Sounds like we should leave the PR checks and testing as is: if it fails jshint, it "breaks the build".
What we can re-evaluate is our .jshintrc defaults: do we want to allow missing ; in our codebase? Seems like this is a very divided issue. I'm pro semicolons, but I hear thats old-school. I'm also pro-choice ;) so I'd be happy to remove that option. I'm also pro-consistency so I'd rather we use semicolons everywhere or nowhere. Sounds like I'm bikeshedding with myself! -Michal On Fri, Aug 22, 2014 at 2:42 PM, Jesse <[email protected]> wrote: > okay, that is what I did ... > I will leave things as is, I just wanted to judge if everyone felt that 1 > missing (though not required) ';' is enough to break the build. > > @purplecabbage > risingj.com > > > On Fri, Aug 22, 2014 at 9:40 AM, Lorin Beer <[email protected]> wrote: > > > +1 to fixing the test > > -1 to warnings > > > > warnings are either treated like errors, or ignored entirely. I'm for > > treating warnings like errors in the first place > > > > > > On Thu, Aug 21, 2014 at 1:53 PM, Marcel Kinard <[email protected]> > wrote: > > > > > I agree. My personal philosophy is generally to treat warnings as > errors. > > > > > > On Aug 21, 2014, at 9:27 AM, Mark Koudritsky <[email protected]> > wrote: > > > > > > > If some tests are too strict or too flaky, we can and should fix or > > > remove > > > > them altogether. I don't believe there is any value in warnings, they > > > will > > > > accumulate and then people will largely ignore them. > > > > > > > > >
