Messages sent while mouse is down?

[Quentin Long]

sez Tim Selander <selander at tkf.att.ne.jp>:
> I miss the Zynga "Pathwords" game on Facebook, so for my own 
> amusement I'm trying to recreate it in Livecode.
> 
> For those who don't know the game, it had a solid screenful of 
> Scrabble-like lettered tiles. Click and drag the mouse through 
> adjacent letters to make words.
> 
> Without clicking, simply moving the mouse through the tiles 
> (fields) triggers mouseenter, mouseleave, etc. which makes it 
> easy to pick up the letters.
> 
> But when the mouse is down, it seems mouseloc() is the only thing 
> I can get. Using a variable what has all the field rectangles, I 
> can use the mouseloc() to ultimately identify the field under the 
> pointer, but it's too slow...
> 
> Does a moving mouse with the button down trigger any other 
> messages besides mouseloc()?
I see that Mike Bonner has already provided a solution which seems to do what you want. But just in case there are other people out there who might need a different solution, here's my stab at it…

If you have a "screenful of Scrabble-like lettered tiles", these "tiles" are presumably arranged in a rectangular grid, with neatly aligned rows and columns. If this is the case, the locations of the row-tiles are going to be separated by X number of pixels, such that row-tile 1 has X-coördinate A; row-tile 2 has X-coördinate (A + X); row-tile 3 has X-coordinate (A + 2*X); and so on.

Column-tiles will work similarly. Their locations will be separated by Y number of pixels, such that column-tile 1 has Y coördinate B; column-tile 2 has Y coördinate (B + Y); column-tile 3 has Y coördinate (B + 2*Y); and so on.

If the grid's horizontal spacing is identical to its vertical spacing, the separation-values X and Y will be the same, of course. Given the fact that pixels are not *necessarily* square, it would be imprudent to *assume* that the grid's horizontal and vertical separation-values are identical, and I will not make that assumption here.

So.

My solution to Tim Selander's problem completely ignores most of the mouse[whatever] messages, depending strictly on mouseLoc. Like so:

local dX = 25 -- if the horizontal-spacing value is not 25, put the real value here
local dY = 25 -- again, replace 25 with the real value as needed
local TimeSlice = 50 -- how often, in milliseconds, the code checks the mouseLoc. adjust as needed for response time
local GridLocPulse

global GridCell = "1,1"

on GridLoc
  if (GridLocPulse) then send GridLoc to me in TimeSlice milliseconds
  put the mouseLoc into ThisLoc
  put (1 + (item 1 of ThisLoc div dX)) into item 1 of GridCell -- may need tweaking to account for edge effects
  put (1 + (item 2 of ThisLoc div dY)) into item 2 of GridCell -- ditto
end GridLoc

on GridLocOn
  put true into GridLocPulse
  GridLoc
end GridLocOn

on GridLocOff
  put false into GridLocPulse
end GridLocOff

The above code can go into the script of the card where the tile-grid lives.

Once every (TimeSlice) milliseconds, this code looks at the mouseLoc and converts the mouse coördinates into grid coördinates, which are stored in the global variable GridCell. GridCell being a global, its contents should be accessible to any handler in any script which includes the line "global GridCell".

It's probably a good idea to *not* have the GridLoc handler burning clock-cycles *all the time*. Thus, the local variable GridLocPulse, and the subsidiary handlers GridLocOn and GridLocOff. GridLocOn activates the GridLoc handler, and GridLocOff turns GridLoc off.

Hope this helps…

   
"Bewitched" + "Charlie's Angels" - Charlie = "At Arm's Length"
    
Read the webcomic at [ http://www.atarmslength.net ]!
    
If you like "At Arm's Length", support it at [ http://www.patreon.com/DarkwingDude ].