Hello, Andrey. > I've just found that I should waste my time fixing styles in user code that > (may be) reproduce some bug, just to validate the bug or fix provided by the > user.
Why do you ignore code style rules while developing «quick reproducer»? What specific code style rule is an issue for you? If we have some rules is just a waste of the time - may be it better to remove them? I’m ++1 to fail the build on code style errors. Code style errors == compile errors for me. > 8 июня 2020 г., в 16:40, Andrey Mashenkov <andrey.mashen...@gmail.com> > написал(а): > > Konstantin, > > +1 > > I've just found that I should waste my time fixing styles in user code that > (may be) reproduce some bug, just to validate the bug or fix provided by > the user. > I'm ok with the idea to block commits with style errors to master branch, > but not for other branches\PR. > > Can anybody explain why commiters should waste their time for this? > Why we even have such a rule to fail build on TC if there is some kind of > style error? > It looks (like a bullshit) counterintuitive as it is still possible to > commit to master with having style error, but impossible to just build a > project or to run a test. > > > On Thu, May 21, 2020 at 12:23 PM Konstantin Orlov <kor...@gridgain.com> > wrote: > >> Hi Ivan, >> >> Thanks for your reply! It’s better to get an answer late than never :) >> >> -- >> Regards, >> Konstantin Orlov >> >> >>> On 18 May 2020, at 09:04, Ivan Pavlukhin <vololo...@gmail.com> wrote: >>> >>> Hi Konstantin, >>> >>> Surprisingly, I found your message in a Spam folder (gmail). >>> >>> We had discussions about the subject before. The most recent one and >>> reflecting a current state is [1]. You can find many thoughts and >>> arguments in another discussion [2] (it might be better to start >>> reading from a bottom). >>> >>> [1] >> https://lists.apache.org/thread.html/6995a4e789117ba3f5577651866cfa99a6ffcc208cf60330d17d5a48%40%3Cdev.ignite.apache.org%3E >>> [2] >> http://apache-ignite-developers.2346864.n4.nabble.com/Code-inspection-td27709i80.html#a41297 >>> >>> 2020-04-20 11:00 GMT+03:00, Konstantin Orlov <kor...@gridgain.com>: >>>> Igniters, >>>> >>>> Currently we have code sanity checks [1][2] integrated within a build >> task >>>> [3]. Do we really need to fail the build (and therefore the other >> tasks) if >>>> there is a minor flaw like a missing line at the end of a file or an >> unused >>>> import? As for me it could be separated from the build task. >>>> >>>> What do you think? >>>> >>>> [1] >>>> >> https://ci.ignite.apache.org/buildConfiguration/IgniteTests24Java8_CheckCodeStyle >>>> < >> https://ci.ignite.apache.org/buildConfiguration/IgniteTests24Java8_CheckCodeStyle >>> >>>> [2] >>>> >> https://ci.ignite.apache.org/buildConfiguration/IgniteTests24Java8_LicensesHeaders >>>> < >> https://ci.ignite.apache.org/buildConfiguration/IgniteTests24Java8_LicensesHeaders >>> >>>> [3] >>>> >> https://ci.ignite.apache.org/buildConfiguration/IgniteTests24Java8_BuildApacheIgnite >>>> < >> https://ci.ignite.apache.org/buildConfiguration/IgniteTests24Java8_BuildApacheIgnite >>> >>>> >>>> -- >>>> Regards, >>>> Konstantin Orlov >>>> >>>> >>>> >>> >>> >>> -- >>> >>> Best regards, >>> Ivan Pavlukhin >> >> > > -- > Best regards, > Andrey V. Mashenkov
