Previously, the whole walk tree was returned. This can be convenient
because you get the whole descendant tree all at once, but it's
unusual since all the other individual .json endpoints only return a
single object, not a list.
The cached_endpoint decorator was detecting that the response content
kept changing, so it never returned 304. Oops. At the moment the client
doesn't even use this key, so if we need it back we can use the etag or
another http header.
The word caching can be ambiguous, and what's worse is this file
previously contained a decorator for server-side caching of a
response and a class for client-side caching of files. It was
confusing. This new separation and naming should make it easier
to find what you're looking for.
Previous called get_data which is dangerous for large but
indeterminate response sizes, and the bail chain was more
difficult to reason about than a simple should_gzip true/false.
Previous version had a bug when the URL contained percent-encoded
spaces because url.replace() was looking for spaces and not replacing
the %20. Constructing the url from parts is more reliable.
With the recent addition of search_embed iframes on other pages, we
had photo cards appearing but the photo_clipboard module was not
imported thus the checkboxes did nothing. I don't want to import
photo_clipboard onto every single page, I'd rather they click through
to the full search UI. Otherwise every single page will have the tray
and often not a good enough reason for it.
So, since the functionality of the checkbox is completely reliant on
the photo_clipboard.js module anyway, there's no reason not to have it
generated by that module.
I was adding messages as strings because that's how they get shown on
the web interface. But it's better to return the real exception objects
and have the interface deal with it.
Move save_clipboard into a 0-timeout. Check checkboxes immediately
instead of relying on apply_check_all / update_pagestate to make a
redundant loop through everything.
In narrow mode, it's very possible to wind up with only 1 photo per row,
and it looks silly when they are left-justified. Centered looks a little
more natural to me in this case.
This prevents the specific tag from being included in the results in
the first place, and has the knock-on effect that the descendants'
lines won't all start with the parent tag's name.
Primarily affects the "unlink" / "delete" buttons on /tags.
Without the padding we can get an extra pt of font size, and these
buttons are somewhat of a readability struggle.
Instead of defining the function with an internal if that will always
pick the same path, let's use that if to instead define a
straightforward function that just does what we want.
Since meta.status would be undefined for an incomplete request,
I would have to check response.completed && response.meta.status,
which is too much burden. Let's set the status immediately, and
if a callback wants to do further diving we'll use completed then.
I want to reduce some complexity around here, part of which is that
launch imported entrypoint imported backend, all to do some proxy
wrapping which isn't necessary for the dev case anyway. Less
layers of wrapping and importing is good. Plus I think this naming
is more clear.
Previously, then the tags list was very long and dominated the scroll
height of the page, the #right and thus the photo would be floating
halfway down the page. By making it sticky, the photo always occupies
the correct position in the viewport no matter how long #left gets.
