On Fri, Feb 18, 2011 at 12:47 PM, Nicolas Chauvat < [email protected]> wrote:
> On Thu, Feb 17, 2011 at 03:52:24PM -0800, Toshio Kuratomi wrote: > > > Right. The biggest reason I hear for not using pylint is how noisy it > is by > > > default and how hard to configure it to be useful it is. > > > > > I think there's two different enhancement requests here with different > merits: > > How useful would pylint be if it complained as much or less than one of > the > > other tools (pychecker, pyflakes, etc) and was hard to configure to be > more > > strict? Making pylint easier to configure would likely help no matter > which > > default existed. OTOH, I think that pylint's niche right now is that it > is > > comprehensive. > > Michael, Toshio, > > Would you be so kind as suggest ways to make pylint easier to > configure? > > Is the problem that it is comprehensive, hence has a lot of options, > hence is difficult to configure ? > > There are already message categories, but maybe "policies" could be > added to group options and enable/disable the whole group of messages > at once. Nicolas, I think it was sparked by my suggestion that if pylint were to start warning (legitimately) about local imports (in functions), it should be part of a group of extra strict warnings that are disabled by default (and turned on with a single option). I'm not seeing the need for introducing such functionality now though, as there aren't any checks I see as too ambiguous, in the way that checking for local imports would be. Arve Arve
_______________________________________________ Python-Projects mailing list [email protected] http://lists.logilab.org/mailman/listinfo/python-projects
