Potential Accordion Bug

I have an accordion control in my flex app and I am
experiencing problems with a checkbox in one of the accordion
items.
I have written some actionscript to dynamically add items to
the accordion control. I forget the terminology, but I have a
'template' document which specifies the contents of each item in
the accordion. Each item has a label control and a checkbox
control. When I check a checkbox in one of the accordion items and
scroll down so the item is not in view, then scroll back up to the
item, the checkbox is displayed as unchecked! Has anyone come
across this behaviour before?
Thanks,
Alan

Post an example.
Tracy

Similar Messages

  • Lion - Changing Custom Website File Location - Potential Lion Bug

    I had an interesting problem while setting up and transferring my custom website from a Windows 2003 server to my new Mac Mini 2011 Lion server.  I was able to get the site transferred and working correctly.  Took a bit of manipulating /etc/apache2...* files but good to go.  Now for the interesting part.
    I decided to RAID 1 my two intertnal drives so I used SuperDuper (not affiliated with them ... use it if you want) and cloned my boot partition to a MyBook drive I had laying around, rebooted, Raided and cloned back.  All good there.  For some reason I had the bright idea to move my custom web site files to the MyBook.  Did that and all good.
    I later came to my senses and tried to move the custom site files back to the RAID 1 partition.  Could not do it.  Neither through the Server.app interface or via manual entry of /etc/apache2...* files.  The solution appears to be due to a potential bug in Lion and Lion Server.app.  I found that under directory /etc/apache2/extra the file httpd-vhosts.conf was not being updated by Server.app.  Once I changed the file location there all is well.
    Hope this helps any of you with the same problem.
    JKH

    Hello, James here.
    The reason why you can't find the "System Sounds" folder is because Apple have hidden it in a slightly different location in OS X Lion 10.7.x
    To get to the directory with the "Volume Mount" file you need to go to...
    Macintosh HD > System > Library > Components > CoreAudio (right-click and show package contents) > Contents > SharedSupport > System Sounds > system
    Hope this helped!

  • Potential concurrency bug in OC4J 10.0.3 WebServices TestPageProvider

    Hi,
    I was wondering whether anybody else was having problems with the OC4J 10.0.3 preview when placed under a light load of concurrent users.
    We have downloaded the JAX-RPC stateless Web service WebServices sample from http://www.oracle.com/technology/tech/java/oc4j/1003/how_to/how-to-ws.html and placed it under varying loads using LoadRunner, when we run a single user over 100 iterations there is no problem but when we invoke 3+ users against the webservice testpage we get random HTTP 500 Internal Server errors, the exact message is :
    =h= Received 2692 bytes from server
    <HTML><HEAD><TITLE>500 Internal Server Error</TITLE></HEAD><BODY><H1>500 Internal Server Error</H1><PRE>java.lang.ArrayIndexOutOfBoundsException
    at java.lang.System.arraycopy(Native Method)
    at oracle.security.jazn.util.SecureRandom.engineNextBytes(Unknown Source)
    at oracle.security.jazn.util.SecureRandom.nextBytes(Unknown Source)
    at oracle.j2ee.util.UUID.<init>(UUID.java:59)
    at com.evermind[Oracle Application Server Containers for J2EE 10g (10.0.3.0.0) - Developer Preview].server.http.EvermindHttpSession.encodeSession(EvermindHttpSession.java:484)
    at com.evermind[Oracle Application Server Containers for J2EE 10g (10.0.3.0.0) - Developer Preview].server.http.HttpApplication.createSession(HttpApplication.java:1586)
    at com.evermind[Oracle Application Server Containers for J2EE 10g (10.0.3.0.0) - Developer Preview].server.http.EvermindHttpServletRequest.getSession(EvermindHttpServletRequest.java:2441)
    at com.evermind[Oracle Application Server Containers for J2EE 10g (10.0.3.0.0) - Developer Preview].server.http.EvermindHttpServletRequest.getSession(EvermindHttpServletRequest.java:2341)
    at oracle.j2ee.ws.server.TestPageProvider.doInvokeService(TestPageProvider.java:621)
    at oracle.j2ee.ws.server.TestPageProvider.doPage(TestPageProvider.java:149)
    at oracle.j2ee.ws.server.JAXRPCServlet.doGet(JAXRPCServlet.java:585)
    at javax.servlet.http.HttpServlet.service(HttpServlet.java:743)
    at javax.servlet.http.HttpServlet.service(HttpServlet.java:856)
    at com.evermind[Oracle Application Server Containers for J2EE 10g (10.0.3.0.0) - Developer Preview].server.http.ServletRequestDispatcher.invoke(ServletRequestDispatcher.java:771)
    at com.evermind[Oracle Application Server Containers for J2EE 10g (10.0.3.0.0) - Developer Preview].server.http.ServletRequestDispatcher.forwardInternal(ServletRequestDispatcher.java:298)
    at com.evermind[Oracle Application Server Containers for J2EE 10g (10.0.3.0.0) - Developer Preview].server.http.HttpRequestHandler.processRequest(HttpRequestHandler.java:829)
    at com.evermind[Oracle Application Server Containers for J2EE 10g (10.0.3.0.0) - Developer Preview].server.http.HttpRequestHandler.run(HttpRequestHandler.java:270)
    at com.evermind[Oracle Application Server Containers for J2EE 10g (10.0.3.0.0) - Developer Preview].server.http.HttpRequestHandler.run(HttpRequestHandler.java:112)
    at com.evermind[Oracle Application Server Containers for J2EE 10g (10.0.3.0.0) - Developer Preview].util.ReleasableResourcePooledExecutor$MyWorker.run(ReleasableResourcePooledExecutor.java:291)
    at java.lang.Thread.run(Thread.java:534)
    It seems to us that there is a concurreny bug that is showing up when some kind of a session identifier is being assigned to each incoming request. Note that the webservice itself is marked as stateless and we are testing at the moment to prove that when the service is stateful the problem does not exist.
    I know that this is not the ideal way of invoking the webservice since it doesn't actually exercise the soap engine on submission but it may be a little bug worth looking into.
    Regards,
    pj.

    Hi,
    Are you trying to do Load testing with 10gApp Server? We have to do this too. Can you tell me if there is any documentation for this Load Testing for 10gAppserver . Or if yo have done can you tell me how to start doing this please. If you don't have a problem.
    Your help will be greatly appreciated.
    Thanks,
    Asha

  • Accordion Bug

    I have an application where the base state is a login page,
    on successful login it takes you to a state called 'home'. I have
    an accordion in a panel in the home state and it doesn't work. Is
    this a known issue or am I missing something?
    Getting the following error:
    TypeError: Error #1009: Cannot access a property or method of
    a null object reference.
    at mx.managers::HistoryManager$/save()
    at mx.containers::Accordion/::commitSelectedIndex()
    at
    mx.containers::Accordion/mx.containers:Accordion::commitProperties()
    at mx.core::UIComponent/validateProperties()
    at mx.managers::LayoutManager/::validateProperties()
    at mx.managers::LayoutManager/::doPhasedInstantiation()
    at mx.core::UIComponent/::callLaterDispatcher2()
    at mx.core::UIComponent/::callLaterDispatcher()
    Created a simple app to show this error:
    <?xml version="1.0" encoding="utf-8"?>
    <mx:Application xmlns:mx="
    http://www.adobe.com/2006/mxml"
    layout="absolute">
    <mx:states>
    <mx:State name="home">
    <mx:RemoveChild target="{hbox1}"/>
    <mx:AddChild position="lastChild">
    <mx:HBox x="0" y="0" width="100%" height="100%">
    <mx:Panel width="250" height="100%" layout="absolute">
    <mx:Accordion x="0" y="0" width="100%" height="100%">
    <mx:VBox label="test1" width="100%" height="100%">
    </mx:VBox>
    <mx:VBox label="test2" width="100%" height="100%">
    </mx:VBox>
    </mx:Accordion>
    </mx:Panel>
    <mx:Panel width="100%" height="100%"
    layout="absolute">
    </mx:Panel>
    </mx:HBox>
    </mx:AddChild>
    </mx:State>
    </mx:states>
    <mx:HBox x="0" y="0" width="100%" height="100%"
    id="hbox1">
    <mx:Button label="Button"
    click="currentState='home';"/>
    </mx:HBox>
    </mx:Application>

    Did you ever get this working? If you remove the surrounding
    containers from the accordian, it does work, so it is not related
    to different states. I am having the same problem, so if you find a
    solution will you post a reply here for all to see? I will do the
    same.
    Here is your sample app working:
    <?xml version="1.0" encoding="utf-8"?>
    <mx:Application xmlns:mx="
    http://www.adobe.com/2006/mxml"
    layout="absolute">
    <mx:states>
    <mx:State name="home">
    <mx:RemoveChild target="{hbox1}"/>
    <mx:AddChild position="lastChild">
    <mx:Accordion x="0" y="0" width="100%" height="100%">
    <mx:VBox label="test1" width="100%" height="100%">
    </mx:VBox>
    <mx:VBox label="test2" width="100%" height="100%">
    </mx:VBox>
    </mx:Accordion>
    </mx:AddChild>
    </mx:State>
    </mx:states>
    <mx:HBox x="0" y="0" width="100%" height="100%"
    id="hbox1">
    <mx:Button label="Button"
    click="currentState='home';"/>
    </mx:HBox>
    </mx:Application>

  • Mac OSX (10.9.5) IDCC2014 - potential caption bug - can anyone replicate this fault?

    Hello everybody.
    I have encountered a strange situation with live captions in InDesign. To demonstrate this issue, I have made a new indesign file, and the information for the live caption uses the Camera option from the metadata dropdown in the live caption setup.
    I placed an image that had been saved from the digital camera into InDesign and that displayed the correct name for the camera. All good so far.
    I then opened the file in photoshop, did no manipulations to it other than save it as a slightly different filename and save it as a JPG. When I placed this into indesign and applied the live caption to this image, the caption is replaced with <No data from link>.
    I then opened THAT file back in photoshop, did no manipulations to it other than to save it as a PSD. When I placed this into indesign and applied the live caption to this image, the caption displays the correct name of the camera.
    Finally, I then opened the photoshop file, once again made no manipulations to it other than to save it as a JPG with a slightly different name. Again, when placed into indesign and applied the live caption to the image, it was again replaced with <No data from link>.
    I have checked Bridge and the Camera information is the same in the Camera Data (Exif) portion of the metadata in Bridge. Similarly, I have gone into the links palette in InDesign and selected utilities/XMP file info, and all camera data is displaying the correct camera information, even though the caption is not displaying the camera information.
    Confused? Maybe this screenshot of what I have done will help.
    To me, it looks like the live caption feature is not importing XMP information correctly. Has anyone else encountered a similar situation? If not, is anyone able to replicate the fault that I have experienced?

    I feared as much. I've also tried saving as a tiff file, made no difference. I've submitted as a bug to Adobe - Feature Request/Bug Report Form , they must think I have nothing else to do at the moment.
    I should add for those reading this post that this doesn't mean all information is being disregarded by the live captions, it looks like just specific camera information. Filenames, descriptions... these come in fine.

  • Potential clientgen bug

    As far as I can tell the clientgen ant tool that generates the client
    side code for Weblogic's web services doesn't handle multiple services
    from a single ear file very well. For each service a type-mapping .xml
    file containing a reference to the detail objects is generated on the
    client. Unfortunately the file is always the same regardless of service.
    In fact what clientgen does is to always take the type-mapping from the
    first service in the ear file. I circumvented the problem by using the
    ant task autotype for each service and using its output as input to
    clientgen. This is the maven/ant script I used, that gave an incorrect
    result:
    <ant:clientgen
    ear="${clientgen.ear.file}"
    autotype="False"
    packageName="${clientgen.package}"
    useServerTypes="True"
    warName="${clientgen.war.name}"
    clientJar="${maven.build.dest}"
    keepGenerated="${wsgen.keepGenerated}"
    serviceName="${clientgen.service.name}">
    <ant:classpath>
    <ant:path refid="maven.dependency.classpath" />
    <ant:pathelement path="${clientgen.ear.file}" />
    </ant:classpath>
    </ant:clientgen>
    I am using WLS 8.1sp1. If the above script is incorrect please let
    me know.
    /Björn

    The following code doesn't compile with sun java
    compiler 1.5.0_03 but eclipse doesn't seem to have a
    problem with it.
    Thanks for reporting this issue. I have filed bug 6284158.

  • Adobe Content Viewer - Potential major bug

    I saw that the new Adobe Content Viewer was just approved by Apple, and updated the app on my iPad 3. When I opened the app and downloaded a folio (folio made using r23), none of my Vector MSO's showed up as vector, they're all still raster. I repeated the process on another iPad 3 and had the same results. It seems like the r24 Content Viewer won't show vector MSO's. Anyone else having the same problem?

    Here are three screenshots. 1 - normal viewing, 2 - first zoom, then I zoomed out, then 3 - second zoom

  • Time Machine BUG: Without warning, one or more of your disks stops being backed up...

    Hi all
    Just discovered an interesting and potentially dangerous bug in TM running 10.7.4 (like many others I've downgraded from 10.7.5 because of the Spotlight/TM problems noted in this forum).
    Here's the bug: You set up TM and specify certain items -- files, folders or volumes -- to be excluded in the "Options" pane of TM Preferences. If you exclude a volume (e.g. a cloned backup volume), that volume must of course be mounted in order for it to be added to the "Exclude" list. But if you subsequently umount the excluded volume, TM SIMPLY REPLACES IT WITH SOME OTHER MOUNTED VOLUME without warning. The result: unknown to you, one of your disks, that you think is being backed up, isn't. You can actually see this happen if you unmount the disk you want to exclude while your exclusions list is open.
    This bug is 100% reliable. Tried permissions repair, clear caches, verify backups, reset TM by deleting /Library/Preferences/com.apple.TimeMachine.plist and re-entering excluded items -- same result.
    Question: Have others found this bug? Has anyone found a way of adding a volume to the TM exclusions list so that it "sticks" regardless of whether or not the volume is mounted??
    The only workaround I can see: Never unmount any disks in your exclusions list -- if you do, something else will be excluded instead. If like me you want easy access to the disks you work with but not others (e.g. backup clones), one option is to make the excluded disk mounted-but-not-visible.
    Appreciate others comments before reporting to Apple in case there's something weird in my setup (don't think there is).
    Pondini: Checked your great site, couldn't see this noted - apologies if I missed it.
    Ta much.

    For Windows 7 I corrected the problem using the steps below.  Since Windows 8 is setup for touch screen you will need to find these controls. 
    Corrective Action.  Note that several of these steps require the computer to process the action and may take 20 minutes to complete, depending on the configuration. 
    DELETE INTERNET EXPLORER PERSONAL RECORDS
    (This step may not be necessary)
    Start Windows Operating System (OS) in Safe Mode (Restart windows using F8)
    Select “Start” button
    Select “Control Panel”
    Select “Internet Options”
    Select “Advanced” tab
    Select “Reset” tab for “Reset Internet Explorer settings”
    Select “Delete Personal Records”
    Select “Reset”
    DISABLE INTERNET EXPLORER
    If not in WINDOWS SAFE MODE start Windows OS in Safe Mode (Restart windows using F8) Select “Start” button
    Select “Control Panel”
    Select “Programs and Features”
    Select “Turn Windows Features On or Off” (Left pane)
    Uncheck IE# checkbox (# will be the IE version number)
    Select “OK”
    Reboot the computer
    RE-ENABLE INTERNET EXPLORER
    Windows OS will restart in normal mode, or select normal operation
    Select “Start” button
    Select “Control Panel”
    Select “Programs and Features”
    Select “Turn Windows Features On or Off” (Left pane)
    Check “IE# checkbox” (# will be the IE version number)
    Select “OK”
    Reboot the computer
    RESTORE YOUR DESKTOP
    Select “Start” button
    Select “Control Panel”
    Select “System & Security”
    Select “System”
    Select “System Protection”
    Select “System Restore”
    Select “Chose a different restore point”
    Select a restore point at least 2 weeks ago (Check “Show more restore points” checkbox, if necessary)

  • Oracle BUG... Snapshots in 8.1.7

    I accidently issued a RENAME on a number of snapshots in my DB. It created a TABLE for each of these objects and I can not drop these objects without getting told "ORA-12083 must use DROP MATERIALIZED VIEW to drop "OWNER.TABLE"".
    If I try to use drop MAT VIEW I recieve messaage telling me that object is not a snapshot.
    Think this is a potential Oracle Bug. How can I drop these tables?
    Thanks,
    Miller

    This is the Oracle Database Lite forum. Can you please repost your question on the general database forum: General Database Discussions

  • Bug Fix Request - To Admins

    Hey there Logic Pro forum Admins,
    I've got a small bug fix request for you. I'm using Logic Studio to lay down VO, and I ran across a potentially distressing bug. When you are recording in the arrange window and you click on the playhead, it erases your record take. It doesn't just remove it from the arrange window and place it in the bin, it erases the take completely.
    In my case, it was good that I caught it early on in a take. But imagine if I had clicked the on the playhead after a 5 - 10 minute take, or perhaps longer? We all know that time is money, particularly in VO sessions. The actor would have had to redo the take, and I would have had to pay the VO director and actor accordingly.
    To be entirely clear, it wasn't as if I had purposefully clicked on the playhead while recording. I was actually trying to rename a marker and accidentally hit the playhead instead. The effect this has had on my session is that I'm now deathly afraid to do anything (including cutting previously recorded regions or renaming them) while rolling. I'm sure you can appreciate my dilemma. Yes I could simply not click on the playhead, but you know how it is in the heat of a session. Things need to happen fast, and your clicking accuracy isn't what it normally is etc.
    Anyway, in future iterations of Logic Studio, it would be really really nice (not to mention heart attack preventing) if the user could disable this 'feature'. Please please please fix this!
    Kind regards,
    -Justin
    Message was edited by: Inteleky

    I'm not sure if what you're experiencing is a bug. I just tried this. While recording, I hit the playhead, and it simply stoped the recording where it was when I hit it, although the session continued to play. Try it yourself and watch the transport - the record button will go off, but the play will continue. When I did it just now, it DID NOT erase the recording that was in progress. Everything that had been recording until I hit the playhead was still there. This action makes sense to me. If the playhead gets touched and jumps to another part of the project, the action of recording SHOULD stop.
    If you're really concerned about this, I'd do some experimenting. Just sit down and start recording, and see what happens if you hit the transport, or if you do other actions you're wanting to do. Again, everything that had recorded until the moment I hit the playhead, was there. If your ENTIRE RECORDING gets deleted when you hit the playback, then something's wrong with your Logic. Are you on 8.0.2?
    Why are you feeling the need to edit while recording, anyway? Especially when you're recording important stuff??
    On a sidenote, you're not going to reach anyone from Apple on this forum. You need to go here:
    http://www.apple.com/feedback/logicpro.html
    Good Luck!

  • Bug in Project 2013 when turning off screenupdating in vb

    Below a short (3-line) piece of code that turns off screen updating, applies the "Gantt Chart" view (though any view or table produces the same effect), then turns screen updating on again.
    When you run the code, the result is all task names and gantt chart bars disappear - only the dependencies remain in the Gantt chart. making ScreenUpdating = True restores them again.
    Has anyone else seen this potentially major bug? It seems like turning screen updating on again after turning it off doesn't refresh the screen.
    Sub screen_updating_problem()
        ' Make the next line "True" to fix the problem, i.e. not update the screen
        ScreenUpdating = False
        ViewApply "Gantt Chart"
        ScreenUpdating = True
    End Sub

    Regardless of the need to turn screen updating off and on, clearly it's not working and it's a bug. This is my point.
    If I turn screen updating off, change the view, then turn it on again, regardless of whether this is needed or not, turning it back on removes the Gantt bars and text in the tables. This should not happen.
    Let me give you an example:
    1. Open all tasks (removing the filter and opening all summaries)
    2. Close the lower pane
    3. Go through a set of operations to determine the 1st, 2nd, 3rd, etc. critical paths from a selected milestone
    4. Add a column to the table
    5. Filter on the critical path to the selected milestone
    If I do this when screen updating is on (especially in step 3), the user can see each step of the process. We don't want this. We want to turn screen updating off so the next thing the users sees is the result, and non of the intermediate operations that
    involves changes to the screen. Turning screen updating off while it goes through each step makes each step also makes it run faster, because the screen doesn't have to be updated.
    What we are doing is relatively complex. Again, right or wrong, turning screen updating on after being off doesn't behave correctly.

  • SJSAS EE 8.1_02 / Bug in jss4.jar breaks our new release

    +[Full environment description below]+
    Problem description: our application server contains several applications including one J2EE application made of several WARs and EJBs. One of these EJBs starts some internal services that use an open source library called jsch [1] to make SFTP transfers (we use jsch 0.1.31). This jsch library relies on javax.crypto and negotiates encryption with the remote SFTP server. Doing so, it selects a cryptographic cipher. This works well from our development and build environments but fails on our test server which uses Sun App Server 8.1_02. This server doesn't deploy a JavaCAPS application internally (we have a separate App. Server for it).
    For some reasons (that I do not yet completely understand), our test (and production) Sun AS contains in the classpath a library (jss4.jar) and this library contains a cipher that is used for the transfer. The library may be related to Sun Application Manager or for some security setup but I am not yet sure. It may be standard with Sun AS EE.
    As shown in the appendices below, this library seems to contain a bug causing the SFTP connection to fail. If the code is indeed the one that is used in jss4.jar, then the library is obviously broken. But the only Sun patch I've seen that upgrades the library was http://sunsolve.sun.com/search/document.do?assetkey=1-21-119211-14-1&searchclause=jss4.jar . Tt doesn't seem to address our problem (at least in the patch documentation).
    I think that we now need some help/feedback from Sun. We have to either
    * find out if we can remove this library from the classpath
    * find a way to bypass the use of the cryptographic component
    * or get a fix for the library itself.
    Is it possible for someone at Sun to help us fix this issue ? We've to release again the application with a fix ASAP.
    Cheers,
    Jerome
    Appendix 1 ** Environment:
    SJSAS EE 8.1_02
    java -versionjava version "1.5.0_07"
    Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_07-b03)
    Java HotSpot(TM) Server VM (build 1.5.0_07-b03, mixed mode)
    uname -aSunOS ast-b2bdev 5.10 Generic_118822-27 sun4u sparc SUNW,Sun-Fire-V210
    ls -la /usr/share/lib/mps/secv1/jss4.jarlrwxrwxrwx   1 root     root          11 Dec 12  2005 /usr/share/lib/mps/secv1/jss4.jar -> ../jss4.jar
    ls -la /usr/share/lib/mps/jss4.jar -rw-r--r--   1 root     bin       657089 Jul 29  2005 /usr/share/lib/mps/jss4.jar
    md5 /usr/share/lib/mps/jss4.jar 1c9f9b09a1d632f31458caef5cfe4812        /usr/share/lib/mps/jss4.jarAppendix 2 ** stack trace that we see while running our program
    I've finally managed to get a full stack trace:
    [#|2007-11-05T18:23:00.980+0100|WARNING|sun-appserver-ee8.1_02|javax.enterprise.system.stream.err|_ThreadID=230 ;|
    javax.crypto.ShortBufferException: 32 needed, 20480 supplied
            at org.mozilla.jss.provider.javax.crypto.JSSCipherSpi.engineUpdate(JSSCipherSpi.java:338)
            at javax.crypto.Cipher.update(DashoA12275)
            at com.jcraft.jsch.jce.TripleDESCBC.update(TripleDESCBC.java:81)
            at com.jcraft.jsch.Session.encode(Session.java:800)
            at com.jcraft.jsch.Session._write(Session.java:1135)
            at com.jcraft.jsch.Session.write (Session.java:1132)
            at com.jcraft.jsch.UserAuth.start(UserAuth.java:52)
            at com.jcraft.jsch.UserAuthNone.start(UserAuthNone.java:36)
            at com.jcraft.jsch.Session.connect(Session.java:350)
            at com.jcraft.jsch.Session.connect(Session.java:145)
            at org.apache.commons.vfs.provider.sftp.SftpClientFactory.createConnection(SftpClientFactory.java:210)
            at org.apache.commons.vfs.provider.sftp.SftpFileProvider.doCreateFileSystem (SftpFileProvider.java:90)
            at org.apache.commons.vfs.provider.AbstractOriginatingFileProvider.findFile(AbstractOriginatingFileProvider.java:81)
            at org.apache.commons.vfs.provider.AbstractOriginatingFileProvider.findFile (AbstractOriginatingFileProvider.java:62)
            at org.apache.commons.vfs.impl.DefaultFileSystemManager.resolveFile(DefaultFileSystemManager.java:641)
            at org.apache.commons.vfs.impl.DefaultFileSystemManager.resolveFile (DefaultFileSystemManager.java:582)Appendix 3 ** potential matching bug in JSSCipherSpi as identified by jsch author
    ---------- Forwarded message ----------
    From: Atsuhiko Yamanaka <[email protected]>
    Date: Nov 6, 2007 2:12 AM
    Subject: Re: [JSch-users] javax.crypto.ShortBufferException in Session with 0.1.31 ?
    To: [email protected]
    Cc: [email protected]
    - Hide quoted text -
      +-From: "Jerome Lacoste" <[email protected]> --
      |_Date: Mon, 5 Nov 2007 18:39:50 +0100 _______________
      |
      |I've finally managed to get a full stack trace:
      |[#|2007-11-05T18:23:
      |00.980+0100|WARNING|sun-appserver-ee8.1_02|javax.enterprise.system.stream.err|_ThreadID=230
      |;|
      |javax.crypto.ShortBufferException: 32 needed, 20480 supplied
      |        at org.mozilla.jss.provider.javax.crypto.JSSCipherSpi.engineUpdate(
      |JSSCipherSpi.java:338)
      |        at javax.crypto.Cipher.update (DashoA12275)
    Sign, I knew.
    Before writing my last message, I had found the following buggy code[1],
       public int engineUpdate(byte[] input, int inputOffset, int inputLen,
           byte[] output, int outputOffset) throws ShortBufferException
           byte[] bytes = engineUpdate(input, inputOffset, inputLen);
           if( bytes.length < output.length-outputOffset ) {
               throw new ShortBufferException(bytes.length +  " needed, " +
                   (output.length-outputOffset) + " supplied");
           System.arraycopy(bytes, 0, output, outputOffset, bytes.length);
           return bytes.length;
    and I had strongly believed that the problem had come from it,
    but it is contradict to your success with your own compiled jar.
    [1] http://www.koders.com/java/fidF0D62D7385480B906B3CC88210B02AF4A84FF467.aspx

    +[Full environment description below]+
    Problem description: our application server contains several applications including one J2EE application made of several WARs and EJBs. One of these EJBs starts some internal services that use an open source library called jsch [1] to make SFTP transfers (we use jsch 0.1.31). This jsch library relies on javax.crypto and negotiates encryption with the remote SFTP server. Doing so, it selects a cryptographic cipher. This works well from our development and build environments but fails on our test server which uses Sun App Server 8.1_02. This server doesn't deploy a JavaCAPS application internally (we have a separate App. Server for it).
    For some reasons (that I do not yet completely understand), our test (and production) Sun AS contains in the classpath a library (jss4.jar) and this library contains a cipher that is used for the transfer. The library may be related to Sun Application Manager or for some security setup but I am not yet sure. It may be standard with Sun AS EE.
    As shown in the appendices below, this library seems to contain a bug causing the SFTP connection to fail. If the code is indeed the one that is used in jss4.jar, then the library is obviously broken. But the only Sun patch I've seen that upgrades the library was http://sunsolve.sun.com/search/document.do?assetkey=1-21-119211-14-1&searchclause=jss4.jar . Tt doesn't seem to address our problem (at least in the patch documentation).
    I think that we now need some help/feedback from Sun. We have to either
    * find out if we can remove this library from the classpath
    * find a way to bypass the use of the cryptographic component
    * or get a fix for the library itself.
    Is it possible for someone at Sun to help us fix this issue ? We've to release again the application with a fix ASAP.
    Cheers,
    Jerome
    Appendix 1 ** Environment:
    SJSAS EE 8.1_02
    java -versionjava version "1.5.0_07"
    Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_07-b03)
    Java HotSpot(TM) Server VM (build 1.5.0_07-b03, mixed mode)
    uname -aSunOS ast-b2bdev 5.10 Generic_118822-27 sun4u sparc SUNW,Sun-Fire-V210
    ls -la /usr/share/lib/mps/secv1/jss4.jarlrwxrwxrwx   1 root     root          11 Dec 12  2005 /usr/share/lib/mps/secv1/jss4.jar -> ../jss4.jar
    ls -la /usr/share/lib/mps/jss4.jar -rw-r--r--   1 root     bin       657089 Jul 29  2005 /usr/share/lib/mps/jss4.jar
    md5 /usr/share/lib/mps/jss4.jar 1c9f9b09a1d632f31458caef5cfe4812        /usr/share/lib/mps/jss4.jarAppendix 2 ** stack trace that we see while running our program
    I've finally managed to get a full stack trace:
    [#|2007-11-05T18:23:00.980+0100|WARNING|sun-appserver-ee8.1_02|javax.enterprise.system.stream.err|_ThreadID=230 ;|
    javax.crypto.ShortBufferException: 32 needed, 20480 supplied
            at org.mozilla.jss.provider.javax.crypto.JSSCipherSpi.engineUpdate(JSSCipherSpi.java:338)
            at javax.crypto.Cipher.update(DashoA12275)
            at com.jcraft.jsch.jce.TripleDESCBC.update(TripleDESCBC.java:81)
            at com.jcraft.jsch.Session.encode(Session.java:800)
            at com.jcraft.jsch.Session._write(Session.java:1135)
            at com.jcraft.jsch.Session.write (Session.java:1132)
            at com.jcraft.jsch.UserAuth.start(UserAuth.java:52)
            at com.jcraft.jsch.UserAuthNone.start(UserAuthNone.java:36)
            at com.jcraft.jsch.Session.connect(Session.java:350)
            at com.jcraft.jsch.Session.connect(Session.java:145)
            at org.apache.commons.vfs.provider.sftp.SftpClientFactory.createConnection(SftpClientFactory.java:210)
            at org.apache.commons.vfs.provider.sftp.SftpFileProvider.doCreateFileSystem (SftpFileProvider.java:90)
            at org.apache.commons.vfs.provider.AbstractOriginatingFileProvider.findFile(AbstractOriginatingFileProvider.java:81)
            at org.apache.commons.vfs.provider.AbstractOriginatingFileProvider.findFile (AbstractOriginatingFileProvider.java:62)
            at org.apache.commons.vfs.impl.DefaultFileSystemManager.resolveFile(DefaultFileSystemManager.java:641)
            at org.apache.commons.vfs.impl.DefaultFileSystemManager.resolveFile (DefaultFileSystemManager.java:582)Appendix 3 ** potential matching bug in JSSCipherSpi as identified by jsch author
    ---------- Forwarded message ----------
    From: Atsuhiko Yamanaka <[email protected]>
    Date: Nov 6, 2007 2:12 AM
    Subject: Re: [JSch-users] javax.crypto.ShortBufferException in Session with 0.1.31 ?
    To: [email protected]
    Cc: [email protected]
    - Hide quoted text -
      +-From: "Jerome Lacoste" <[email protected]> --
      |_Date: Mon, 5 Nov 2007 18:39:50 +0100 _______________
      |
      |I've finally managed to get a full stack trace:
      |[#|2007-11-05T18:23:
      |00.980+0100|WARNING|sun-appserver-ee8.1_02|javax.enterprise.system.stream.err|_ThreadID=230
      |;|
      |javax.crypto.ShortBufferException: 32 needed, 20480 supplied
      |        at org.mozilla.jss.provider.javax.crypto.JSSCipherSpi.engineUpdate(
      |JSSCipherSpi.java:338)
      |        at javax.crypto.Cipher.update (DashoA12275)
    Sign, I knew.
    Before writing my last message, I had found the following buggy code[1],
       public int engineUpdate(byte[] input, int inputOffset, int inputLen,
           byte[] output, int outputOffset) throws ShortBufferException
           byte[] bytes = engineUpdate(input, inputOffset, inputLen);
           if( bytes.length < output.length-outputOffset ) {
               throw new ShortBufferException(bytes.length +  " needed, " +
                   (output.length-outputOffset) + " supplied");
           System.arraycopy(bytes, 0, output, outputOffset, bytes.length);
           return bytes.length;
    and I had strongly believed that the problem had come from it,
    but it is contradict to your success with your own compiled jar.
    [1] http://www.koders.com/java/fidF0D62D7385480B906B3CC88210B02AF4A84FF467.aspx

  • Where can I submit a potential security issues for OSX10.8.4?

    This morning I found a potential security bug that allowed me to access files, applications and settings for my MBP before the OS had time to confirm my login credentials. I'm sure there are people looking to exploit this so I want to know if there is a place that I can submit the issue online or via telephone.

    http://www.apple.com/feedback/macbookpro.html

  • Dangerous iPhone headphone volume bug

    I've discovered a potentially dangerous bug with the volume in IOS7.1 on my iPhone 5S
    If I engage the volume limit (or the EU limit), this causes the volume to jump to max whenever I play a new track or skip a track in Music. This leads to very dramatic jumps in playback level.
    Anyone else experiencing this?

    Try killing and re-starting the Music app, or simply Reset your iPhone, or maybe a full Restore.

  • How do I use Sun Web Server 7.0u1 reverse proxy to change public URLs?

    Some of our installations use the Sun Web Server 7.0 (update 1, usually)
    for hosting some of the public resource and reverse-proxying other parts
    of the URI namespace from other backend servers (content, application
    and other types of servers).
    So far every type of backend server served a unique part of the namespace
    and there was no collision of names, and the backend resources were
    published in a one-to-one manner. That is, a backend resource like, say,
    http://appserver:8080/content/page.html would be published in the internet
    as http://www.publicsite.com/content/page.html
    I was recently asked to research whether we can rename some parts of
    the public URI namespace, to publish some or all resources as, say,
    http://www.publicsite.com/data/page.html while using the same backend
    resources.
    Another quest, possibly related in solution, was to make a tidy url for the
    first page the user opens of the site. That is, in the current solution when
    a visitor types the url "www.publicsite.com" in his or her browser, our web
    server returns an HTTP-302 redirect to the actual first page URL, so the
    browser sends a second request (and changes the URL in its location bar).
    One customer said that it is not "tidy". They don't want the URL to change
    right upon first rendering the page. They want the root page to be rendered
    instantly i the first HTTP request.
    So far I found that I can't solve these problems. I believe these problems
    share a solution because it relies on ability to control the actual URI strings
    requested by Sun Web Server from backend servers.
    Some details follow, now:
    It seems that the reverse proxy (Service fn="service-passthrough") takes
    only the $uri value which was originally requested by the browser. I didn't
    yet manage to override this value while processing a request, not even if
    I "restart" a request. Turning the error log up to "finest" I see that even
    when making the "service-passthrough" operation, the Sun Web Server
    still remembers that the request was for "/test" (in my test case below);
    it does indeed ask the backend server for an URI "/test" and that fails.
    [04/Mar/2009:21:45:34] finest (25095) www.publicsite.com: for host xx.xx.xx.83
    trying to GET /content/MainPage.html while trying to GET /test, func_exec reports:
    fn="service-passthrough" rewrite-host="true" rewrite-location="true"
    servers="http://10.16.2.127:8080" Directive="Service" DaemonPool="2b1348"
    returned 0 (REQ_PROCEED)My obj.conf file currently has simple clauses like this:
    # this causes /content/* to be taken from another (backend) server
    NameTrans fn="assign-name" from="/content" name="content-test" nostat="/content"
    # this causes requests to site root to be HTTP-redirected to a certain page URI
    <If $uri =~ '^/$'>
        NameTrans fn="redirect"
            url="http://www.publicsite.com/content/MainPage.html"
    </If>
    <Object name="content-test">
    ### This maps http://public/content/* to http://10.16.2.127:8080/content/*
    ### Somehow the desired solution should instead map http://public/data/* to http://10.16.2.127:8080/content/*
        Service fn="service-passthrough" rewrite-host="true" rewrite-location="true" servers="http://10.16.2.127:8080"
        Service fn="set-variable" set-srvhdrs="host=www.publicsite.com:80"
    </Object>
    I have also tried "restart"ing the request like this:
        NameTrans fn="restart" uri="/data"or desperately trying to set the new request uri like this:
        Service fn="set-variable"  uri="/magnoliaPublic/Main.html"Thanks for any ideas (including a statement whether this can be done at all
    in some version of Sun Web Server 7.0 or its opensourced siblings) ;)
    //Jim

    Some of our installations use the Sun Web Server 7.0 (update 1, usually)please plan on installing the latest service pack - 7.0 Update 4. these updates addresses potentially critical bug fixes.
    I was recently asked to research whether we can rename some parts of
    the public URI namespace, to publish some or all resources as, say,
    http://www.publicsite.com/data/page.html while using the same backend
    resources.> now, if all the resources are under say /data, then how will you know which pages need to be sent to which back end resources. i guess, you probably meant to check for /data/page.html should go to <back-end>/content/page.html
    yes, you could do something like
    - edit your corresponding obj.conf (<hostname>-obj.conf or obj.conf depending on your configuration)
    <Object name=¨default¨>
    <If $uri = ¨/page/¨>
    #move this nametrans SAF (for map directive - which is for reverse proxy within <if> clause)
    NameTrans.. fn=map
    </If
    </Object>
    and you could do https-<hostname>/bin/reconfig (dynamic reconfiguration) to check out if this is what you wanted. also, you might want to move config/server.xml <log-level> to finest and do your configuration . this way, you would get enough information on what is going on within your server logs.
    finally,when you are satisfied, you might have to run the following command to make your manual change into admin config repository.
    <install-root>/bin/wadm pull-config user=admin config=<hostname> <hostname>
    <install-root>/bin/wadm deploy-config --user=admin <hostname>
    you might want to check out this for more info on how you could use <if> else condition to handle your requirement.
    http://docs.sun.com/app/docs/doc/820-6599/gdaer?a=view
    finally, you might want to refer to this doc - which explains on ws7 request processing overview. this should provide you with some pointers as to what these different directives mean
    http://docs.sun.com/app/docs/doc/820-6599/gbysz?a=view
    >
    One customer said that it is not "tidy". They don't want the URL to change
    right upon first rendering the page. They want the root page to be rendered
    instantly i the first HTTP request.
    please check out the rewrite / restart SAF. this should help you.
    http://docs.sun.com/app/docs/doc/820-6599/gdada?a=view
    pl. understand that - like with more web servers - ordering of directives is very important within obj.conf. so, you might want to make sure that you verify the obj.conf directive ordering is what you want it to do..
    It seems that the reverse proxy (Service fn="service-passthrough") takes
    only the $uri value which was originally requested by the browser. I didn't
    yet manage to override this value while processing a request, not even if
    I "restart" a request. Turning the error log up to "finest" I see that even
    when making the "service-passthrough" operation, the Sun Web Server
    still remembers that the request was for "/test" (in my test case below);
    it does indeed ask the backend server for an URI "/test" and that fails.
    now, you are in the totally wrong direction. web server 7 includes a highly integrated reverse proxy solution compared to 6.1. unlike 6.1, you don´t have to download a separate plugin . however, you will need to manually migrate your 6.1 based reverse proxy settings into 7.0. please check out this blog link on how to set up a reverse proxy
    http://blogs.sun.com/amit/entry/setting_up_a_reverse_proxy
    feel free to post to us if you need any futher help
    you are probably better off - starting fresh
    - install ws7u4
    - use gui or CLI to create a reverse proxy and map one on one - say content
    http://docs.sun.com/app/docs/doc/820-6601/create-reverse-proxy-1?a=view
    if you don´t plan on using ws7 integrated web container (ability to process jsp/servlet), then you could disable java support as well. this should reduce your server memory footprint
    <install-root>/bin/wadm disable-java user=admin config=<hostname>
    <install-root>/bin/wadm create-reverse-proxy user=admin uri-prefix=/content server=<http://your back end server/ config=<hostname> --vs=<hostname>
    <install-root>/bin/wadm deploy-config --user=admin <hostname>
    now, you can check out the regular express processing and <if> syntax from our docs and try it out within <https-<hostname>/config/<hostname>-obj.conf> file and restart the server. pl. note that once you disable java, ws7 admin server creates <vs>-obj.conf and you need to edit this file and not default obj.conf for your changes to be read by server.
    >
    I have also tried "restart"ing the request like this:
    NameTrans fn="restart" uri="/data"
    ordering is very important here... you need to do this some thing like
    <Object name=default>
    <If not $restarted>
    NameTrans fn=restart uri from=/¨ uri=/foo.
    </If>

Maybe you are looking for