[ https://issues.apache.org/jira/browse/SOLR-326?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12519146 ]
Hoss Man commented on SOLR-326: ------------------------------- > In general, I think it is discouraged to change code unless there is > something wrong with it the general mantra i like to follow is that commits should either change code, or fix formatting -- not both, because it's too hard to tell which is which. I personally don't think there is anything wrong with a commit where the sole purpose is to reduce warnings (either from the compiler, or from javadoc, or from bug finding tools) -- the only negative is that it sometimes makes older patches harder to apply, but in the case of unused imports, i don't think that's a big deal. one minor personal peeve i have is editors which sort the imports when removing the unused ones -- it makes the diffs a lot more confusing then ones with simple removals (also: i personally don't think alphabetized import lists are as useful as import lists where related classes in the same packages (or related packages in the same namespace) are listed next to each other -- but that's my own personal issue, and not an ideologue i would try and force on anyone else.) > cleanup eclipse warnings > ------------------------ > > Key: SOLR-326 > URL: https://issues.apache.org/jira/browse/SOLR-326 > Project: Solr > Issue Type: Improvement > Reporter: Paul Sundling > Assignee: Ryan McKinley > Priority: Minor > Attachments: remove_unused_imports_patch.txt > > > On default settings, Eclipse had 628 warnings. This patch removes 119 of > those warnings related to unused imports. These are the safest warnings to > fix and shouldn't require any testing other than confirming building still > works. > The general idea of removing warnings is both cleaner code, but also making > it easier for interesting warnings to get hidden by uninteresting warnings. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.