Look at http://planet.racket-lang.org/package-source/Dima_/Tray.plt/1/0/main.rkt
You can get GTK/win32/cocoa handle from any widget via get-handle ( http://docs.racket-lang.org/gui/window___.html#%28meth._%28%28%28lib._mred%2Fmain..rkt%29._window~3c~25~3e%29._get-handle%29%29 ) Wed, 10 Dec 2014 21:35:54 -0800 от Alexis King <lexi.lam...@gmail.com>: >The Racket GUI system is nice, but as with almost any GUI system, sometimes I >yearn for the simplicity of HTML rendering. CSS and JavaScript let >properly-styled HTML be easily inserted into almost any application without >anyone realizing it isn’t native, so long as it’s fast and seamless. > >Basically, I want to try and create a wrapper for a Chromium web view to be >able to add to a Racket GUI in the same way that I can add, say, a >horizontal-panel%. However, even putting the logistics of the complications >that would involve aside, the question becomes… is it even possible to >implement a new native UI component without sticking it into the internals of >MrEd? > >Currently, all of the plumbing that handles the linkage of the GUI classes >with their underlying native counterparts is tightly woven into the gui >library itself. There’s no way for an external package to hook into that >system and add something new, as far as I can tell. Ideally, I’d like to be >able to write a native library for each platform to define a GTK widget, a >Cocoa view, and a Windows… whatever they’re called in Windows. (I don’t do >Windows development.) Then I’d be able to write my own plumbing to connect >those native components with a Racket class that would provide the adequate >abstractions. > >I’m willing to work out those problems, but in order to do that, I do need to >figure out a suitable method for actually adding a GUI component. I’d love for >adding a web view to be as easy as typing (new web-view% [parent frame] [url " >http://google .com "]) and have it just work. > >Is this possible? Is it foreseeably possible? If not, what would be the best >way to integrate such functionality so that it’s at least semi-composable with >the existing libraries? > >(As a note, rendering to a canvas% or something like that is most certainly >out of the question. Chromium needs access to the window itself to get the >hardware-accelerated performance necessary to actually make it feasible.) > >Alexis >____________________ > Racket Users list: > http://lists.racket-lang.org/users > -- Roman Klochkov
____________________ Racket Users list: http://lists.racket-lang.org/users
