On Tue, Jul 22, 2014 at 8:21 PM, Andreas Kupries <[email protected]> wrote:
> Having heard about termbox only through this thread I just had a quick > look over the API ... > > The blit() function seems to be the core for windowing, copying a > cell-rectangle into the main termbox buffer. > > Have a display-list of windows = rectangles, and blit() them into the > main buffer, from back to front before presenting the result. > It currently looks like that's the approach i'll need to take. i'll first need a Window class where i can store the necessary bits for each window and keep track of child windows for GC purposes, but that won't be hard to do. > Missing features to help with this > - Can't blit() rect's out of the main buffer (to save state, before > blitting in a modal window, for example) > i (coincidentally) emails the author asking about the ability to save/restore blocks. No answer yet, but it was only an hour or so ago. Looking at the "godit" link from Gour, it seems it is indeed possible to implement multi-windowing, popup window support, etc., using only the available APIs. See the screenshots - it looks a lot like emacs and is implemented in "go": https://github.com/nsf/godit - Can't redirect the put()/change() APIs to work on rect's outside of > the main buffer. > That's where the Window class (fairly analog to a curses Window) would come in. It would need its own variants which translate the coordinates into the global ("stdscr") space. The fact that there's only one output routine (tb_change_cell(); tb_put_cell() is just a convenience wrapper around that one) should mean that higher-level code has to wrap less stuff. In curses there are many routines which have to be specialized for windows vs pads vs panel. BTW: termbox got the coordinates right: (x,y) instead of (y,x). Curses is dead. Long live curses. Still, something to keep track off. > Might be interesting to see how it meshes with "linenoise". > termbox provides all the low-level bits you'd need to implement linenoise-like editing, but it would be a bit of work to do, at least in C - might be easier in script code. Speaking of: it took me all of a couple hours to bind the 12-or-so functions to script space, and i'm quite certain you could implement a TCL module for it faster than that. My point is only - in the off chance that you want to experiment with it, doing so would not require a huge time sink. Will post screenshots if/when i've got something interesting to show. The core API binding is finished (plus a few extras, e.g. clearing the screen using given fg/bg attributes), but i'm not yet sure if i'm in the mood to hack on it further tonight. -- ----- stephan beal http://wanderinghorse.net/home/stephan/ http://gplus.to/sgbeal "Freedom is sloppy. But since tyranny's the only guaranteed byproduct of those who insist on a perfect world, freedom will have to do." -- Bigby Wolf
_______________________________________________ fossil-users mailing list [email protected] http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
