Import Snapshot in iOS

Richard Gaskin ambassador at fourthworld.com
Wed May 15 09:59:33 EDT 2013


Mark Wilcox wrote:

 > Yes, the code I looked at definitely took a rect and not an object.
 > I suspect the new method is not taking a snapshot of the screen at
 > all but rendering the object into an image. Wouldn't using the new
 > method and asking for the current card work for most scenarios?

I would imagine so.  In my understanding, the newer "from <object>" 
option does indeed work as you described, which is why I guessed it 
might be helpful in cases where the "from <rect>" option doesn't work.

I think Björnke's suggestion that this be reported as a bug is valid, 
since it's probably possible to change the buffer from which the "from 
<rect>" image data is obtained, from what I presume is the rendered card 
buffer to the final OS-rendered screen buffer instead.  This may be 
slightly more difficult on non-mobile systems, since the card buffer 
will be using different coordinates from the screen buffer, but even 
then those coords should be translatable.

In fact, this may help us understand why John's results are so different:

 > There is not a problem with importing a snapshot from a rectangle !..
 > and it doesn't matter is acceleratedRendering is true or false.. or
 > has even left the room ..:-)

I wonder if successful uses of "from <rect>" are using global 
coordinates, while unsuccessful ones are using local coordinates.

If this is true, I'd still consider it a bug but at least it becomes 
somewhat understandable, in that perhaps local coords pull from the card 
buffer (which is in tiles with acceleratedRendering turned on) while 
global coords perhaps pull from the final composited screen buffer.

That's just a wild guess, though.  But since John's seeing good results 
while others are seeing blank images, there must be some difference 
somewhere.

Perhaps we should note engine version numbers with results as this 
discussion continues?  It may be that ongoing enhancements to the 
rendering subsystem introduced a change in a relatively recent version 
which is the cause of this.

--
  Richard Gaskin
  Fourth World
  LiveCode training and consulting: http://www.fourthworld.com
  Webzine for LiveCode developers: http://www.LiveCodeJournal.com
  Follow me on Twitter:  http://twitter.com/FourthWorldSys





More information about the Use-livecode mailing list