| Thu, 31 Oct 2019 15:32:13 +0000 |
Chris Cannam |
Further replacements of QRect::right() with the right (arf) calculation
|
| Thu, 31 Oct 2019 15:32:01 +0000 |
Chris Cannam |
Use x + width() instead of QRect::right(), which doesn't return what we need and leaves us a pixel short; also subtract the resolution from the start frame to avoid scrolling scraggy single-pixel lines when we should be drawing nice fat ones
|
| Thu, 17 Oct 2019 14:44:22 +0100 |
Chris Cannam |
Permit getScaleProvidingLayerForUnit to return a dormant layer if there is no visible alternative. This is necessary to avoid the scale disappearing in Tony when the spectrogram is toggled off.
|
| Thu, 17 Oct 2019 11:12:54 +0100 |
Chris Cannam |
Pass a y-coord, optionally, to Layer::snapToFeatureFrame. This is necessary for BoxLayer which needs to coordinate its snaps with the box it is highlighting for editing. Then in BoxLayer, merge getPointToDrag and getLocalPoints into a single getLocalPoint and use this throughout.
|
| Thu, 17 Oct 2019 10:26:21 +0100 |
Chris Cannam |
Update getPointToDrag to prefer boxes that are containing the mouse position in height as well as width
|
| Wed, 16 Oct 2019 16:20:12 +0100 |
Chris Cannam |
Make audio dial more visible in Mac dark mode
|
| Wed, 16 Oct 2019 13:02:52 +0100 |
Chris Cannam |
Return true from getValueExtents always, just with no unit in the case where we don't have a nice neat scale. This should preserve the property of preventing other layers auto-aligning to us, while also ensuring we don't get overlooked for the purposes of drawing our own scale in a situation where a scale-less layer is on top of us
|
| Wed, 16 Oct 2019 13:01:23 +0100 |
Chris Cannam |
Debug output
|
| Wed, 16 Oct 2019 12:23:36 +0100 |
Chris Cannam |
Overhaul the way WaveformLayer reports value and display extents. The previous logic has been nonsense for a very long time, excused only by the fact that it was very seldom relevant (only really if another layer with unit V came along that wanted to auto-align with it). Partial fix for #1954 Peculiar alignment for Amplitude Follower y-scale in Auto-Align mode
|
| Wed, 16 Oct 2019 12:19:04 +0100 |
Chris Cannam |
Fix getScaleProvidingLayerForUnit to make it only return a layer that actually has display extents. Modify getVisibleExtentsForUnit to make it more like the behaviour in 3.x: where no layer with display extents is found, use the union of the value extents of layers with the right unit. Partial fix for #1954 Peculiar alignment for Amplitude Follower y-scale in Auto-Align mode.
|
| Wed, 16 Oct 2019 12:13:28 +0100 |
Chris Cannam |
Debug output tweaks
|
| Tue, 15 Oct 2019 14:53:25 +0100 |
Chris Cannam |
Overhaul choice of layer to show scale, in Pane::drawVerticalScale, making it call out to the same logic used by View::getVisibleExtentsForUnit when appropriate. Also refuse to accept scale from dormant top layer.
|
| Tue, 15 Oct 2019 14:51:58 +0100 |
Chris Cannam |
Avoid compiler warning
|
| Tue, 15 Oct 2019 11:40:56 +0100 |
Chris Cannam |
Clarify naming of some view-related methods. Rename LayerGeometryProvider::getValueExtents to getVisibleExtentsForUnit, and View::getTextLabelHeight to getTextLabelYCoord. Add View::getVisibleExtentsForAnyUnit to be used to determine which unit to adopt in a new e.g. box layer.
|
| Tue, 15 Oct 2019 09:32:24 +0100 |
Chris Cannam |
Fix #1951 Selecting row in data editor with multiple items having same frame always selects the first
|
| Mon, 14 Oct 2019 14:20:02 +0100 |
Chris Cannam |
Tidy / debug
|
