>From the peanut gallery, in-line. > -----Original Message----- > From: Gilles [mailto:gil...@harfang.homelinux.org] > Sent: Monday, August 8, 2016 08:50 > To: dev@commons.apache.org > Subject: Re: [5/7] [math] Fix "FastMath#round(..)" to comply to changed > contract of "Math#round()" in Java 8 > > On Mon, 8 Aug 2016 16:40:04 +0200, Emmanuel Bourg wrote: > > Le 8/08/2016 à 16:22, Gilles a écrit : > > > >> There are pro and contra; IMO, saving a few characters is not worth > >> wondering upon reading whether "assertEquals" is from (JUnit) > >> "Assert" > >> or Commons Math "TestUtils". > > > > Seriously, there is little doubt that "assertEquals" comes from JUnit > > in > > a FooTest class. Everybody is used to that. [orcmid]
It is the presumption of tacit knowledge, in an ever-expanding ring, that complicates the on-ramp for newcomers, until they manage to gulp down all of the kool-aid and become perpetrators themselves. > > My point was that there exists "TestUtils.assertEquals": also for use > in a unit test class! > Someone might want to "import static" those too. In some files > "Assert" > would be implied, and in others "TestUtils". [No big deal: just what > you gain somewhere you loose somewhere else.] [orcmid] Except it is another dependency that someone has to be careful to check, and we're not talking about small test files here. > > I have no problem deciding that static import of Junit's Assert methods > is allowed. But let's be clear that it does extend to other classes. > > >> When people get accustomed to git, they won't see it as a > >> constraint, > >> rather as second nature. > > > > I'm accustomed to Git and I see this as a constraint that I don't > > have > > the time to deal with. > > Hence, you assume that reviewers are either not necessary, or that > they their time is worth less than yours. > > We can (have to, actually) understand that long-time committers > will take a short-cut sometimes, but whenever it happens it should > not be a massive commit. Otherwise anything goes. I hope that you > can agree that it should not be advertised as good practice. [orcmid] +1 with regard to paying attention to sustainability and community. > > > Regards, > Gilles > > > > > Emmanuel Bourg > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > For additional commands, e-mail: dev-h...@commons.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
