From: Gustaf N. <ne...@wu...> - 2016-12-06 08:42:13
|
Am 06.12.16 um 05:05 schrieb Colin McCormack: > > Am I being unreasonable in thinking that recent evidence of failures > in a piece of code does not itself make a very strong case for keeping > that code, but rather makes the opposite case? > it is as well my experience that a code, that nobody dares to touch (because it is complex/fragile/...) deserves be touched/redesigned first. > > I also question how sociable it is to seek to dissuade a widely > beneficial innovation > i am starting to wonder whether this is really the case (on all platforms) > > merely because it might disenfranchise one part (the Cocoa users) of > one part (the Mac users) of one part (the Tk users) of the collection > (I hesitate to use the word 'community') of Tcl users. > previously, the mac users had the option to build Tcl on Mac OS with and without Apple's core framework. The proposed code removed (among other things) all traces of CFRunLoop() from macosx/tclMacOSXNotify.c, by a developer seemingly having no possibility to test the consequences. CFRunLoop() was there for handling the interactions with the core framework. So far, I see more signals of of "ignorant surgery" than of "redesign". -g |