020
frontend, I really am winding that down, but adding a couple lil API
parameters are *by far* the easiest way to get Alt Styles working in
the main app because of how it calls the 2020 API. So here we are, just
using new parameters for DTI 2020 that I'm gonna deploy to impress-2020
first!
Specifically, I was looking at the new "Stormy Cloud Kacheek" items,
and was surprised to find that, in the outfit editor, they all get
grouped under "Markings" (and therefore the UI treats them as
mutually-exclusive via hidden radio button and only bolds one at a
time), but they aren't actually conflicting because they occupy
different zones named "Markings".
In this change, we make the zone groups actually just be *by zone*
rather than jumbling all of the zones with the same label together; but
in most cases, we still keep the same simplified display. In the case
of the "Stormy Cloud Kacheek" items though, we now get a few groups:
`Glasses`, `Markings (#6)`, and `Markings (#16)`. Glasses is chosen
by coincidence because it's the first zone label for that item
alphabetically (even though that item also occupies a third "Markings"
zone), and then the other two know to disambiguate from each other.
There's an opportunity here to cheat things further, like to
*intentionally* select items like "Glasses" that are less ambiguous
when possible. I'm not aware of enough other cases like this for that
to really matter, though, so I'm just leaving it as-is!
I tested this a *bit* on other outfits, and everything looked fine at
a glance, so I'm just moving forward—but I'll make an announcement to
ask people to help take a look!
Two motivations here:
1. Unconverted pets should no longer exist on Neopets.com (and we
especially don't expect new ones), so this logic helps no one.
2. The Baby Pteri keeps getting overridden to be marked as Unconverted
because it has only one asset, which is incorrect.
A decent heuristic for a bygone era, goodbye!
This wasn't causing big problems because we made resilient little
choices in a lot of little places… but it's confusing as a potential
state, and the Styles chooser wasn't selecting the "Default" option
after you switch—which is what tipped us off!
Yay it works(*)! But two major missing pieces:
- Outfit saving doesn't persist it at all
- Item compatibility is unaffected: items will still appear in search
and in the preview, even when they don't fit anymore.
To help with space, I'm just showing the word "Nostalgic" (or "???" if
it's from a series we don't recognize, this is hardcoded by ID), and
trusting that from context it will be obvious that it's the "Nostalgic
Faerie" case or whatever. (Moreover, in both the button and the select
we're omitting the species name, by similar reasoning!)
Note that this _still_ doesn't actually apply the style to the outfit
whatsoever; this is all just local state as we're continuing to play
with UI concepts. Actually applying it is probably next though! (Though
there's a couple more UI things I want to do, like some affordances to
clarify that a Style is applied and that Expression changes won't work.)
Oh right, React Query's API is slightly different, fixed! (Previously,
this would cause the PosePicker to show before all the data was ready,
so alt styles would sometimes pop in after the popover was already
open.)
On small screens, the PosePicker opens down, and we put the tabs on the
top, to be near the button.
On large screens, the PosePicker opens up, and we put the tabs on the
bottom, to be near the button.
Previously, we always set `placement="bottom-end"`, which on small
screens behaved as written, and on large screens there would not be
space to open downward so it would open upward instead.
Now, we set the placement explicitly based on a media breakpoint, and
we change the `flexDirection` of the tabs container on the same media
breakpoint.
I'm playing with using text to call more attention to this button, but
I'm not altogether pleased with the design yet. I'll leave it there for
me and Support users, but hide it for most people until we've got a
more complete concept.
That way, I can stop being on a branch and be working on main, and
deploy stuff to preview live, without having to share it with everyone
just yet! (This was the motivation for finally adding Support tooling
to main DTI lol!)
A little architecture trick here! DTI 2020 authorizes support staff
requests by means of a secret token, instead of user account stuff. And
our support tools still all call DTI 2020 APIs.
So here, we bridge the gap: we copy DTI 2020's support secret to this
app's environment variables (I needed to update
`deploy/files/production.env` and run `bin/deploy:setup` for this!),
then users with the new `support_secret` flag have it added to their
HTML documents in the meta tags. Then, the JS reads the meta tag.
I also fixed an issue in the `deploy/setup.yml` playbook, where I had
temporarily commented some stuff out to skip steps one time, and forgot
to uncomment them after oops lol!
To activate this, I created a `.env.development` file in my project
root, with the following content:
```env
IMPRESS_2020_ORIGIN=http://localhost:4000
```
Then, I started impress-2020 with `yarn dev --port=4000`.
Now, the app loads from there, hooray!! It even fixes that obnoxious
pet state ID bug that happens when you run against the production db lol
Something in the Rails loader doesn't like that I have both a gem and
a lib folder named `RocketAMF`, I think? It'll often work for the first
pet load request, then on subsequent ones say `Envelope` is not
defined, I'm guessing because it scrapped the gem's module in favor of
mine?
Idk, let's just simplify all this by making our own module. I feel like
this old lib could use an overhaul and simplification anyway, but this
will do for now!
Like in 0dca538, this is preliminary work for being able to drop the
`zone_translations` table! We're copying the field over first, to be
able to migrate DTI 2020 safely before dropping anything.
Non-English languages haven't been supported on Neopets for a while, so
I'd like to remove this extra cross-cutting complexity, especially
since it's now inconsistent with the real site anyway!
The main motivation is that I'd like to do this for items too, because
I have a hunch that all the complexity of `globalize` to read
`item.name` is part of what's making large user lists so slow to
render: lots of little objects getting created down the stack, and
needing to be garbage-collected later.
I'm not sure that's why! But I figure removing this complexity is a
simplicity win anyway, so let's do it!
Note that this doesn't *finish* the migration, it just starts it! The
`Species::Translation` and `Color::Translation` models still exist, and
still have their data, and not all references to them are scrubbed yet.
I especially don't want to delete the backing tables until both DTI and
DTI 2020 are ready for it!
So this change will someday be paired with another change to actually
drop the tables - after backing up the data for future records just in
case, of course!
Using good ol'-fashioned cookies! The JS sets it, and then Rails reads
it on pageload. That way, there's no flash of content for it to load in
after JS loads.
If your first wanted list was created before your first owned list,
then `false` would come before `true` in the keys of
`current_user_lists`.
I both fixed this to be more consistent at the model level, because who
likes unpredictable behavior? But also downstream at the view I
hardcoded that true should come before false, because that's a UI
concern that I want to be encoded in the view regardless of what's
upstream.
Oh right, I just remembered how tabs work, they join with the page when
selected! lol
Something about this still feels off to me, but idk, I find it charming
and I wanna just let it rock.
They can just ellipsis, that's fine, the full name is not essential
when you're just scanning the list; you're gonna pop it open and see
the full name during contact anyway.
It was a bit tricky to figure out the right API for this, since I'm
looking ahead to the possibility of splitting these across multiple
pages with more detail, like we do in DTI 2020.
What I like about this API is that the caller gets to apply, or not
apply, whatever scopes they want to the underlying hanger set (like
`includes` or `order`), without violating the usual syntax by e.g.
passing it as a parameter to a method.
I made two mistakes here! One is that you need to round up subpixel
line heights, and the other is that `clientHeight` is what I want, I
shouldn't include the margin!
I guess I deleted this a while ago without really noticing… I think I'd
at some point like to replace this with like, the DTI 2020 improved
table layout thing, but I figured this would be pretty quick to throw
in and make the page not feel like a pain to use lmao
Oh yeah, a long-standing limitation. Good thing we're better at stuff
now!
This is also probably the real cause of the weird number of slight
discrepancies between main DTI and DTI 2020 when I eyeballed stuff lol
oh, well, that and the missing default-lists. A bit messy!
Ahh I see, the way we got away with not having a `trading` scope before
was a weird metaprogramming `{owned/wanted}_trading` situation. Okay,
let's trash that in favor of our new stuff! And that helps us bulk the
queries too which is nice.
In impress-2020, we do a big slow query to figure out which users have
been active in trades recently. Now, we cache that timestamp on the
User model.
This won't have any immediate effect; it's to clear the way for Classic
DTI to receive the better trade ratios feature people like from 2020.
I also added some unit testing infra because I finally wanted it! for
all the ways you can trigger this timestamp lol
Note too that this is a bit of an unusually complex migration, but my
hope is that the batching and query structure and such helps it run
surprisingly fast! 🤞
Poking at the site, I didn't love that clicking any of the NC items on
the homepage took a few seconds to load cuz of the OWLS request taking
a couple seconds.
If the request were faster to load I might decide otherwise, but for
now, let's just keep that cache long.
Another approach we could take would be to ask for an endpoint that
returns all the values at once, so we could cache and update them all
together, instead of having it all split into separate cache keys!
So this was a slightly wrong error message, what was happening was:
1. Trying to load the image hash for this pet, by looking them up at
https://pets.neopets.com/cpn/PET_NAME/1/1.png and seeing what URL it
redirects to.
2. But pets.neopets.com was rejecting our User-Agent string, which
would've been just "Ruby", since we hadn't set it otherwise. I guess
that's an explicitly banned string?
I also found that the kind of more-helpful User-Agent string I like to
write was being rejected, and I could only get it to accept something
very simple? So that's what we're using now, I guess!!
Building toward replacing more of the 2020 data sources! I think this is
an endpoint that benefits from bulk loading, esp with the way the item
page previews work. I also like taking the concept of "canonical" out of
the GQL interface, and instead just loading for each of the 50 species
and letting the client decide. (And then it can fast-swap between them!)
