🚀 go-pugleaf

In article <BzKtBn.F0F.1@cs.cmu.edu>, jayg+@CS.CMU.EDU (Jay Gowdy) writes:
|> I have an ActiveHandler subclass that set's it's body as follows:
|> 
|> body(layout.fixed_span(kit.outset_frame(character_glyph
|>                        size, size));
|> 
|> The problem is, events such as enter and leave only take place
|> over the character glyph!  It seems that kit.outset_frame sets up a 
|> margin around the glyph it contains, and events are not reported in
|> the margins of glyphs for some odd reason.  Now it seems to me that
|> this behaviour would be hard to implement, therefore there _must_ be
|> some way to turn it off (yeah, I'm an optimist)!
|> 
|> By the way, for some other odd reason kit.bright_inset_frame does 
|> _not_ put a margin around it's glyphs, and the following works fine:
|> 
|> body(layout.fixed_span(kit.bright_inset_frame(
|>             layout.hbox(layout.hglue(),
|>                         character_glyph,
|>                         layout.hglue())), size, size));
|> 
|> I've noticed this behaviour in other glyphs that marginalize things,
|> but its only been annoying, not critical, as it is in this case.  
|> Please tell me 1) why this is happening 2) how I can stop it.

1)  The why is a bit obscure.  It has to do with the last two parameters
to the BevelFrame constructor and their affect on BevelFrame::pick.
I'm not sure I can defend the pick behavior of bright_inset_frame and
outset_frame being different, but that is why.  The reason the margin
usage is different is that bright_inset_frame tries to allocate the margin
from the body itself, as opposed to adding it.

2)  This is easier.  The Target class is very useful for controlling pick
behavior.  You can wrap a target around the body so that the entire
body will be picked.  The TargetPrimitiveHit tag is the most useful.