There was an issue where the lines did not redraw their targets right away
when entering drag node, which could make it possible for a quick drag node
to try to connect to its parent line. With the chooseEdge exclusion it
would not connect to the parent nearby, but in another weird part of the line.
Some highlights
- `getSiblingAndChildVertices` are expensive, so they're saved and called less frequently
- draw touch targets for all the visible vertices
- remove redundant css classes and `setClass` function
This is more work to further isolate the layers that entities draw to.
It makes it easier to debug what is going on, and can eventually lead to
deferred drawing, if each draw function is in its own place and not dependant
on anything else.
I've started to replace the vertex-hover with an explicit layer for touch
targets.
Also had to change a lot of the svg tests, which are really brittle.
Things would happen like - the surface would be created, it would kick of a
deferred redraw, which would notice that the zoom was 0 and call
editOff, which would remove the osm layers that were just created and
that the tests were trying to draw to. These tests need proper zoom and
projection otherwise nothing works.
(re: #3003 / #4602)
For now, drawHover is commented out. Still not sure what I will do with it.
This means that things flicker a bit when dragging, also connecting nodes
(and closing lines) does not currently work.
There was lot going on preventing the vertices from rendering while dragging.
1. `modeDragNode` needed a proper `selectedIDs()` function that works like other
modes.. Many other places in iD (including the vertex renderer) call `context.selectedIDs()`..
This means that `modeDragNode` needs a new function `restoreSelectedIDs()` to do what
`selectedIDs()` was previously doing (a place to store selectedIDs so that we
can reselect those entities after the user is done dragging a node in select mode)
2. Just so many things in svg/vertices.js
- siblingAndChildVertices was missing some things for points that we render
as vertices (points in wireframe, points with directions)
- the sibling vertices weren't being included in the `filter` function, so
would disappear when doing differenced/extent redraws
- probably some other things
Previously it was up to the caller to draw whatever they want into the
footer flash. With this change, uiFlash creates an icon and a text, so
the caller doesn't need to do as much work.
(closes#4533)
There are several issues here.. Here's a quick brain dump:
- the requestIdleCallbacks in map.js (scheduleRedraw) seem to be causing
crashiness in Firefox on fast zoom/unzoom, mousewheel, etc.. anything where
the view transform changes a lot and the redraws don't catch up.
(commented out, reverted back to lodash throttle)
- the requestIdleCallback worker queue in idle_worker.js seems to be causing
crashiness in Firefox and Chrome when exiting the walkthrough. Something about
deferring the tile parsing as the user leaves the intro has a problem.
It might be an infinite `while` loop, not sure.
(commented out, reverted back to for loop)
- the requestIdleCallback in call_when_idle.js is only used to defer
tile loading in context.loadTiles() - this one seems fine, and actually
has maybe the biggest benefit for improving performance.
(left in)
1. All services are disabled in testing now to prevent network accesses
2. Only services are enabled when needed to test something
3. Many changes throughout code to allow iD to run with services disabled
(e.g. check for osm service instead of assuming context.connection() will work)
4. Actually export the services so we can disable and enable them
