• Icon: Bug Bug
    • Resolution: Fixed
    • Icon: Major Major
    • core
    • jenkins.war v2.79

      favicon.ico in the root of the packaged war file (from http://mirrors.jenkins.io/war/) is being mangled since Jenkins v2.79.

      Compared to previous releases, the favicon file has doubled in size and now can no longer be read by browsers or by image editing software such as GIMP.

      The favicon in the source on github hasn't changed in 7 years, so I would guess this is as a result of some kind of change in war packaging, such as an encoding issue.

          [JENKINS-47127] Favicon in war broken since 2.79

          It's the maven-war-plugin, works with 3.0.0 and fails with 3.1.0, looking into it.

          Baptiste Mathus added a comment - It's the maven-war-plugin , works with 3.0.0 and fails with 3.1.0, looking into it.

          Baptiste Mathus added a comment - This caused the regression https://github.com/apache/maven-plugins/commit/4ebaa8bac7e20772a6b6438b273c48582c63550e#diff-acf964fa3df0af170507b414b9ea005f cc khmarbaise

          Baptiste Mathus added a comment - https://github.com/jenkinsci/jenkins/pull/3048 in review

          kboz added a comment - - edited

          I would suggest rather reverting maven-war-plugin to v3.0.0 until v3.2.0 lands (https://issues.apache.org/jira/projects/MWAR/versions/12341372) with MWAR-407

          kboz added a comment - - edited I would suggest rather reverting maven-war-plugin to v3.0.0 until v3.2.0 lands ( https://issues.apache.org/jira/projects/MWAR/versions/12341372 ) with MWAR-407

          Daniel Beck added a comment -

          I agree with kboz, nothing wrong with reverting part of a general cleanliness change if it's bad (and even an acknowledged upstream regression!).

          batmat FWIW in $previousJob when I performed similar changes I used binary identical file comparison, recursing into jars/war/zips, and then checking differences for more than the usual timestamps, random order of lists, etc. – a suggestion for the future.

          Daniel Beck added a comment - I agree with kboz , nothing wrong with reverting part of a general cleanliness change if it's bad (and even an acknowledged upstream regression!). batmat FWIW in $previousJob when I performed similar changes I used binary identical file comparison, recursing into jars/war/zips, and then checking differences for more than the usual timestamps, random order of lists, etc. – a suggestion for the future.

          kboz danielbeck agreed, I started receiving the notifications of MWAR-404 being reverted during the night just before going to sleep, and I thought that anyway we still have still to let it mature.

          Though the upcoming 3.2.0 release seems like it will contain a fix, I don't think this will be released before the end of the week though. So I'll still leave it some thoughts & hours, and I'll probably file a "revert to MWAR 3.0.0" PR later tonight or tomorrow for instance.

          If you think this must be expedited, just tell me so and I can file the PR more quickly.

          Baptiste Mathus added a comment - kboz danielbeck agreed, I started receiving the notifications of MWAR-404 being reverted during the night just before going to sleep, and I thought that anyway we still have still to let it mature. Though the upcoming 3.2.0 release seems like it will contain a fix, I don't think this will be released before the end of the week though. So I'll still leave it some thoughts & hours, and I'll probably file a "revert to MWAR 3.0.0" PR later tonight or tomorrow for instance. If you think this must be expedited, just tell me so and I can file the PR more quickly.

          Daniel Beck added a comment -

          We just need it done by the weekend.

          Daniel Beck added a comment - We just need it done by the weekend.

          Code changed in jenkins
          User: Baptiste Mathus
          Path:
          pom.xml
          http://jenkins-ci.org/commit/jenkins/d7dc9bc324bf0bc567c5b2a5b1ecfb71b99cf1e2
          Log:
          JENKINS-47127 Downgrade to m-war-p 3.0.0

          MWAR-404 introduced the regression in maven-war-plugin 3.1.0that makes
          every files be filtered, including binaries.

          MWAR-407 has been filed to revert MWAR-404, but that will be only in
          m-war-p 3.1.1 or 3.2.0 which are not yet released.

          So we downgrade to 3.0.0 for now.

          SCM/JIRA link daemon added a comment - Code changed in jenkins User: Baptiste Mathus Path: pom.xml http://jenkins-ci.org/commit/jenkins/d7dc9bc324bf0bc567c5b2a5b1ecfb71b99cf1e2 Log: JENKINS-47127 Downgrade to m-war-p 3.0.0 MWAR-404 introduced the regression in maven-war-plugin 3.1.0that makes every files be filtered, including binaries. MWAR-407 has been filed to revert MWAR-404, but that will be only in m-war-p 3.1.1 or 3.2.0 which are not yet released. So we downgrade to 3.0.0 for now.

          Code changed in jenkins
          User: Daniel Beck
          Path:
          pom.xml
          http://jenkins-ci.org/commit/jenkins/52342b26e51df5b989ab60ef77bd217b263df1ff
          Log:
          Merge pull request #3052 from batmat/JENKINS-47127-revert-to-m-war-p-3.0

          JENKINS-47127 Downgrade to m-war-p 3.0.0

          Compare: https://github.com/jenkinsci/jenkins/compare/34b8eb0628a7...52342b26e51d

          SCM/JIRA link daemon added a comment - Code changed in jenkins User: Daniel Beck Path: pom.xml http://jenkins-ci.org/commit/jenkins/52342b26e51df5b989ab60ef77bd217b263df1ff Log: Merge pull request #3052 from batmat/ JENKINS-47127 -revert-to-m-war-p-3.0 JENKINS-47127 Downgrade to m-war-p 3.0.0 Compare: https://github.com/jenkinsci/jenkins/compare/34b8eb0628a7...52342b26e51d

          Daniel Beck added a comment -

          Fixed in 2.82.

          Daniel Beck added a comment - Fixed in 2.82.

            batmat Baptiste Mathus
            kboz kboz
            Votes:
            0 Vote for this issue
            Watchers:
            5 Start watching this issue

              Created:
              Updated:
              Resolved: