On 13/11/17 08:59, Mark Thomas wrote: > On 11/11/17 21:47, Konstantin Kolinko wrote: >> 2017-11-11 23:48 GMT+03:00 Mark Thomas <ma...@apache.org>: >>> On 27/10/17 14:13, Mark Thomas wrote: >>>> Now Checkstyle 8.3 has been released, I think we need to do something >>>> about these. There are quite a few options. >>>> >>>> Do we want the minimalist approach (run Rainer's script, tweak the >>>> Eclipse recommendations to get as close as we can to what Checkstyle >>>> wants) or do we take a broader 'What was Mark thinking when he suggested >>>> this import order?' approach and take the opportunity to rethink the >>>> order (and probably choose something simpler / more standard)? Or >>>> something else? >>>> >>>> Thoughts? >>>> >>>> I have no strong feelings either way but I'm happy to contribute to >>>> whatever option we choose. >>> >>> Ping? >>> >>> I'm not sure if the lack of response is because folks are ignoring the >>> Gump e-mails or because no-one has any strong opinions. >>> >>> In the absence of further feedback I plan to upgrade to Checkstyle 8.3, >>> fix the warnings and back-port. I hope to get to this early next week. >> >> Beware that Checkstyle 7.0+ requires Java 8 as runtime environment. >> Thus such upgrade cannot be easily backported to Tomcat 8.5 or earlier. >> >> I see the following options: >> >> 1) I think the easiest option is to >> - Disable these checks at Gump (except for Tomcat trunk) and enable >> them at Buildbot. >> >> Essentially, Gump bugs (regressions) in Checkstyle library and >> notifies us about them in advance. >> An issue is that we do not have resources to propagate these bugs to >> Checkstyle project. (Nobody stepped up to do that). >> >> Buildbot will find bugs in our own project, running the version of the >> library that we chose. > > Something along those lines makes sense to me. I should be able to start > work on that today.
Hmm. As I started to look at this I realised that a large number of the classes with errors are only using the static import a few times. Some of them are using a mix of static and non-static imports. Given that switching to the non-static usage also fixes the issue, I'm going to apply that fix first and then see what is left. Generally, I plan to apply it when switching to the non-static usage results in roughly the same amount of code or less. Mark > > Mark > > >> 2) I see that "Backport" section at Checkstyle home page [1] mentions >> some unofficial fork of Checkstyle that can be run on Java 1.6. I have >> not looked further. >> >> [1] http://checkstyle.sourceforge.net/#Backport >> >> >> 3) Another idea of mine is to run all Ant builds with Java 8, >> and explicitly configure paths to JDKs that are used for compilation. >> In Tomcat 7 we have "java.7.home". The idea is to add explicit >> "java.6.home" along that. >> http://markmail.org/message/6qck3d4zqb6nh5ja >> >> Best regards, >> Konstantin Kolinko >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org >> For additional commands, e-mail: dev-h...@tomcat.apache.org >> > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org > For additional commands, e-mail: dev-h...@tomcat.apache.org > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
