Uploaded image for project: 'JDK'
  1. JDK
  2. JDK-4425642

Regression: Bad conversion of bytes above 127....

XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Not an Issue
    • Icon: P3 P3
    • None
    • 1.3.0, 1.4.2_13
    • core-libs



      Name: boT120536 Date: 03/14/2001


      java version "1.3.1-beta"
      Java(TM) 2 Runtime Environment, Standard Edition (build 1.3.1-beta-b15)
      Java HotSpot(TM) Client VM (build 1.3.1beta-b15, mixed mode)

      No, its not the operating system character set, and no its not fixed in the
      1.3.1 beta. It is however, broken in java 1.3 on Solaris.
      This little program:

      public class DifferentAnswers {
          public static final char rs = 0xc8;
          public static void main(String argv[]) {
              System.out.println("a" + rs);
          }
      }

      When run on a solaris 8 machine, running Java 1.2.X and redirected to a file
      will create a file, with an a, and an E with an accent, ("a?"). This is what it
      should do. On that very same Solaris machine, Java 1.3 and Java 1.3.1 will
      produce a file with an a and a question mark. Every character in the extended
      ascii set gets converted to a question mark.
      If you build a web system, that accepts input from users, extended ascii, will
      be in your input stream. Web pages like this one for example....
      Or all of our JSP pages....

      If I write a jsp running under JRUN on a solaris 8 machine, that writes a
      web page that a user sees that includes extended ascii characters and it
      works while running Java 1.2.X, then I upgrade to java 1.3 or 1.3.1beta,
      and it starts spewing out question marks instead of the extended ascii
      characters it cannot be the machine, the font, the character set or
      anything other than the java implementation.

      I posted this as a bug, and got a reply that it was fixed in the next
      release. I downloaded 1.3.1 beta (the next release) and it is not fixed. I
      posted again, and got a bunch of junk about operating system character
      sets. I did not change operating systems or character sets, I just changed
      versions of jdk.

      In any case, if I look at the output stream and I do not see bytes with
      values above 127, and instead I see the question mark, I know something is
      very wrong. This is not a display issue, it is a morphing oir masking issue.


      (Review ID: 118757)
      ======================================================================

            ilittlesunw Ian Little (Inactive)
            bonealsunw Bret O'neal (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            0 Start watching this issue

              Created:
              Updated:
              Resolved:
              Imported:
              Indexed: