This gives what is to me smoother and more predictable snapping motion with
respect to the current mouse position.

For example-- imagine the grid cell size is 10 and the selected object under
the mouse already has its top left corner aligned with vertical grid line #2.

If I begin dragging that object to the left, within a single pixel of motion
it will snap to vertical grid line #1 which is 9 pixels away from the
object's current position.

With this commit, the user must drag the object cellsize / 2 pixels before it
snaps to a new position.

To me this seems smoother as it rounds to the nearest grid line rather than
to the grid line with the smallest coordinate value.

We also keep a partially transparent version of a grid for editmode. This
allows us to keep editmode visually distinct from runmode, even if snap-to-grid
isn't turned on.

If needed it's pretty easy to make the following changes:

* set the "big" cell size to something that isn't 100 (but I'm not sure how
  much control over the grid we actually want to give users-- e.g., setting
  the big cell to something for which the small cell isn't a factor
* allow snap-to-grid to be set independently for each canvas. (But again,
  what would be the use case for this?

also run for extended periods of time.

For some reason, GUI operations which run for extended periods of time
may cause the engine to crash with a SIGHUP error on Linux. Apparently
this is caused by the watchdog bailing out and taking the engine with
it. This affects the new print operation as well as the existing message
operation in the file menu, possibly also search index generation for
the help browser, and maybe some other operations that still need to be
identified.

The new gui-busy method lets the GUI set an internal flag in the engine
which makes the engine keep the watchdog happy while the GUI does its
thing, using the same mechanism that is already in place to ping the
watchdog when running GUI-less. To these ends, the GUI needs to invoke
"pd gui-busy 1" when starting the time-consuming operation, and "pd
gui-busy 0" when the operation is done.

Massaged the routine a bit to bring it more in line with what vanilla
currently has, and backported some recent changes in rev. 1f383be3 and
rev. ff9e52c6 from vanilla which are supposed to fix some bugs in
sys_domicrosleep.

Sam Thursfield authored 4 years ago and Albert Gräf committed 4 years ago

This causes tar_em_up.sh to download an older version of nw.js suitable
for OSX 10.8.

It was accidentally removed by me in 4bffb36f.

Disable zexy/lpt on Windows for now, it doesn't work there at present and causes a segfault in the regression tests.