Is there a bug in env_open.c?

Similar Questions

• Clock on windows 8 is wrong time, I tried to set the time of the internet but its still not working, think theres a bug in the clock time, anyone have the same problem for windows 8 and know how to fix?

  Clock on windows 8 is wrong time, I tried to set the time of the internet but its still not working, think theres a bug in the clock time, anyone have the same problem for windows 8 and know how to fix?

  It sets obtained :-) Sorry for the delay

  Thank you very much! has worked perfectly!

• Is there a bug in the as3 tween class

  Is there a bug in the as3 tween class that causes interpolations is stopped before completing. I read that there is and that a 3rd party tween class should be used. I'm starting to learn as3 and experimentation, but it would be good to know when I move on more complex projects.

  The solution for this would also make your overall function. It simply means place your feature in the top level of your film.

  I did some research on this topic, but I found a lot of conflicting answers.

  Any help is welcome thanks in advance.

  interpolations should NOT be local to a function.  otherwise, they are likely to be gc would be before they finish.

• Is there a bug in shuffle the songs? iOS 10.0.1

  Love that I can remove (delete) the next song in shuffle mode, but the songs I delete seem to reappear after a song or two were played. The first person I delete is skipped, but suddenly are playing songs that I know that I removed.

  Is there a setting that I forgot, or is this one bug others have as well?

  

  I also have this problem.

  I get the following (to allow anyone to follow the bug):

  When the music plays.

  -Delete a song at random from the following list of top.

  -A this time the song removed disappears and won't play in principle.

  -When the current song ends and move on to the next, then the deleted song reappears in its original position.

  This happens with any combination: the random mode, non-aleatoire mode, playing a playlist or songs in the phone.

  So, there seems to be a clear bug in the new music app.

• Can't buy a single image. Is there a bug fix?

  Hello

  I was unable to buy a single image for a few days now. When I click on the button to buy the image, I get a prroger that says: "good news! '. We noticed that you already have a creative cloud subscription. Here are your upgrade options. "But there is no upgrade option. Then I tried to buy the 10 images per month option (even if I didn't it), I got the same message. I really need this image, the brochure I am working on needs to go to the printer at the end of the week. Do you know when this bug?

  Hi Timothy,.

  We are aware of the problem and have even reported to the engineering team. We do not have an ETA for the moment, but the solution is in the testing phase so will inform you when we have more information.

  ~ David

• Is there a bug in the Camera Raw 9.3.1.520 files corruptrices of the .dng since a Drone from Parrot Bebop?

  I am facing a very reduced problem. When you open a .dng file taken with my parrot Bebop Drone using 2015 Lightroom or Photoshop 2015 cc cc, the first time it opens fine. If you exit without saving the file, try to open the file, Adobe Camera Raw (ACR) will report that the file format is not compatible. It seems that simply open the file will cause Adobe to create a bunch of xmp metadata embedded in the file and write it in the file, even if it has been opened and not registered. Looks like a terribly bad practice for any software. I tried to find an arrangement in Adobe that could prevent this from happening without success. I filed a case with Adobe. Case # 187349196. I spent nearly an hour with a person of foreign support that kept insisting that ACR does not support drones Parrot when in fact it is listed as a supported in ACR 9.2 file type. Adobe would remove support in 9.3 +? I was finally transferred to a level 2 tech who has experienced the same problem and tabled the report. I expected better support for Creative cloud because I pay every month.

  Here is the temporary workaround...

  Before opening one of your Parrot records, change the permissions on the file read-only (you can do this using file information on a Mac, not sure about Windows). The file cannot be altered after this step. You can save the file to another format to avoid this problem, .psd works very well.

  This image shows the information file before and after opening with Photoshop. Note there is no 'save' on my part, I just opened up and quit smoking. The file corrupted on the left shows a bunch of metadata that are not in the original on the right file.

  I would like to hear from Adobe soon on this issue because it is very frustrating and time consuming.

  

  Hi lagunahils,

  Thank you for reporting the problem and it has been reported as a Bug.

  Kind regards

  Tanuj

• Is there a bug with handling binary_double columns?

  
  Hello.

  I use Apex 4.2.1 against Oracle 11 g 2 and mod_plsql.

  I tried to create a new page via the 'Form on a Table or view' option and using an Oracle table that contains more than one column of type BINARY_DOUBLE.  The new wizard, automatically highlighted all my columns in table to be included in the form and the form has been created without any errors.  However, when I went to edit the newly created form, I saw that all the columns of type BINARY_DOUBLE disappeared.  The other columns in the table (INTEGER, VARCHAR2, DATE, etc.) were all there.

  At first, I thought that, maybe, I'd made a mistake.  And so, I deleted my page and tried with the same exact results.

  I then added manually columns missing BINARY_DOUBLE, ensuring to designate 'source' as a 'database'.  Apex had no problem in creating these.

  Subsequently, I have tried insert rows in my table using this form.  I found that I could create new rows in my table with all fields filled, including the BINARY_DOUBLE type.  However, to my surprise, when I then tried to change the saved lines, values for all columns of type BINARY_DOUBLE appeared as NULL in the corresponding form fields.

  I use IE10.  Then I went to same IE8 and IE9.  I then went to Firfox 16.x.  All have the same behavior.

  Then, I changed my table, changing all BINARY_DOUBLE type NUMBER columns.  When you do that, everything worked as expected.  I recreated my form on the page and found * all * the columns present and accounted for.  And I was able to create and modify lines without problem.

  If anyone know what could cause this strange behavior?  I don't know that this is not a "function".  And so I wonder if this is a bug.

  Thank you.

  Elijah

  Hi Elijah,

  Thank you for reporting this issue. I filed bug #16939492.

  Kind regards

  Christian

• Is there a bug with the maintenance of interaction of the Glossary for the entire project?

  Greetings,

  I've seen in other posts (Cp6) that there is a known bug with the maintenance of interaction available in a project glossary.

  Maybe my problem can help further this question - or maybe someone can post tips on how to solve this problem.

  I use a trial version of the Cp7 and followed the advice in this tutorial: http://www.YouTube.com/watch?v=wTEF-56V1Z4

  I found that if I left 'Visible output' box unchecked for the button closure (as in the tutorial), I don't see the glossary at all.

  Similarly for the button I created for the glossary - I couldn't have hidden or I don't see it.

  Another problem: the close button in the glossary (which is grouped in the interaction) in a first time appeared with the glossary, but then decided to appear on following slides without it (but in the same location).

  Everyone has been able to follow the advice on the above tutorial link successfully and is there another tutorial I can watch or read about it? I really want to provide access to the glossary on multiple slides and during quizzes. (And Yes, I'll buy a license - although the number of bugs I encounter creeps upward!)

  Thank you

  Jacqui.

  Hello

  Unchecking 'visible output' means the Group glossary will be invisible in the output until it is made visible by an action. In this case, the action that make visible is assigned to the Glossary button.

  Make sure that:

  • Glossary button, button Exit and the Widget are defined "Rest of project" accordion Timing.
  • EXIT button is set to Place object on top in the accordion of Timing.

  Anthony

• Is there a bug in FM9 tank applying tags in the tables?

  When I apply my tag "BOLD" character of a word in a regular paragraph, what is happening, it's what I expect must happen--namely, weight = regular remains unchanged in the TFP designer and weight changes in "BOLD" in the tank Designer (and no asterisk override in down - read on).

  BUT, when I apply my "BOLD" word character tag inside a table, weight also passes "BOLD" in the TFP designer and an override asterisk appears at the bottom of the screen to the left of the TFP tag displayed.

  Is this a bug in FrameMaker, or is there something I'm not understand how FM is supposed to work messed up with my "BOLD" tag?  I think it's a bug, or at least an inconsistency - IMHO apply a tag to a word inside or outside a table should at least do the same behavior, shouldn't it?

  Thanks for the ideas!

  -Dave

  Dave,

  If the only content in the paragraph is the word that you apply the "BOLD" tag or if you manage to understand the end of the paragraph mark (or end of stream marks in the table cell), then FM applies this property as an alternative to the whole paratag (because that's what you've actually done).

  To avoid this, make sure that some content in the paragraph (even a space to the right) without the character format applied.

• is there a bug in the documentation?

  Hi guyes,
  
  I'm referring "Oracle® Data Guard Concepts and Administration 10 g Release 2 (10.2) B14239-05 August 2008 ' physical configuration of database the day before on my host even have my primary database.
  
  I told you it was my worst experience using oracle documentation.
  
  I use oracle 10g R2 and resounding throughout the procedure of physical configuration of the standby database, but I think that there are a few bugs in this document.
  
  1. they said to create controlfile to stadby database, but did not modify the controlfile to the new path of the data files.
  
  2. they ask to copy all the files from primary data instead of stdby, but when you try to open the database pending, it dosent match the path of the datafile because stdby same controlfile.
  
  These and there are many bugs are in this document.
  
  they ask to edit the tnsnames.ora file but dosent provide the procedure to do.
  
  I mean this isn't the guide to follow.
  
  one thing without hope
  
  is this a bug?
  
  Thanks and greetings
  VD

  Vikrant dixit says:
  is that I completely agree that a DBA should have knowledge about the functioning internal Oracle when the database was created.

  I am not convinced that service name configuration is internals knowledge - I think a developer to do so.

  but as a reader guess the oracle docs are the basis of their knowledge on any topic.

  We totally believe on these docs... rather that they don't even bother to write this change your path in controlfile for data files.

  If you read my previous answer, as if you read the documentation carefully, you will see that the most likely reason that they do not edit the controlfile to change the path of the data file that is with the advent of DB_FILE_NAME_CONVERT parameter database initialization isn't not required+.

  and according to my knowledge creating controlfile to standby database with the database of heavy production of transaction is not a small thing.

  Caz, it will have to close your primary database.

  Well you wouldn't create a watch from a base of high load on the same hardware that the original post asked, because this would imply the madness. It is possible for the create controlfile operation should be blocked pending a coherent picture of the controlfile, but it is almost always because the archivelogs pass too quickly, in which the solution to resize your logs and/or treat the excessive load on the db before considering the closure of your database to create a standby controlfile

  So it should be described here, if you provide procedure step by step, then u should at least have written this step.

  Once more it is not required anymore.

  I am not against the docs of the oracle, but even for beginners, they should improve, because we consider them as a guide perfect for us...

  We will, they are not perfect which is why you can log bugs against them - not unlike the software itself-, nor that they can meet all the needs, but in this case you seem to complain that some of the documentation is incorrect when it is not. Please can you show a standby database creation where you followed exactly the steps and still it did not work - you'll need to do to connect to a bug in all cases.

  I'll try and demo the contrary, if it might be a blog entry. If you really can demo it, positive the outweigh a negative case and I concede it gladly.

  Niall Litchfield
  Oracle DBA
  http://www.orawin.info/

  Published by: Niall Litchfield on November 24, 2008 19:02

• Is there a bug TestStand converting 4294967295 binary?

  When a binary variable has the value 4294967295 the binary value appears as 0b 10000000000000000000000000000000 and not 0b 11111111111111111111111111111111 as it should.  (4294967295 is 0xFFFFFFFF).  Octal and hexadecimal both work very well.  This can be tested using the NumericFormat SequenceEditor dialog box.  TestStand version 4.1.  Is there a solution?

  

  This problem has been fixed in 4.1.1.

  It is not really a workaround solution, even if the actual stored value is still correct. The digital format is only for display purposes. You can write your own code to display if you wish, but it would probably be easier to simply upgrade to the point 4.1.1 since you already use 4.1.

  This behavior was introduced due to a change in the compiler we use that caused casts from double to int Wrap is no longer and instead result in an overflow value (the value that you see). Essentially, the compiler now uses a new processor instruction that has this behavior. This can also happen in your own code of C/C++, depending on the compiler you use, every time that you start a double to int (TestStand numbers are stored as double) so it's something to know. The solution is first cast to an unsigned integer, we should be careful how do you this or the compiler will optimize out and it won't work in the Release version.

  Hope this helps,

  -Doug

• Is there a bug in CVI 2010 SetTableCellRangeVals and SetTableCellRangeAttribute for strings - particularly in 64-bit mode?

  It seems that I can't get SetTableCellRangeVals to publish string data to a table in 32-bit mode, I can't SetTableCellRangeAttribute to work for arrays of strings and now it seems numbers either. I just get garbage - it could be my code, however.

  Worse in 64-bit mode, I can't seem to post anything (for strings at least), but get a GP fault instead.

  I played with the size of the pointer in the array of strings (from 4 to 8 bytes

  The heart of the test code is here but the files are attached as well.

  int CVICALLBACK Test_callback (int, int int event, control panel,
  void * callbackData, int eventData1, int eventData2)
  {
  int i, numrows = 10, byattrib;
  int bit32 = 1;
  Double * dbl_array;
  char * string_array, * string;
  Rect a_rect;
   
  switch (event)
  {
  case EVENT_COMMIT:
  GetCtrlVal (main_handle, PANEL_BYATTRIB, & byattrib);
  String = malloc (50 * sizeof (char));
  If (bit32)
  string_array = malloc (4 * numrows * sizeof (char));
  on the other
  string_array = malloc (8 * numrows * sizeof (char));
  for (i = 0; i< numrows;="">
  string_array [i] = malloc (50 * sizeof (char));
  dbl_array = malloc (numrows * sizeof (double));
     
  for (i = 0; i< numrows;="">
  {
  dbl_array [i] = i * I * 3.1416.
  FMT (string, "The value is %f", dbl_array [i]);
  strcpy ([i] string_array, string);
  }
     
  a_rect. Top = 1;
  a_rect. Left = 1;
  a_rect. Height = numrows;
  a_rect. Width = 1;
     
  status = DeleteTableRows (main_handle, PANEL_TABLE, 1, -1);
     
  status = InsertTableRows (main_handle, PANEL_TABLE,-1, numrows, VAL_USE_MASTER_CELL_TYPE);
     
  If (byattrib)
  {
  status = SetTableCellRangeAttribute (main_handle, PANEL_TABLE, a_rect, ATTR_CTRL_VAL, dbl_array);
     
  a_rect. Left = 2;
  status = SetTableCellRangeAttribute (main_handle, PANEL_TABLE, a_rect, ATTR_CTRL_VAL, string_array);
  }
     
  on the other
  {
  status = SetTableCellRangeVals (main_handle, PANEL_TABLE, a_rect, dbl_array, VAL_ROW_MAJOR);
     
  a_rect. Left = 2;
  status = SetTableCellRangeVals (main_handle, PANEL_TABLE, a_rect, string_array, VAL_ROW_MAJOR);
  }
     
  free (string);
  for (i = 0; i< numrows;="">
  free (string_array [i]);
  free (string_array);
  free (dbl_array);
  break;
  }
  return 0;
  }

  Thanks for any help,

  Greg

  Greg-

  Certainly, you have found a bug.  However, I think that the issues that you run in a SetTableCellRangeAttribute are the result of a misunderstanding of the function.  This function, when you use it to set the value, sets the values of all the cells in the range to the same value.  Therefore, the var_args parameter must be the value that you want to set all cells to and not a pointer to an array of values.  In your code, you would need dereference the berries, and then, you would see that all the cells are the same value, which is the expected behavior.

  I created a bug report for the accident which follows from the definition of values on string cell types in 64 - bit with ID 284842.  This bug will be fixed in the next version of maintenace of the CVI.  In the meantime, you should be able to use SetTableCellVal to set the value of each cell individually in 64-bit.

  I'm sorry for the inconvenience-

  NickB

  National Instruments

• C++ API: is there a bug with Map::InvokeAll?

  Hello
  
  I try to use the input processors to do like 'put away cases', 'Delete where...
  
  I works very well except for return values. Everything that I do, the map returned by InvokeAll is always empty!
  
  This simple example:

  < coherence/net/cachefactory.hpp > #include
  #include < coherence/util/extractor/KeyExtractor.hpp >
  #include < coherence/util/filter/AlwaysFilter.hpp >
  #include < coherence/util/filter/LikeFilter.hpp >
  #include < coherence/util/filter/NotFilter.hpp >
  #include < coherence/util/filter/PresentFilter.hpp >
  #include < coherence/util/HashSet.hpp >
  #include < coherence/util/processor/ConditionalPutAll.hpp >
  #include < coherence/util/processor/ConditionalRemove.hpp >
  
  namespace ch = coherence::net;
  namespace cu = coherence::util;
  namespace cl = coherence::lang;
  
  namespace {}
  void dumpMap (cu::Set:View vSetResult, const std::string & message)
  {
  «std::CERR < < message < < ": «;»
  for (hIter cu::Iterator:Handle = vSetResult-> iterator(); hIter-> hasNext() ;))
  {
  CU::map:entry:view vEntry = cl::cast < cu::Map:Entry:View > (hIter-> next());
  CL::String:View key = cl::cast < cl::String:View > (vEntry-> getKey());
  CL::string:view val = cl::cast < cl::String:View > (vEntry-> getValue());
  ("std::CERR < <"("<< vEntry-> getKey() <<",) "< < GetValue vEntry - > < < '), '; '.
  }
  std::CERR < < std::endl;
  }
  
  void dumpCache (ch::NamedCache:Handle hCache, const std::string & message)
  {
  dumpMap (hCache-> entrySet(), message);
  }
  
  Sub initCache (ch::NamedCache:Handle hCache)
  {
  hCache-> clear();
  hCache-> put (cl::String:create("01"), cl::String::create("v:01"));)
  hCache-> put (cl::String:create("05"), cl::String::create("v:05"));)
  hCache-> put (cl::String:create("11"), cl::String::create("v:11"));)
  hCache-> put (cl::String:create("15"), cl::String::create("v:15"));)
  hCache-> put (cl::String:create("21"), cl::String::create("v:21"));)
  hCache-> put (cl::String:create("25"), cl::String::create("v:25"));)
  }
  }
  
  Sub testInvokeAll()
  {
  ch::NamedCache:handle hCache(ch::CacheFactory::getCache("first_cache"));)
  
  {
  std::CERR < < "== init" < < std::endl;
  initCache (hCache);
  dumpCache (hCache, "init");
  }
  {
  std::CERR < < "== emulation putIfAbsent" < < std::endl;
  Cu::map:handle = cu::HashMap::create() entries;
  entries-> put (cl::String:create("01"), cl::String::create("v:01MOD"));)
  entries-> put (cl::String:create("==="), cl::String::create("===MOD"));)
  entries-> put (cl::String:create("11"), cl::String::create("v:11MOD"));)
  entries-> put (cl::String:create("+++"), cl::String::create("+++MOD"));)
  CU::InvocableMap:EntryProcessor:handle processor = cu::processor:ConditionalPutAll: create)
  CU::filter:NotFilter:create (Cu::Filter:PresentFilter:GetInstance ())), entries;
  CU::map:view res = hCache-> invokeAll (cl::cast < cu::Collection:View > (entries-> keySet()), processor);
  BUG? must return a result set of 4 inputs but returns an empty set
  std::CERR < < ' size res ' < < res - > size() < < std::endl;
  dumpMap (res-> entrySet(), "res");
  The work has be done correctly
  dumpCache (hCache, "putIfAbsent");
  }
  {
  std::CERR < < "== emulation delete where key in ('11 ',' =') ' < < std::endl;
  Cu::set:handle cu::HashSet::create() = keys;
  keys-> add(cl::String::View("11"));)
  keys-> add(cl::String::View("==="));)
  processor Cu::InvocableMap:EntryProcessor:handle = cu::processor:ConditionalRemove:create (cu::filter:AlwaysFilter:getInstance (), true);
  CU::map:view res = hCache-> invokeAll (cl::cast < cu::Collection:View > (keys), processor);
  BUG? must return a result set of 2 inputs but returns an empty set
  std::CERR < < ' size res ' < < res - > size() < < std::endl;
  dumpMap (res-> entrySet(), "res");
  The work has be done correctly
  dumpCache (hCache, "Delete where");
  }
  {
  std::CERR < < "== emulation delete where key like 2% '" < < std::endl; "."
  should return tickets as the corresponding flag is set to true

  
  processor Cu::InvocableMap:EntryProcessor:handle = cu::processor:ConditionalRemove:create (cu::filter:AlwaysFilter:getInstance (), true);
  CU::filter:view f = cu::filter:LikeFilter:create (cu::extractor:KeyExtractor:create (), cl::String::create("2%"));)
  CU::map:view res = hCache-> invokeAll (f, processor);
  BUG? must return a result set of 2 inputs but returns an empty set
  std::CERR < < ' size res ' < < res - > size() < < std::endl;
  dumpMap (res-> entrySet(), "res");
  The work has be done correctly
  dumpCache (hCache, "remove where2" ');
  }
  
  }
  
  int main (int argc, char * argv)
  {
  Try
  {
  ch::CacheFactory::Configure(ch::CacheFactory::loadXmlFile("client-cache-config.xml"));
  testInvokeAll();
  ch::CacheFactory::Shutdown();
  }
  catch (const cl::Exception:View & e)
  {
  std::CERR < < "Ouch!" < < e < < std::endl;
  }
  return 0;
  }
  The output is:

  == init
  init: (11, v: 11), (v, 01:01), (25, v: 25), (v 05:05), (v, 21:21), (15, v: 15).
  is emulation putIfAbsent
  RES size 0
  RES:
  putIfAbsent: (=, = MOD), (11, v: 11), (v, 01:01), (25, v: 25), (+++, +++ MOD), (v 05:05), (v, 21:21), (15, v: 15).
  == emulation delete where key in ('11 ',' =')
  RES size 0
  RES:
  Delete where: (01, v: 01), (25, v: 25), (+++, +++ MOD), (v 05:05), (v, 21:21), (15, v: 15).
  == emulation delete where key like 2% '
  RES size 0
  RES:
  Remove where2: (01, v: 01), (+++, +++ MOD), (v 05:05), (15, v: 15),

  the cache is modified, as expected, but the cards returned by invokeAll are always empty!
  
  Is this a bug? I misunderstand documentation? Use the API incorrectly?

  Don't forget that when you are running code in the cluster - you actually run the Java code. That is to say. When you call / invokeAll, State of C++ processors gathered unmarshaled in the form of Java in the cache, and only then run.

  That is why - if the java ConditionalPutAll always returns the Null map (which does)... then your version of C++ will most probably also. (The return of a null map is for efficiency purposes)

  Once you recognize that and really really want to return values in a PutAll call - you can create / derive your own implementation of the Java ConditionalPutAll processor.

  The ConditionalRemove also works as advertised - but you should very carefully read the fineprint. Defining the fReturn returns a value only if the entry is NOT deleted.
  In your examples, your filter ALWAYS removes everything, so... nothing is returned.

  Like ConditionalPutAll, you can create your own version to return the deleted entries.

• Is there a bug with radiogroup when page elements are more than 100?

  Hello
  
  I have more than 100 items on one page and in the end there are 6 radiogroups 'yes' or a 'no', the user can select. 'Yes' or 'no' in each radiogroup are mutually exclusive, but the 6 radiogroups are independent of each other. For example:
  
  RadioGroup #1: Yes no
  
  RadioGroup #2: Yes No
  
  RadioGroup #3: Yes no
  
  RadioGroup #4: Yes no
  
  RadioGroup #5: Yes no
  
  RadioGroup #6: Yes no
  
  It works very well. However, recently, I added more elements to the page. And I realize that the user can select only one answer among all 6 radiogroups, which means that not only the 'yes' and 'no' in each radiogroup to exclude each other, but also the 6 radiogroups are exclusive mutual.
  
  I have them defined simply as "radiogroup" as follows:
  
  2812 Radiogroup #1 P20_radiogroup
  
  2813 Radiogroup #2 P20_radiogroup
  
  2814 Radiogroup #3 P20_radiogroup
  
  2815 P20_CHARTAUDITING display text (escape special characters, does not save the State) == > just as a display
  
  2816 P20_radiogroup Radiogroup #4
  
  2817 Radiogroup #5 P20_radiogroup
  
  2818 Radiogroup #6 P20_radiogroup
  
  I'm on ApEx 3.2 with EPG on Oracle 11 g. Any ideas/suggestions would be greatly appreciated!
  
  Thank you very much
  Helen

  You can not have more than 100 electronic items on a single page.

  See Re: Question about the message "number of elements exceeds 100 maximum."
  or search the limit of 100 points.

  Gus...

• Is there a bug with firefox when he supported "aria-labelledby" &lt; span &gt; element attribute?

  I use firefox 10.0.3 and using JAWS 13 to read on the screen.
  I put the attribute "aria-labelledby" for the duration, and only first item's content has been read that even I put three 'id' for this attribute. It can work perfectly in IE8.

  <table role="grid" summary="Details table">
  <tr>
  	<td type="columnTitle"><span tabIndex="0" role="gridcell">Date</span></td>
  	<th id="_NS_hdr1" role="columnheader" type="columnTitle"><span tabIndex="-1" role="columnheader">3/4</span></th>
  	<th id="_NS_hdr2" role="columnheader" type="columnTitle"><span tabIndex="-1" role="columnheader">3/11</span></th>
  	<th id="_NS_hdr3" role="columnheader" type="columnTitle"><span tabIndex="-1" role="columnheader">3/18</span></th>
  	<th id="_NS_hdr4" role="columnheader" type="columnTitle"><span tabIndex="-1" role="columnheader">3/25</span></th>
  </tr>
  <tr>
  	<th id="_NS_hdr5" role="rowheader" type="columnTitle"><span tabIndex="-1" role="rowheader">Count</span></th>
  	<td headers="_NS_hdr1 _NS_hdr5" type="datavalue">
  	<span tabIndex="-1" id="_NS_N158B5CC0.1614A0C80" role="gridcell" aria-labelledby="_NS_hdr1 _NS_hdr5 _NS_N158B5CC0.1614A0C80">0</span>
  	</td>
  	<td headers="_NS_hdr2 _NS_hdr5" type="datavalue">
  	<span tabIndex="-1" id="_NS_N158B5CC0.1614A1200" role="gridcell" aria-labelledby="_NS_hdr2 _NS_hdr5 _NS_N158B5CC0.1614A1200">75</span></td>
  	<td headers="_NS_hdr3 _NS_hdr5" type="datavalue">
  	<span tabIndex="-1" id="_NS_N158B5CC0.1614A1780"  role="gridcell" aria-labelledby="_NS_hdr3 _NS_hdr5 _NS_N158B5CC0.1614A1780">231</span>
  	</td>
  	<td headers="_NS_hdr4 _NS_hdr5" type="datavalue">
  	<span tabIndex="-1" id="_NS_N158B5CC0.1614A1D00" role="gridcell" aria-labelledby="_NS_hdr4 _NS_hdr5 _NS_N158B5CC0.1614A1D00">81</span>
  	</td>
  </tr>
  </table>

  Hello!

  I think you don't need the ARIA markup for what you're trying to accomplish. Firefox has very sophisticated algorithms correctly make accessible data tables. You are already using line and column inside the table headers, and you use the attribute headers etc. You don't need the extra markup role and aria-labelledby to create the table works the way you want. JAWS 13 also supports this interface table and should work without all the stuff of WAI-ARIA. I think because we have this sophisticated table interface, the largest part of the ARIA markup is ignored inside the correctly marked data tables.

  Make a test without the ARIA markup and report if you still have problems.

  Note that Firefox 11 improves even a couple of these things in the data tables still further, you can update.