[<<] [<] Page 1 of 1 [>] [>>] | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Subject:
nano-x and mwin
From: Chris Ross ####@####.#### Date: 3 Dec 1999 19:22:39 -0000 Message-Id: <Pine.LNX.4.10.9912031910210.15233-100000@vortex.ukshells.co.uk> i have been looking throught the source to both of these and i had an idea - tell me if it's a bad one. i was thinking that if we made the client libary for nanox have the same fuinction signitures as that of X we could then have our opne as nX.h - this wold mean easier proting of apps. and what would be the feasibilty of having one graphics server and having 2 client libaries - one that implements the Xlib replacement and one that is a Win32 replacment? this would mean the tree is far mor e logical in it's structure. chris ====================================================== | Chris Ross ( Boris` ) ####@####.#### | | http://www.darkrock.co.uk | `----------------------------------------------------' | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Subject:
RE: nano-x and mwin
From: Greg Haerr ####@####.#### Date: 3 Dec 1999 19:41:06 -0000 Message-Id: <796896539E6CD311B0E70060083DFEFB076FAE@NBA-SLAM.CenSoft.COM> having 2 client libaries - one that implements the : Xlib replacement and one that is a Win32 replacment? Chris - This is the idea I've had from the beginning. The whole original point of the Microwindows api was to allow applications to be run without learning a new api, or rewriting the application. With Nano-X, it's somewhat more complicated, since with the X model noone writes applications at the Xlib level, instead writes widgets, and there's no widget standard. However, there are several well known widget sets that would be nice to run on Nano-X. However, there is an additional complication to this original design that I just realized last week. That is, if Nano-X is exactly the Xlib api (good idea) then when people want Nano-X to run under X Windows, they get find symbol collisions with Nano-X and Xlib (nasty trick fixes required). At this point, I still think a viable compromise, since alot of folks on this list want the Nano-X on X11 option, would be to have Nano-X's api be "NXyyy()" where yyy matches the X11 code. However, when you remove the display parameter, in most cases, that's what we've got now, with Gryyy() calls.... So unmodified applications will be tricky with the X11 driver, unless that driver is moved to a shared library and the X11 symbols aren't exported. Greg | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
[<<] [<] Page 1 of 1 [>] [>>] |