Just noticed that Classpath is five errors away from covering all of 1.2 except for swing.text, other than serialVersionUID problems in swing.
- method java.beans.Introspector.getBeanInfo(java.lang.Class, int): missing in classpath
- class javax.swing.plaf.metal.MetalSplitPaneDivider.MetalDividerLayout: public in jdk12, but protected in classpath
- constructor javax.swing.plaf.metal.MetalSplitPaneDivider.MetalDividerLayout(): public in jdk12, but protected in classpath
- field
javax.swing.undo.StateEdit.RCSID: constant [$Id: StateEdit.java,v
1.1.1.1 1998/10/28 06:43:54 sbb Exp $] in jdk12, but constant [] in
classpath
- field javax.swing.undo.StateEditable.RCSID:
constant [$Id: StateEditable.java,v 1.1.1.1 1998/10/28 06:43:54 sbb Exp
$] in jdk12, but constant [] in classpath
The Metal errors are a single one-word patch (s/protected/public/ on
the class, which will fix the constructor for free) and the RCSIDs are
the same ones I've submitted a patch for previously which Mark doesn't
want to accept (I, and also apparently others, still would like to see
it go in, though). I've no idea about the difficulty of the missing
Beans method.
In other spectacular Japi progress, Classpath has passed 97% of 1.4 and the generics branch is fast approaching 90% of 1.5!
Oh, btw, I do still intend to continue blogging about cmeScribble,
but I've been working on more of the mocking framework in nrdo to make
it easy and convenient to populate the mock tables with decent amounts
of data. I'm doing this in a way that will also solve a longstanding
feature request for cmScribe proper...