maven-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Tibo (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (MCHECKSTYLE-341) Introduce an expectedViolation flag
Date Fri, 29 Sep 2017 02:26:02 GMT

    [ https://issues.apache.org/jira/browse/MCHECKSTYLE-341?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16185269#comment-16185269
] 

Tibo commented on MCHECKSTYLE-341:
----------------------------------

I think maven is the right tool.... :)
It's more than discipline, sometimes you fix checkstyle issues without noticing. You want
to know about it ASAP, whatever can be enforce in the build tool should be enforce because
you want a fast feedback.
IMO SonarQube and such are nice tools to display metrics and enforce quality gates that can't
be easily done elsewhere but If I have something I want to enforce, I should enforce it wherever
I can and the closest to the developer, the best.
In this case having a "expectedViolations" flag will give more relevant feedback to our developers
for free (no perf impact, no negative consequences)
I am pretty sure that mostly everyone who is using "maxAllowedViolation" right now would rather
have an "expectedViolation" flag instead...

> Introduce an expectedViolation flag
> -----------------------------------
>
>                 Key: MCHECKSTYLE-341
>                 URL: https://issues.apache.org/jira/browse/MCHECKSTYLE-341
>             Project: Maven Checkstyle Plugin
>          Issue Type: New Feature
>          Components: checkstyle:check
>    Affects Versions: 2.17
>            Reporter: Tibo
>            Priority: Minor
>
> We are trying to fix our tech debt step by step using the maxAllowedViolation flag and
reducing the number slowly.
> We have 400+ maven module in our project and developer never update this flag, So basically
when someone is fixing checkstyle error without updating the flag, it leaves room for another
developer to introduce new errors...
> I would like an expectedViolation flag just to force people who are fixing issues to
also update the count... It could be called "expectedViolation". The difference with the maxAllowedViolation
flag is that this one would also fail when the number of actual violation is less than the
"expectedViolation" flag.
> I believe the maxAllowedViolation should have been an expectedViolation from the start.
I don't believe anyone wants to leave room for violation, you just want, for an existing project,
to explicitly specify the current number of violation and disallow through pull request the
number to go up.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Mime
View raw message