utf-x
  1. utf-x
  2. UTF_X-47

5 failures when running rt on windows xp

    Details

    • Type: Bug Bug
    • Status: Resolved
    • Priority: Critical Critical
    • Resolution: Fixed
    • Affects Version/s: 0.0.7 alpha
    • Fix Version/s: 0.0.8
    • Component/s: CanonicalForm
    • Labels:
      None
    • Environment:

      Operating System: Windows XP
      Platform: PC

    • Issuezilla Id:
      47

      Description

      jvm 1.5.0_06

        Activity

        Hide
        jacekrad added a comment -

        in fact there are only 4 failures. This is due to OS specific line ending.
        either these tests never worked on Windows (I find it hard to believe) or the
        implementation of java.io.PrintWriter has changed. In either case the
        CanonicalForm class doesn't correctly handle (translate) line endings on windows
        XP SP2
        F:\workspace\utf-x-0-0-x>java -version
        java version "1.5.0_06"
        Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_06-b05)
        Java HotSpot(TM) Client VM (build 1.5.0_06-b05, mixed mode, sharing)

        this is a serious issue and must be resolved before 0.0.8 is released

        Show
        jacekrad added a comment - in fact there are only 4 failures. This is due to OS specific line ending. either these tests never worked on Windows (I find it hard to believe) or the implementation of java.io.PrintWriter has changed. In either case the CanonicalForm class doesn't correctly handle (translate) line endings on windows XP SP2 F:\workspace\utf-x-0-0-x>java -version java version "1.5.0_06" Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_06-b05) Java HotSpot(TM) Client VM (build 1.5.0_06-b05, mixed mode, sharing) this is a serious issue and must be resolved before 0.0.8 is released
        Hide
        jacekrad added a comment -

        oops, imeant p2

        Show
        jacekrad added a comment - oops, imeant p2
        Hide
        lachdrache added a comment -

        I built the dist/$

        {dist.name}_src_full.zip file with the release target on Windows and copied it to Linux,
        unzipped it and ran the regression tests. Now the same 4 failures happen on Linux. When I build the ${dist.name}

        _src_full.zip on Mac OS X and copy it over to Linux everything is fine.

        I.e. one has to be careful on which platform to build the release.

        Show
        lachdrache added a comment - I built the dist/$ {dist.name}_src_full.zip file with the release target on Windows and copied it to Linux, unzipped it and ran the regression tests. Now the same 4 failures happen on Linux. When I build the ${dist.name} _src_full.zip on Mac OS X and copy it over to Linux everything is fine. I.e. one has to be careful on which platform to build the release.
        Hide
        lachdrache added a comment -

        fixed 3 of the 4 failures

        Show
        lachdrache added a comment - fixed 3 of the 4 failures
        Hide
        lachdrache added a comment -

        also fixed the last one

        Show
        lachdrache added a comment - also fixed the last one
        Hide
        jacekrad added a comment -

        Alex, just had a look at the the changes and at the first glance it seems that
        the original implementation was correct. The issue is with CVS.

        Show
        jacekrad added a comment - Alex, just had a look at the the changes and at the first glance it seems that the original implementation was correct. The issue is with CVS.
        Hide
        jacekrad added a comment -
            • Issue 55 has been marked as a duplicate of this issue. ***
        Show
        jacekrad added a comment - Issue 55 has been marked as a duplicate of this issue. ***
        Hide
        jacekrad added a comment -

        changed all test XML files to binary so that CVS doesn't change the EOL markers;
        reverted to the original (rev 1.2) implementation of CanonicalForm. Still 2
        failures

        Show
        jacekrad added a comment - changed all test XML files to binary so that CVS doesn't change the EOL markers; reverted to the original (rev 1.2) implementation of CanonicalForm. Still 2 failures
        Hide
        jacekrad added a comment -

        fixed CanonicalForm implementation by changing all println() statements to
        print('\n') which is inline with the W3C Canonical XML 1.0 recommendation.

        All tests pass on Linux and Windows. Distribution built on Windows and copied
        to Linux passes rt; distribution built on Linux and copied to Windows passes rt
        as well.

        Show
        jacekrad added a comment - fixed CanonicalForm implementation by changing all println() statements to print('\n') which is inline with the W3C Canonical XML 1.0 recommendation. All tests pass on Linux and Windows. Distribution built on Windows and copied to Linux passes rt; distribution built on Linux and copied to Windows passes rt as well.

          People

          • Assignee:
            jacekrad
            Reporter:
            jacekrad
          • Votes:
            0 Vote for this issue
            Watchers:
            0 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved: