Qyery returning empty result using dbms_xmlgen doesn't work as expected

Hi group.
I'm using the following code in a PSQL procedure
procedure selectXML(consulta in varchar2,xmlout out clob) as
context dbms_xmlquery.ctxtype;
resultado clob;
tam integer;
context_gen dbms_xmlgen.ctxHandle;
begin
IF useDBMS_XMLGEN THEN
BEGIN
context_gen:=DBMS_XMLgen.newContext(consulta);
DBMS_XMLgen.usenullattributeindicator(context_gen,true);
resultado:=DBMS_XMLgen.getxml(context_gen);
tam:=dbms_lob.getlength(resultado);
dbms_lob.copy(xmlout,resultado,tam,1,1);
DBMS_XMLgen.closecontext(context_gen);
ELSE
context:=DBMS_XMLQUERY.newContext(consulta);
DBMS_XMLQUERY.usenullattributeindicator(context,true);
resultado:=DBMS_XMLQUERY.getxml(context);
tam:=dbms_lob.getlength(resultado);
dbms_lob.copy(xmlout,resultado,tam,1,1);
DBMS_XMLQUERY.closecontext(context);
END IF;
end;
If the query used returns nothing (for example 'SELECT * FROM dual WHERE rownum=0') using the DBMS_XMLQUERY code, the function getxml() returns
<?xml version = '1.0'?>
<ROWSET/>
but using the DBMS_XMLGEN code, the function getxml() returns nothing.
Is that behaviour normal or am I doing anything wrong?
I'm testing the code in Oracle9i Enterprise Edition Release 9.2.0.7.0 but it must also be run Oracle 8 and 10
Thanks in advance
Javier

Is that behaviour normal or am I doing anything wrong?That seems to be normal behaviour per design:
SQL> select 1 id, dbms_xmlgen.getxml('select * from dual where 1=0') xml from dual
union all
select 2, dbms_xmlquery.getxml('select * from dual where 1=0',0) xml from dual
union all
select 3, dbms_xmlquery.getxml('select * from dual where 1=0',1) xml from dual
union all
select 4, dbms_xmlquery.getxml('select * from dual where 1=0',2) xml from dual
ID XML
1
2 <?xml version = '1.0'?>
    <ROWSET/>
3 <?xml version = '1.0'?>
    <!DOCTYPE ROWSET [
    <!ELEMENT ROWSET (ROW)*>
    <!ELEMENT ROW (DUMMY)>
    <!ATTLIST ROW num CDATA #REQUIRED>
    <!ELEMENT DUMMY (#PCDATA)>
    ]>
    <ROWSET/>
4 <?xml version = '1.0'?>
    <DOCUMENT xmlns:xsd="http://www.w3.org/2000/10/XMLSchema">
       <xsd:schema xmlns:xsd="http://www.w3.org/2000/10/XMLSchema">
          <xsd:element name="ROWSET">
             <xsd:complexType>
                <xsd:sequence>
                   <xsd:element name="ROW" minOccurs="0" maxOccurs="unbounded">
                      <xsd:complexType>
                         <xsd:sequence>
                            <xsd:element name="DUMMY" type="xsd:string" nullable="true" minOccurs="0"/>
                         </xsd:sequence>
                         <xsd:attribute name="num" type="xsd:integer"/>
                      </xsd:complexType>
                   </xsd:element>
                </xsd:sequence>
             </xsd:complexType>
          </xsd:element>
       </xsd:schema>
       <ROWSET xmlns:xsi="http://www.w3.org/2000/10/XMLSchema-instance" xsi:noNamespaceSchemaLocation="#
    /DOCUMENT/xsd:schema[not(@targetNamespace)]"/>
    </DOCUMENT>
4 rows selected.

Similar Messages

I have Iphone 4G. I bought when I studied in Usa. Then I returned to Uzbekistan this Iphone doesn't work. How I can use this phone? Can you help me?

I have Iphone 4G. I bought when I studied in Usa. Then I returned to Uzbekistan this Iphone doesn't work. How I can use this phone? Can you help me?

shohjahon wrote:
I have Iphone 4G.
No, you do not. It simply is not possible to have a device that does not exist.
Most likely, you have an iPhone 4. There is no 4G iPhone.
shohjahon wrote:
I bought when I studied in Usa. Then I returned to Uzbekistan this Iphone doesn't work.
Yes, it does work. It is carrier locked to AT&T, just like all GSM iPhones sold in the US since their introduction.
shohjahon wrote:
How I can use this phone?
You can use the phone as it is, but it will be very expensive with AT&T international roaming plans.
There is no official way to unlock the phone for use with another carrier.
Your options:
Sell the device and purchase one intended for use in your country.
Find unofficial method to unlock
Use as an iPod Touch.
No one here can provide you with assistance on getting you AT&T locked iPhone to work on another carrier.

My new iPad air is suddenly asking for a pass code. I have only just finished restoring it from my iPad 2 and I didn't set one. The one I used before doesn't work. What can I do?

My new iPad air is asking for a pass code. I have only just finished restoring it via iTunes from my ipad2 and I didn't set one. The one I used before doesn't work. What can I do?

http://support.apple.com/kb/HT1212

I've had my ipod shuffle (2nd generation) for a long time now and after a year and a bit of not being in use it doesn't work. When I connect it to the computer it doesn't come up. Can my ipod be charged or fixed?

I've had my ipod shuffle (2nd generation) for a long time now and after a year and a bit of not being in use it doesn't work. When I connect it to the computer it doesn't come up. Can my ipod be charged or fixed? Please help!

That is about normal. You can maximize battery life by changing some settings. See:
Apple - Batteries - iPod
Apple will exchange your iPod for a refurbished one for with a new battery for $79. They do not fix yours.
Apple - iPod Repair price
A third-party place like the following will replace your battery for less. Google for more.
iPhone Repair, Service & Parts: iPod Touch, iPad, MacBook Pro Screens
Replace the battery yourself if you are up to it. You can purchase a replacment for about $10
iPod Touch Repair – iFixit

Got M-Cell that gives 5 bars, got full-strenth wifi, Internet on phone doesn't work, any app that uses internet doesn't work, seems like its started when upgraded to 7.1.1

I have an AT&T M-Cell that gives 5 bars, got full-strenth wifi. Same wi-fi that my computer uses, that my Kindle uses. So not capacity issue. but my Internet on phone drops out continuously and frequently, any app that uses internet doesn't work at all. It seems like it started when I upgraded to 7.1.1 I have connectivity like crazy, so why doesn't the internet work?

Reboot your router by powering it down for 30-60 seconds.

ArrayDeque as a stack doesn't work as expected with complex objects

Trying to use ArrayDeque as a strorage for complex values <PSList<PSol>> (i.e. Arraylists of structured Values PSol), this doesn't work as expected. The code below should produce different values of pSLWk, being stored on bkStack, which are then to be retrieved by pop() to the variables pSL1, pSL2, pSL3.
However, retrieval only ends up with three identical data sets (variables) pS1,pS2, pS3.
        public PSList<PSol> pSL;
        private ArrayDeque<PSList<PSol>> bkStack=new ArrayDeque<PSList<PSol>>();
        pSLWk=new PSList<PSol>();       // Constructor copies some Array (static field) to the PSLists
        pSL=new PSList<PSol>();
        pSLAux=new PSList<PSol>(pSLWk);   // Constructor copies from existing PSList
        pSLWk.checkResult("pSLWk prior to setDefaults - modifies pSLWk !");
        setDefaults();                                                            // Modifies pSLWk only
        pSLWk.checkResult(" pSLWk after setDefaults");              // .. got changes (o.k.)
        pSL.checkResult(" pSL after setDefaults");                  // .. unchanged (o.k.)
        pSL.checkResult(" pSLAux after setDefaults");               // .. unchanged (o.k.)
        bkStack.push(new PSList<PSol>(pSLWk));                      // store changes in bkStack
        pSLWk.getEl(77).setVal(new StringBuffer("4"));              // change pSLWk again (value 4 @ 77)
        pSLWk.checkResult("pSLWk, after PUSH, THEN modify 4@77");   // .. got change (o.k.)
        pSL.checkResult("pSL after setVal 77");
        bkStack.push(pSLWk);                      // store changes in bkStack
        pSLWk.getEl(80).setVal(new StringBuffer("8"));              // change pSLWk again (value 8 @ 80)
        pSLWk.checkResult("pSLWk after setVal 8@80");               // .. got change (o.k.)
        pSL.checkResult("pSL after setVal 80");
        bkStack.push(new PSList<PSol>(pSLWk));                      // store changes in bkStack
        pSL1=new PSList<PSol>(bkStack.pop());
        pSL1.checkResult("pSL1 after 1st pop");
//      pSL1=bkStack.pop()                                          // Straightforward way doesn't work either...
        pSL2=new PSList<PSol>(bkStack.pop());
        pSL2.checkResult("pSL2 after 2nd pop");
        pSL3=new PSList<PSol>(bkStack.pop());
        pSL3.checkResult("pSL3 after 3rd pSLWk=..pop()");Here the result from the code above:
debug:
1 8 9   4 5 42 9 7 5 2 6 4 923     1 7   3         86   7 5 3 69     2   63 <C.R.>pSLWk prior to setDefaults - modifies pSLWk !
128 96 4635 42 9 7945 2 6 47923     1 7   3 9       86   7 5 3 69     2   63 <C.R.> pSLWk after setDefaults
1 8 9   4 5 42 9 7 5 2 6 4 923     1 7   3         86   7 5 3 69     2   63 <C.R.> pSL after setDefaults
1 8 9   4 5 42 9 7 5 2 6 4 923     1 7   3         86   7 5 3 69     2   63 <C.R.> pSLAux after setDefaults
128 96 4635 42 9 7945 2 6 47923     1 7   3 9       86   7 5 3 69     2 463 <C.R.>pSLWk, after PUSH, THEN modify 4@77
1 8 9   4 5 42 9 7 5 2 6 4 923     1 7   3         86   7 5 3 69     2   63 <C.R.>pSL after setVal 77
128 96 4635 42 9 7945 2 6 47923     1 7   3 9       86   7 5 3 69     2 4638<C.R.>pSLWk after setVal 8@80
1 8 9   4 5 42 9 7 5 2 6 4 923     1 7   3         86   7 5 3 69     2   63 <C.R.>pSL after setVal 80
128 96 4635 42 9 7945 2 6 47923     1 7   3 9       86   7 5 3 69     2 4638<C.R.>pSL1 after 1st pop
128 96 4635 42 9 7945 2 6 47923     1 7   3 9       86   7 5 3 69     2 4638<C.R.>pSL2 after 2nd pop
128 96 4635 42 9 7945 2 6 47923     1 7   3 9       86   7 5 3 69     2 4638<C.R.>pSL3 after 3rd pSLWk=..pop()
1 8 9   4 5 42 9 7 5 2 6 4 923     1 7   3         86   7 5 3 69     2   63 <C.R.>pSLWk prior to setDefaults - modifies pSLWk !
128 96 4635 42 9 7945 2 6 47923     1 7   3 9       86   7 5 3 69     2   63 <C.R.> pSLWk after setDefaults
1 8 9   4 5 42 9 7 5 2 6 4 923     1 7   3         86   7 5 3 69     2   63 <C.R.> pSL after setDefaults
1 8 9   4 5 42 9 7 5 2 6 4 923     1 7   3         86   7 5 3 69     2   63 <C.R.> pSLAux after setDefaults
128 96 4635 42 9 7945 2 6 47923     1 7   3 9       86   7 5 3 69     2 463 <C.R.>pSLWk, after PUSH, THEN modify 4@77
1 8 9   4 5 42 9 7 5 2 6 4 923     1 7   3         86   7 5 3 69     2   63 <C.R.>pSL after setVal 77
128 96 4635 42 9 7945 2 6 47923     1 7   3 9       86   7 5 3 69     2 4638<C.R.>pSLWk after setVal 8@80
1 8 9   4 5 42 9 7 5 2 6 4 923     1 7   3         86   7 5 3 69     2   63 <C.R.>pSL after setVal 80
128 96 4635 42 9 7945 2 6 47923     1 7   3 9       86   7 5 3 69     2 4638<C.R.>pSL1 after 1st pop
128 96 4635 42 9 7945 2 6 47923     1 7   3 9       86   7 5 3 69     2 4638<C.R.>pSL2 after 2nd pop
128 96 4635 42 9 7945 2 6 47923     1 7   3 9       86   7 5 3 69     2 4638<C.R.>pSL3 after 3rd pSLWk=..pop()What's the problem with this ?
Rem: I tried the simple approach as well:
bkstack.push(pSLWk);
...

Thank you for your comments, although I see we still don't have a common understanding of the problem.
Firstly, I add the code for the PSList and the PSol classes, so you might find some problem with that:
     public class PSol     {
          private StringBuffer val;
          private int zI;
          private int sI;
          private int bI;
                    // == Konstruktor
          public PSol( StringBuffer v, int z, int s, int b )     {
               this.val=v;
               this.zI=z;
               this.sI=s;
               this.bI=b;
                    // == Getter,Setter
          public StringBuffer getVal()     {return val;}
          public int getZ()     {return zI;}
          public int getS()     {return sI;}
          public int getB()     {return bI;}
          public int getVSize()     {return val.length();}
          public void setVal(StringBuffer v)     {val=v;}
          public boolean hasVChar( StringBuffer ch, boolean delCh )     {
               boolean bT=false;
               StringBuffer fSt=getVal();
               if (!(fSt.indexOf( ch.toString() )     == -1))     {
                    bT=true;
                    if (delCh)     {
                         setVal(fSt.deleteCharAt(fSt.indexOf( ch.toString() )));
               return bT;
     }     // PSol
     public class PSList<E> extends ArrayList<PSol>     {
               /**     Construktor 1: PSList(v,z,s,b) - makes list from single arrays
          private static final long serialVersionUID = 4711L;                         // ### JAVAC Warning! ###
        public PSList (String[] vS, int[] z, int[] s, int[] b) {
               StringBuffer[] v=new StringBuffer[valDim];
            for (int i=0;i<valDim;i++) {
                v=new StringBuffer(vS[i]);
//ArrayList<PSol> pSL=new ArrayList<PSol>;
               for (int i=0; i<valDim; i++) {
                    this.add( new PSol( v[i], z[i], s[i], b[i] ) );
/** Konstruktor2 : makes list from matrix array
public PSList () {
for (int j=0; j<nDim; j++) {
for (int i=0; i<nDim; i++) {
this.add( new PSol( new StringBuffer(sGuiArr[i][j]), i, j , i/locDim + (j/locDim)*locDim) );
                    /**     ------- Construktor 3 : PSList(PSList pS) - makes list as a copy of an existing one
public PSList ( PSList<PSol> pX )     {
               super (pX); // ArrayList-Constructor (Collection)
// get Element <PSol>
public PSol getEl ( int i )     {return get(i);}
     public int getCount()     {return size();}
     public int getTValLg()     {
               int lg=0;
               for (int i=0; i<getCount(); i++)     {
                    lg=lg + getEl(i).getVal().length();
               return lg;
                    /**     ------- checkResult()     -     Check if alll elements are single char +dump
     public boolean checkResult(String messg)     {
               boolean allOne=true;
               for (int i=0; i<size(); i++)     {
                    if ( getEl(i).getVal().length() > 1 )     {
                         allOne=false;
                         System.out.print(" ");
               else     {
                    System.out.print(getEl(i).getVal());
               System.out.println("<C.R.>"+messg);
               return allOne;
     }     // Class PSList
Secondly, I don't really see what you mean by pointing out to 'only one "pSLWk" instance of PSList'. The variable pSLWk is the variable to be worked upon; after some change of the contents, I want to save this state of contents to the stack. When I pop that variable from the stack, I wouldn't want to restore it to pSLWK, but to some other variable, e.g. by public PSList<PSol> pSL1;
pSL1=new PSList<PSol>(bkStack.pop());Again - to my understanding (which comes from old days of microprocessor coding... - there shouldn't be a need to know how the data came there, or what was the name of the variable who stored it there. And : the implementation of ArrayDeque returns 'elements' of class E, not references !
Thirdly, you're right, that the method of using a copy constructor for retrieval looks 'weird'. However - I had some other versions that didn't work either, e.g. the straightforward one, as I pointed out.
And fourthly: yes, I'm almost sure that I'm messing up something somewhere. I went to this forum hoping to clarify that ... :)
If you don't mind, could you please sketch a few lines of code, how to 'push' a complex variable to a ArrayDeque stack, and retrieve it - by 'pop()' - to some to other variable of the same class later ?
Might make our discussion much easier, to see how things REALLY work.
Thank you !

VoiceOver navigation buttons doesn't work as expected on latest version of Firefox

VoiceOver navigation buttons doesn't work as expected on latest version of Firefox. VO+Left/Right not jump from element to element
Version of OS: OS X Yosemite 10.10.1
Version of Chrome: 42.0.2311.90 (64-bit)
Version of Firefox: 37.0.2

Did this work in a previous version of Firefox?
There is an open bug to improve accessibility on OS X. (Bug 336306) I'll update that bug with any relevant information you provide here.

Fn_my_permissions returns empty result yet my account is a member of sysadmin?

I was trying to see what permissions I had on another server using fn_my_permissions
USE Database_name_here
SELECT * FROM sys.fn_my_permissions(NULL, 'SERVER')
USE Database_2_name_here
SELECT * FROM sys.fn_my_permissions(NULL, 'SERVER')
The one DB that returns the expected list of permissions is the only DB where I am the owner. Others return an empty result.
My (domain) account is in the local administrators group which is in the sysadmin server role.
Whats going on here? Does my account have to be explicitly granted permissions to those schemas? Is role membership not enough?

Hi,
SELECT * FROM fn_my_permissions(NULL, 'SERVER') returns a list of the effective permissions of the caller on the server.
For the server itself, a null value (meaning "current server") is required and fn_my_permissions cannot check permissions on a linked server.
In my option, the server level permission should not change with the databases you specified. You can re-run the code and see if it returns empty by chance.
To return a list of the effective permissions of the caller on the AdventureWorks2012 database:
USE AdventureWorks2012;
SELECT * FROM fn_my_permissions (NULL, 'DATABASE');
GO
sys.fn_my_permissions (Transact-SQL)
http://technet.microsoft.com/en-us/library/ms176097.aspx
Tracy Cai
TechNet Community Support

Where used list doesn't work for global data

Hi,
When i try to do where used list on a constant in an include, it doesn't give the results for classes.
The class contains the include in a method that uses the constant, and the where used list for that constant doesn't find the appearance of the constant in the class.
I tried to look for oss note that solves the problem but i couldn't find any.

I tried to update the navigation index but it still doesn't work. from my debuging i found out that it just doesn't look in classes, all classes. I mean even in the where used itself there is not an option for used in classes. I checked it and it's a generic problem of the where used list for global data.
Edited by: Cohen Lior on Apr 14, 2010 6:52 PM

I bought app .it coast me 24.99 . And when I wanna use it doesn't work?

I bought Arab app and I paid 24.99 $ . But the app doesn't work with me? Thanx

Have you contacted the developer of the app about the problem?

MAX function to get recent date doesn't work as expected

Hi,
I recently started working on Oracle and came across this problem. I coded below given query to get the latest date from table and compare it with Sysdate + 1 timestamp. I am not able to understand what I coded wrong but the query gives me wrong results as explained below :
select NVL(MAX(to_char(max(last_update),'mm/dd/yyyy hh12:mi:ss AM')), to_char(sysdate + 1,'mm/dd/yyyy')||' 12:00:00 AM')
from Audit_Table_Name
where col1 = 'AA'
and location_type = 'STATE'
and original_flag = 'Y'
group by col1,location_type,original_flag;
Here Max(last_update) from Audit Table Contains value --> 08/25/2009 12:00:00 AM
Note : Data type of last_update is Date
so ideally speaking it should give 08/26/2009 12:00:00 AM but surprisingly it is giving 08/25/2009 12:00:00 AM as output. Anyone can explain why this is happening and what we need to do to get correct results?
Regards,
Amol

1) Why would you expect this to return a date of 8/26? NVL returns the first parameter unless it is NULL. If the first parameter is NULL, it returns the second parameter. Since
MAX(to_char(max(last_update),'mm/dd/yyyy hh12:mi:ss AM'))returns a non-NULL value, that value will be returned. Oracle never needs to evaluate the second parameter
to_char(sysdate + 1,'mm/dd/yyyy')||' 12:00:00 AM'You state that you want to compare the MAX( last_update ) to SYSDATE + 1. Your code isn't doing a comparison now. And it's not clear from your description what comparison you actually want to do. Can you elaborate a bit on what you want to compare and what the output should be depending on the result of that comparison?
2) You almost certainly want to do as much of the logic as possible using dates and only convert to a string at the very end. Otherwise, you risk comparing two strings that represent dates using string comparison semantics and getting a result you don't expect.
3) My guess (and it is just a guess) is that you want something like
SELECT GREATEST( max(last_update), trunc(sysdate+1) )or
SELECT TO_CHAR( GREATEST( max(last_update), trunc(sysdate+1) ), 'mm/dd/yyyy hh12:mi:ss AM' )Justin

Connecting to other computers on network doesn't work as expected.

This is kind of interesting.
I do this: From Menu Go > Network
Then I choose either my iMac or MacBook Pro and double click the user name for my home folders on the other computers.
A window opens up and the bottom right of the window the little twirling busy thing just goes and goes and goes.
I notice the server icon on the Desktop and double click it and everything is there an access is fast, just like I expected from the first window that popped up. But it's still there, blank, with a little twirly thing (very technical).
I get what I'm after, but it doesn't work the way it should.

Try keeping your shares from sleeping during the expected period of use--assuming they are. If this procedure works, then you'll need to decide between the "lesser of the two evils."

Webutil: webutil_file.file_open_dialog with filter doesn't work as expected

Hi
I just tried to open files by using webutil_file.file_open_dialog. Since I only want to select files with the prefix "brp" (e.g. brp0001.txt, brp0002.lis), I fill the parameter file_filter with the value: brp*.*
This doesn't work. The dialog only shows folders, but not the brp*.* files. I tried with switching to capitals, which doesn't solve anything. With the filter set to *.* the dialog shows all folders and files, so that seems to work ok.
Is anyone aware of this problem and know a solution? I'm using webutil 1.0.2 (beta).

K..here's some logging. As far as I can see it's all normal, but it doesn't work (no matter if I use a last | or not). Should be easy to reproduce.
2003-okt-29 09:42:44.547 WUF[setProperty()] Setting property WUF_GFN_DIRNAME to y:\data\
2003-okt-29 09:42:44.567 WUF[setProperty()] Setting property WUF_FILENAME to false
2003-okt-29 09:42:44.567 WUF[setProperty()] Setting property WUF_FILTER to (brp_archief*.*)|brp_archief*.*
2003-okt-29 09:42:44.587 WUF[setProperty()] Setting property WUF_GFN_MESSAGE to bestandsnaam selecteren
2003-okt-29 09:42:44.597 WUF[setProperty()] Setting property WUF_GFN_MULTISELECT to FALSE
2003-okt-29 09:42:44.608 WUF[getProperty()] Getting property WUF_GFN_OPENFILE
2003-okt-29 09:42:45.679 WUF[gfnDialog()] Creating Custom File Filter : Mask=brp_archief*.*, Label=(brp_archief*.*)
2003-okt-29 09:42:45.679 WUF[gfnDialog()] Open File mode
2003-okt-29 09:45:27.203 WUF[getProperty()] Value of WUF_GFN_OPENFILE=

Vi.lib\Utility\config.llb doesn't work as expected in 6.1

I have to modify part of a LV project written in 6.1. I started of writing a patch/VI for the project in 7.1.1, tested it and it worked fine. Then I save it from 7.1.1 > 7.0 > 6.1 and the code doesn't work. I have attached the 6.1 and 7.1.1 code that simply modifies a config data file.
Run both the VI version on the attached data file (save copies of it beforehand as the VI overwites/modifies the .dat file) and you should see for e.g.
4.6.2.1 Name=Pass
Change to
4.6.2.1 Name="4.6.2.1 Extend Ground Cycle"
with 7.1.1
and no change with 6.1
Attachments:
Modify Configuration File.zip ‏52 KB

The problem is at the very end. The 'Close Config Data.vi' changed between 6.1 and 7.0. The input 'Write Configuration File' on 7.0 and higher is default to True, any version before that was default to False. With this set to false, the file is not actually written.
So on 6.1, you need to connect a boolean constant or control set to True to this termianl to write the file.
Ed
Ed Dickens - Certified LabVIEW Architect - DISTek Integration, Inc. - NI Certified Alliance Partner
Using the Abort button to stop your VI is like using a tree to stop your car. It works, but there may be consequences.

Finder 'Back' button doesn't work as expected

Try this in Leopard Finder: In Finder View Options (Command-J), set Applications to always open in icon view, and Documents to always open in list view. Now open a fresh window and click each in the sidebar. They open as expected, one in icon view, and the other in list view. Now click the back button, and you're looking at the folder you just had open half a second earlier, but it's suddenly shifted to a different view - a view you explicitly told the Finder you didn't want this folder to open in. How does this make sense?
(Thanks to grasshopperus for posting this first at arstechnica: http://episteme.arstechnica.com/eve/forums/a/tpc/f/8300945231/m/982000038831/p/1)

Back button behaviour can be pretty wierd. I usually use the keyboard combination, but for some reason it doesn't work on PDF's and I have to end up using the graphical back button.
There are some websites out there that for some reason or another break the Back button's behaviour. Sometimes I have to hold down the mouse button to get the back button menu in order to work around that problem.
In another case there was this page in Amazon that wouldn't allow me to go back to. Everytime I went to that page the browser to automatically kick me to the next page. I had to disable javascript to work around that problem.

Qyery returning empty result using dbms_xmlgen doesn't work as expected

Similar Messages

Maybe you are looking for