On Thu, Mar 21, 2013 at 11:10 AM, Tom Hacohen <tom.haco...@samsung.com>wrote:
> On 21/03/13 11:00, Mike McCormack wrote: > > On 03/21/2013 09:25 PM, Tom Hacohen wrote: > > > >> Silencing warnings just for silencing is not good. We use them to spot > >> bugs, that's why we like them so much. Silencing useful warnings is > >> counter-productive. > > > > Usually a developer, upon checking out a code base and building it on > > their system and observing a nice warning free compile will have a sense > > of confidence in a project. > > Yes. And we don't want to break that trust by creating false confidence > just by silencing the warnings, even ones that my point to bugs. > > > > Usually a developer, on making a modification to their code, checks for > > warnings and makes sure none have been introduced before pushing their > code. > > Yes, that's cedric's fault, no doubt, but still, just blindly silencing > these warnings will not help cedric fix the broken code. > > > > There are some people who wish to compile their code with every frickin > > warning in the world turned on, then proceed to ignore said warnings, > > and commit code with warnings to their revision control systems. > > I didn't quite get what you are talking about here. > > > > That gcc somehow changed which warnings it emits in a newer version > > changes none of the above. > > > > Enjoy your warnings. > > Dude, you know me, you know I'm pro-fixing warnings. However, shushing > important warnings that point out bugs just to silence the build is not > a good thing... I think you subconsciously linked this thread with the > one you had with demarchi. We are talking about completely different > things. > > Cedric just pointed out, and I agreed, that people shouldn't just shush > important warnings just for the sake of it. They should either fix them, > or make whoever introduced them fix them. That's it. > > -- > Tom. > pretty sure neither of the cases were "blindly silencing warnings". one of the changes made was to use the correct free function, preventing a magic failure and a leak. another was to add a format string and prevent undefined behavior with fprintf; the only issue here was that I added a trailing newline. since we're exaggerating here, I guess I had better delete my commit access since the newline broke build, prevented E from starting, and caused an orphanage to catch fire somewhere in Africa. starting a troll war on the mailing list without knowing what you're fighting about is not a productive use of anyone's time. ------------------------------------------------------------------------------ Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_d2d_mar _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
