[Checkins] SVN: z3c.form/trunk/CHANGES.txt Added change comment.

[Stephan Richter]

Log message for revision 76653:
  Added change comment.
  

Changed:
  U   z3c.form/trunk/CHANGES.txt

-=-
Modified: z3c.form/trunk/CHANGES.txt
===================================================================
--- z3c.form/trunk/CHANGES.txt	2007-06-13 04:48:54 UTC (rev 76652)
+++ z3c.form/trunk/CHANGES.txt	2007-06-13 05:01:54 UTC (rev 76653)
@@ -13,35 +13,40 @@
   selection and omission of prefixed fields and still specify the short
   name. Thanks to Nikolay Kim for the idea.
 
-- Feature: HTML element ids containing dots are not very good, because 
-  then the "element#id" CSS selector does not work and at least in 
-  Firefox the attribute selector ("element[attr=value]") does not work 
+- Feature: HTML element ids containing dots are not very good, because
+  then the "element#id" CSS selector does not work and at least in
+  Firefox the attribute selector ("element[attr=value]") does not work
   for the id either. Converted the codebase to use dashes in ids instead.
 
-- Bug/Feature: The IWidgets component is now an adapter of the form 
+- Bug/Feature: The IWidgets component is now an adapter of the form
   content and not the form context. This guarantees that vocabulary
   factories receive a context that is actually useful.
 
-- Bug: The mode selection made during the field layout creation 
-  was not honored and the widget manager always overrode the 
-  options providing its value. Now the mode specified in the 
+- Bug: The readonly flag within a field was never honored. When a field is
+  readonly, it is displayed in "display" mode now. This can be overridden by
+  the widget manager's "ignoreReadonly" flag, which is necessary for add
+  forms.
+
+- Bug: The mode selection made during the field layout creation
+  was not honored and the widget manager always overrode the
+  options providing its value. Now the mode specified in the
   field is more important than the one from the widget manager.
 
-- Bug: It sometimes happens that the sequence widget has the 
-Â  no-value token as one element. This caused ``displayValue()`` to 
-Â  fail, since it tried to find a term for it. For now we simply 
+- Bug: It sometimes happens that the sequence widget has the
+Â  no-value token as one element. This caused ``displayValue()`` to
+Â  fail, since it tried to find a term for it. For now we simply
 Â  ignore the no-value token.
 
-- Bug: Fixed the converter when the incoming value is an 
-  empty string. An empty string really means that we have no value  
+- Bug: Fixed the converter when the incoming value is an
+  empty string. An empty string really means that we have no value
   and it is thus missing, returning the missing value.
 
-- Bug: Fix a slightly incorrect implementation. It did not cause any harm in 
-  real-world forms, but made unit testing much harder, since an API 
+- Bug: Fix a slightly incorrect implementation. It did not cause any harm in
+  real-world forms, but made unit testing much harder, since an API
   expectation was not met correctly.
 
-- Bug: When required selections where not selected in radio and checkbox 
-  widgets, then the conversion did not behave correctly. This also 
+- Bug: When required selections where not selected in radio and checkbox
+  widgets, then the conversion did not behave correctly. This also
   revealed some issues with the converter code that have been fixed now.
 
 - Bug: When fields only had a vocabulary name, the choice terms adaptation