Call HitTest in just one place
Can now preserve repeatedly hit UIHandle objects during pre-click motion
Fields of HitTestResult besides the handle pointer are now unused
The need to repaint a track during mouse movement can be indicated when
constructing a UIHandle or when updating it for move; HitPreview no longer
does this
And the last allows simplifications of LabelTrack glyph highlighting
Also move the temporary state for label glyph dragging out of LabelTrack
... Rather, construct them during hit tests (also capturing more state sooner
rather than at Click time, and adding some accessors for later use)
This also fixes bug 1677 by other means and avoids similar problems.
A cell may be implemented to re-use a previously hit handle object, not yet
clicked, in a later hit test, by remembering a weak pointer, but TrackPanel
holds the strong pointers that determine when the object is destroyed.
And the objects will surely be destroyed after drag-release, or ESC key.
For now they are also destroyed whenever not dragging, and hit-testing is
re-invoked; that will be changed later, so that the re-use mentioned above
becomes effective, but still they will be destroyed when the pointer moves
from one cell to another.
... except TrackSelectHandle. Cursor changes to a hand only after button-down.
HitTests give a UIHandle, whenever they also give any cursor or status, even
when it's unsafe to click and drag; Click override is reponsible for cancelling.
SelectHandle::Preview introduces much duplication, but the original in the hit
test will later be deleted.
... Let cell hit tests, and handle preview, know states only, not transitions.
Cell hit tests are passed a mouse state that does not always match the current,
but anticipates the button click to come; usually left, but if the Control
[sic] key on Mac is down, then right.
Thus, pressing and releasing Mac Control in multi-tool switches in and out of
the magnifier cursor.
I would prefer these strings be added to the po files and then extracted from there to make a .desktop file.
That way translators will only have to think about one file.
Problem: With no tracks selected, the commands cursor to track start/end set the cursor position to a silly high value or undefined respectively.
Simple fix: Set the flags on the commands to TracksSelectedFlag, as they were before the menus were reorganised.
Note that after this fix, with tracks present, but none selected, these commands cause a disallowed dialog to open, saying that there was a problem with the last action. I don't know if this behaviour is now expected.
After a focused track is removed by pressing shift+c, the new focus is often not correctly read by screen readers, especially nvda.
The fixes:
1. In AudacityProject::RemoveTrack, set the new focus after the track has been deleted, rather than before. (If it is set before, then the childId can be wrong after the track is deleted.
2. In TrackPanelAx::SetFocus, send an object focus event if there are no tracks. This is so the focus is correctly set when there are no tracks after a track has been deleted.
3. In TrackPanelAx::GetFocus, given the change in 2. , only call SetFocus if the focus has changed, to avoid sending unnecessary focus events. (Screen readers are normally tolerant of this, but Window Eyes became a bit too talkative.)
Enable EXPERIMENTAL_MIDI_OUT on all platforms
Add Portmidi dependencies (libasound2-dev, alsa-utils, alsa-oss) to travis
Regenerate configure scripts and Makefile.in's
Attempt to add the rest of the makefile for portmidi
Re-add automake files for portmidi
Fix missing include for allegro
These were created by rbdannenberg, in commit
932ca88255a322bff833e3eb4e3931a971aa97d4, but were accidently deleted
later in commit bb63fa0d07dad9a9f02b75eda455576e6e8b4da4 (which updated
portmidi). I can't say whether they work or not, and even when they were
introduced it was stated that they may be wrong. But it's better than not
having any automake files at all, and will at least provide a base for
them.
I didn't add some of the sourcefile changes from the first commit (some
were already added to portmidi, and others weren't). Some of them may be
needed to get building to work (or, we could update portmidi, if there are
any changes upstream). I also didn't add any of the libscorealign changes
other than one, since libscorealign already had makefiles (other than the
uninstalled one).
These configure scripts may be really old; regenerating them is probably a
good idea.
... an inline function in a header was declared to return bool, but returned
nothing. Now its type is void.
This function's undefined return value was not used anywhere.
If you hover exactly above the split line, you now get a pointing hand icon and status message, allowing you to start a selection at the split line more easily.
This allows selection and clicking on split lines to merge clips to act independently. It also means we don't need a portmanteau explanation in the status line, and the action to select up to a split line is simpler and easier to explain.
The expat site says versions like 2.1.0 with an odd number for minor version
should be considered "beta". This is instead "production grade".
This version promises to fix a possible "infinite loop DoS" attack, though
that is unlikely to happen with Audacity which does not download XML content.
