================================================================================
  Licensed to the Apache Software Foundation (ASF) under one or more
  contributor license agreements.  See the NOTICE file distributed with
  this work for additional information regarding copyright ownership.
  The ASF licenses this file to You under the Apache License, Version 2.0
  (the "License"); you may not use this file except in compliance with
  the License.  You may obtain a copy of the License at

      http://www.apache.org/licenses/LICENSE-2.0

  Unless required by applicable law or agreed to in writing, software
  distributed under the License is distributed on an "AS IS" BASIS,
  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  See the License for the specific language governing permissions and
  limitations under the License.
================================================================================

$Id: STATUS.txt 1056957 2011-01-09 15:00:54Z markt $

                         =================================
                         Apache Tomcat 6.0 Patch Proposals
                         =================================


RELEASE SHOWSTOPPERS:


PATCHES ACCEPTED TO BACKPORT:
  [ start all new proposals below, under PATCHES PROPOSED. ]

PATCHES PROPOSED TO BACKPORT:
  [ New proposals should be added at the end of the list ]

* Fix the maven stuff for the maven repo.
  Before it does't find tomcat-juli.jar and the remoteRepository seems broken .
  http://people.apache.org/~jfclere/patches/maven.patch.100711
  +1: jfclere
  -1: fhanik - easier to pass in the root path (lib/bin) to the macro instead of hacking around it
               if we remove the SCP auto feature, then there should be something to replace it with
               (http://ant.apache.org/manual/OptionalTasks/scp.html)
  +1: kkolinko: +1 for the updated patch (maven.patch.100711)
      I still think that it is not so good to remove the old code of <remoteRepository/>,
      because it ignores ${maven.repo.url} parameter provided by deploy-snapshot,
      deploy-staging and deploy-release targets. Maybe leave old code as a comment
      and fix it later if needed? I think that actually nobody besides the release manager
      uses this, so I am letting this pass.

* Expose the new WebappLoader flag in the VirtualWebappLoader,
  but allow alternative name searchVirtualFirst to make it
  consistent with the "virtual" terminology.
  Now you can decide, whether the virtual paths will
  be searched before the webapp or after it.
  If searched before, external resources take precendence
  over internal ones. Before that change one couldn't overwrite
  resources already present in the webapp.
  http://svn.apache.org/viewvc?view=revision&revision=936825
  http://people.apache.org/~rjung/patches/2010-05-14-loader-backport-r936825.patch
  +1: rjung, funkman
  -1:
  -0: kkolinko: The patch itself is OK, but I think having a synonym will
   cause confusion. I'd prefer not to invent a new name, but mention the
   one that we already have when documenting virtualClasspath.

* Fix path parameter handling. Currently the following URL fails with a 404:
  http://localhost:8080/examples/jsp/snp;x=y/snoop.jsp
  http://people.apache.org/~kkolinko/patches/2010-11-17_tc6_path-params.patch
  +1: kkolinko, markt
  -1:

   kkolinko: The old parseSessionId() method calls "clearRequestedSessionURL(request);"
     when there is no sessionid in the path. The new code does not do that. Was that
     call needed? It looks that it was not needed, but I might miss something.

   kkolinko: Discussed in Re:r1005192 thread on dev@


* Backport JSP unloading patch (BZ48358).
  The patch has substantially changed since the original version.
  Original revisions are: 937787, 1028377, 1028389, 1028396, 1028861, 1028862, 1028863,
  1028935, 1028939, 1028940, 1028944, 1028950, 1030014, 1030037
  Combined TC 6 patch:
  http://people.apache.org/~rjung/patches/BZ48358-JSP_Unloading-TC6_20101118.patch
  +1: rjung
  -0: markt - http://markmail.org/thread/777be426ulcfmdd3 suggests there may be
              a memory leak in this code somewhere. I'd like to get to the
              bottom of that before porting this
      rjung:  I started a discussion about JSPs and memory on the dev list.
              The updated patch fixes a race condition.
              We can stall this item until we get some feedback about 7.0.5.
  -1:

* Add session creation / expiration rate statistics to the session managers
  Adds ~10% to the session creation/destruction process when that is all Tomcat
  is doing. Should be less overhead with less contention.
  http://people.apache.org/~markt/patches/2010-11-18-session-rate-stats.patch
  +1: markt, rjung
  -1:
  rjung: getSessionExpireRate() and getSessionCreateRate() could share a common
         implementation (same method body expcept for the input list). The member
          "duration" of SessionTiming doesn not seem to get used.

* Backport AprEndpoint shutdown patch (BZ49795 and similar).
  http://people.apache.org/~mturk/tomcat/patches/tomcat-6.0.x-aprshutdown.patch
  +1: mturk
  -0: markt - Patch doesn't apply cleanly to tc6.0.x/trunk
  -1:

* Fix https://issues.apache.org/bugzilla/show_bug.cgi?id=48822
  Include context name in reload and stop log statements.
  http://people.apache.org/~slaurent/patches/2010-12-16_BZ48822_context_name_on_reload-tc6.patch
  +1: slaurent
  +0: kkolinko: Good idea, though logName() is
      "org.apache.catalina.core.ContainerBase.[engine].[host].[path]"
      which is not just a context name and looks ugly. The patch is not
      clean - it includes BZ 50459 bindThread patch that was proposed above.
  -0: markt: The patch is not clean (includes BZ50459)
             logName() is the wrong thing to be using. getName() should be
             sufficient
  -1:

* Fix https://issues.apache.org/bugzilla/show_bug.cgi?id=50550
  Requests for a newly created directory using MKCOL should not result in a 404
  http://svn.apache.org/viewvc?rev=1055975&view=rev
  +1: markt, funkman
  +1: fhanik - what does the boolean 'result' signify if the first unload() is false, and the second is true?
               markt - it means the resource the caller asked to be unloaded
                       wasn't found. The caller isn't told (to reduce possible
                       confusion) that a similarly named resource was found
                       and was unloaded.
  +1: kkolinko
  -1:
  Additional patch (not important):
  http://svn.apache.org/viewvc?rev=1056828&view=rev
  +1: kkolinko
  -1:
  
* Add helper class to allow a shared data source with different credentials
  http://svn.apache.org/viewvc?view=revision&revision=1055989
  +1: fhanik, funkman
  -1:
   0: kkolinko (have not tested it with TC6 yet)
  Additional patch:
  http://svn.apache.org/viewvc?rev=1056569&view=rev
  +1: kkolinko
  -1:

* Additional patch for bug 50413:
  Fix https://issues.apache.org/bugzilla/show_bug.cgi?id=50413#c6
  Serve the error page regardless of "Range" header in the original request.
  http://svn.apache.org/viewvc?rev=1056889&view=rev
  +1: kkolinko, markt
  -1:
