Closes https://github.com/ppy/osu/issues/30163.
If I'm to be blunt, the decoupled stuff in song select makes my head
spin. I spent a solid 20 minutes thinking how I was going to fix this
one but then finally realised that generally most of the cause there
was the fact that `AdvancedStats` was seeing the new rulesets *before*
the "ensure global selected mods are valid for current ruleset" logic,
and so decided to just _delay_ that until the decoupled transfer
thingamajig happens.
I was honestly considering combining `BeatmapInfo`, `Ruleset`, and
`Mods` into one property on `AdvancedStats`. I figured I'd rather not
push my luck and try the baseline version first, but I honestly think
that direction is going to be required at some point to properly corral
all of the decoupled madness taking place in song select.
Fixes https://osu.ppy.sh/community/forums/topics/1983327.
The cause of the bug is a bit convoluted, and stems from the fact that
the mod select overlay controls all of the game-global mod instances if
present. `ModSpeedHotkeyHandler` would store the last spotted instance
of a rate adjust mod - which in this case is a problem, because on
deselection of a mod, the mod select overlay resets its settings to
defaults:
a258059d43/osu.Game/Overlays/Mods/ModSelectOverlay.cs (L424-L425)
A way to defend against this is a clone, but this reveals another issue,
in that the existing code was *relying* on the reference to the mod
remaining the same in any other case, to read the latest valid settings
of the mod. This basically only mattered in the edge case wherein Double
Time would swap places with Half Time and vice versa (think [0.95,1.05]
range). Therefore, track mod settings too explicitly to ensure that the
stored clone is as up-to-date as possible.
Closes https://github.com/ppy/osu/issues/29738.
This "regressed" in https://github.com/ppy/osu/pull/29639, but if I
didn't know better, I'd go as far as saying that this looks like a .NET
bug, because the fact that PR broke it looks not sane.
The TL;DR on this is that before the pull in question, the offending
`.Contains()` check that this commit modifies was called on a
`List<BeatmapSetInfo>`. The pull changed the collection type to
`BeatmapSetInfo[]`. That said, the call is a LINQ call, so all good,
right?
Not really. First off, the default overload resolution order means that
the previous code would call `List<T>.Contains()`, and not
`Enumerable.Contains<T>()`. Then again, why would that matter? In both
cases `T` is still `BeatmapSetInfo`, right? Well... about that...
It is difficult to tell for sure what precisely is happening here,
because of what looks like runtime magic. The end *symptom* is that the
old code ended up calling `Array<BeatmapSetInfo>.IndexOf()`, and the new
code ends up calling... `Array<object>.IndexOf()`.
So while yes, `BeatmapSetInfo` implements `IEquatable` and
the expectation would be that `Equals<BeatmapSetInfo>()` should be
getting called, the type elision to `object` means that we're back to
reference equality semantics, because that's what
`EqualityComparer.Default<object>` is.
A five-minute github search across dotnet/runtime yields this:
c4792a228e/src/coreclr/vm/array.cpp (L984-L990)
Now again, if I didn't know better, I'd see that "OPTIMIZATION:"
comment, see what transpired in this scenario, and call that
optimisation invalid, because it changes semantics. But I *probably*
know that the dotnet team knows better and am probably just going to
take it for what it is, because blame on that code looks to be years
old and it's probably not a new behaviour. (I haven't tested empirically
if it is.)
Instead the fix is just to tell the `.Contains()` method to use the
correct comparer.
The explicit detach call was removed from `updateBeatmapSet`, causing
this to occur. We could optionally add it back (it will be a noop in all
cases though).
These used to work because there was a huge blocking load operation,
which is now more asynchronous.
Note that the change made in `SongSelect` is not required, but defensive
(feels it should have been doing this the whole time).
The only exception to the rule here was "when screen isn't active apply
without debounce" but I'm not sure we want this. It would cause a
stutter on returning to song select and I'm not even sure this is a
common scenario.
I'd rather remove it and see if someone finds an actual case where this
is an issue.
This reverts a portion of https://github.com/ppy/osu/pull/9539.
The rearrangement in `SongSelect` is required to get the initial filter
into `BeatmapCarousel` (and avoid the `FilterChanged` event firing,
causing a delayed/scheduled filter application).
