We have an "oldBinding" value that we use for checking whether the vertex attribute was already
bound to the current VAO, but we never set the state. Additionally, we're also checking whether
the previous state was already any binding (optional is set), and don't re-enable the vertex
attribute array. Additionally, we now only disable the vertex attribute array when the previous
state was in fact an array attribute. We still don't deduplicate constant glVertexAttrib* calls,
but that's a little trickier.
* Add simple unit tests for SymbolSizeBinder
* Fix bug in symbol size uniform value calculation
For camera functions we were setting the zoom levels in "covering ranges" to
`[(zoom stop <= tile zoom), (zoom stop >= 1 + tile zoom)]`, but then evaluating
the function at `[tile_zoom, tile_zoom + 1]`.
* Check for it != end() before accessing it->first
* [ios][macos] test remove source in use
* [android] test remove source in use
* [core] check source usage before remove
* [core] ensure layer::accept works with non-void return values on gcc
* [android] - remove upgrade runtime exceptions (#9191)
The isClose threshold is switched from 0.000001 pixels to 1 pixel.
As a backup, it checks whether r0 and r1 are finite. It might be
possible to have just the threshold check or just the finiteness check,
but I don't see the harm in having both.
std::abs(w0 - w1) < 0.000001 is removed because it doesn't look like
it's needed. All calculations should run fine even if w0 === w1.
Finally, the point interpolation is tweaked so that at the end of the
flying (when k === 1) it ends up at the exact end point. I didn't see
any bugs related to this, but it seems like a good thing to have
explicitly.
The isClose threshold is switched from 0.000001 pixels to 1 pixel.
As a backup, it checks whether r0 and r1 are finite. It might be
possible to have just the threshold check or just the finiteness check,
but I don't see the harm in having both.
std::abs(w0 - w1) < 0.000001 is removed because it doesn't look like
it's needed. All calculations should run fine even if w0 === w1.
Finally, the point interpolation is tweaked so that at the end of the
flying (when k === 1) it ends up at the exact end point. I didn't see
any bugs related to this, but it seems like a good thing to have
explicitly.
Introduce a second level of immutability, over each of the collections held by a style: sources, images, and layers. Tracking immutability at this level allows us to short-circuit significant portions of the RenderStyle update logic via a simple equality check, greatly improving performance.
