It was a mistake to use original httpRequest as long as it is available because web containers may recycle it and populate with the fields from another request. So make sure that when prepareForDeferredProcessing() is called, we will be only using copied data.

Make AccessEvent.prepareForDeferredProcessing() to create a copy of the request attributes map so attributes are available later even if processing is done in a background thread.

Only Serializable attributes are copied to be on a safe side. Also, do not copy attributes set by Logback's TeeFilter as the very same information is already serialised as request/response content.

LOGBACK-929: daemonized threads created by logback executor service.

LOGBACK-929: fix spacing per Tony's request.

fixed typo in the paragraph relating to ExecutionServiceUtil changes (PR #187)

Fix for LOGBACK-894 cleanByPeriodOffset add time check

- Properly close <p> tags in order to be XML well-formed
- minor edits in the paragraph relating to ExecutionServiceUtil changes
  (PR #187)

Escape XML &,",' in XMLLayout (LOGBACK-728)

Fix LOGBACK-728 and LOGBACK-440.

Add unit test for XMLLayout: validate against log4j.dtd found in log4j
(added as a test dependency).

As log4j-over-slf4j is a test dependency as well we must take care to
have log4j-over-slf4j first in the classpath so that it can take
precedence over log4j.

fixed flaky mocking in AbstractSocketAppenderTest due to multi-threading issues

Use response encoding when creating the printwriter or you will run into...

 LOGBACK-1008 ConfigurationDelegate.groovy should check AppenderAttachabable.

Add support for depth range to CallerDataConverter

Add request method to MDCInsertingServletFilter

Make sure TeeHttpServletResponse uses the response character encoding
when contructing a new writer.

Fixed typo in append-toMySQL-with-datasource.xml

LOGBACK-116: RollingCalendarTest fails when user.timezone is not CEST

simplification of SocketApender connection logic

LOGBACK-909 Fix race condition in SMTPAppenderBase