Re: Application launch detection



On 10 Nov, Peter Astrand scribbled:
->  
->  > Yes, as do WindowMaker and Xalf, but the point of specifying these is
->  > to let toolkits support them natively and dump the LD_PRELOAD hacks.
->  > LD_PRELOAD is unportable, 
->  
->  It works on Linux and Solaris. For other platforms, it's possible to fall
->  back on using "mappingmode" (Xalf term), eg. just check for when an window
->  is mapped. 
->  
->  >slows down the linker
->  
->  True. 
->  
->  >collides with other
->  > LD_PRELOAD hacks, 
->  
->  Sometimes. Linux supports multiple libraries in LD_PRELOAD. This fact is
->  used in Xalf 0.4, and seems to work. 
->  
->  It's also possible to patch XLib directly (and maybe use LD_LIBRARY_PATH).
->  Most people agree on that if XLib could set the PID and hostname as atoms
->  on the windows, this would be a good thing. I've tested to patch Xlib
->  directly, so that XMapWindow sends an signal upon XMapWindow() and it
->  works great. The patch is about 10 lines. Apps could check if the Xlib
->  extension is available, and if not, fall back on LD_PRELOAD etc. 
->  
->  But: Are there *any* chance to get this extension into the official
->  XFree86/OMG source? Is it desirable?

i think getting it into free86 wouldn't be too hard - and that would
solve nigh everyone's problems :) i'll have a chat with mark today and
see if there wouldn't be any objections.

->  >and just generally sucks.
->  
->  :-)
->  

-- 
--------------- Codito, ergo sum - "I code, therefore I am" --------------------
The Rasterman (Carsten Haitzler)    raster rasterman com     raster valinux com
                                    raster enlightenment org raster linux com
				    raster zip com au





[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]