On 04/24/2010 02:53 PM, Gustavo Sverzut Barbieri wrote: > On Sat, Apr 24, 2010 at 9:30 AM, Christopher Michael > <cpmicha...@comcast.net> wrote: >> On 04/24/2010 11:57 AM, Mathieu Taillefumier wrote: >>> Hi, >>> >>> sorry for this big delay, but here it is. All modules uses now log >>> domains >>> for returning messages. I tested the patch for several weeks without any >>> problems and the diff was made starting from the latest version so I do not >>> expect that the patch will cause too much problems when/if applied. >>> >>> The changes are fairly simple. It consisted mostly to define an individual >>> log >>> domain for each modules that did not already included such method. I also >>> fixed >>> a memory leak in the msgbus module which defined a log domain at the wrong >>> place but did not release the resources when terminated. >>> >>> Finally, I did not touch to the illume modules since there is a comment in >>> the >>> code that says that eina_log slows since down on embedded devices. If this >>> comment does not hold anymore I will do the conversion otherwise I let them >>> in >>> the original form. >>> >> Thanks for not adding it to illume modules :) I don't use the eina_log >> functions there for several reasons. If I need to debug or output any >> info there, I just use simple printfs as the eina_log functions are just >> overkill for what I need :) > > eina log will have very minimal cost, particularly to the places we > will use it. One of the proofs of this is that eina, the core library > used everywhere, already uses it and NOBODY ever noticed a slowdown... > it is used extensively in stringshare, array and even list, yet nobody > measured any slowdown. > > Let's also remember that it is possible to compile out these messages > using the same technique in eina (maximum log level), just define a > value and compile with optimization, compilers will remove the block > and checking completely. > Well, all that may be true, but still....overkill for what I need :)
dh ------------------------------------------------------------------------------ _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel