#46593 [NEW]: $object instanceof __CLASS__ is not allowed
From: jtaal at eljakim dot nl Operating system: Irrelevant PHP version: 5.2.6 PHP Bug Type: Feature/Change Request Bug description: $object instanceof __CLASS__ is not allowed Description: The PHP parser currently cannot handle __CLASS__ directly after instanceof. I think this should be possible. Reproduce code: --- class myclass { function mymethod($otherObject) { if ($otherObject instanceof __CLASS__) { $otherObject-mymethod(null); } return 'works!'; } } $a = new myclass(); print $a-mymethod($a); Expected result: works! Actual result: -- Parse error: syntax error, unexpected T_CLASS_C, expecting T_STRING or T_VARIABLE or '$' in ... on line 4 -- Edit bug report at http://bugs.php.net/?id=46593edit=1 -- Try a CVS snapshot (PHP 5.2): http://bugs.php.net/fix.php?id=46593r=trysnapshot52 Try a CVS snapshot (PHP 5.3): http://bugs.php.net/fix.php?id=46593r=trysnapshot53 Try a CVS snapshot (PHP 6.0): http://bugs.php.net/fix.php?id=46593r=trysnapshot60 Fixed in CVS: http://bugs.php.net/fix.php?id=46593r=fixedcvs Fixed in CVS and need be documented: http://bugs.php.net/fix.php?id=46593r=needdocs Fixed in release: http://bugs.php.net/fix.php?id=46593r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=46593r=needtrace Need Reproduce Script: http://bugs.php.net/fix.php?id=46593r=needscript Try newer version: http://bugs.php.net/fix.php?id=46593r=oldversion Not developer issue: http://bugs.php.net/fix.php?id=46593r=support Expected behavior: http://bugs.php.net/fix.php?id=46593r=notwrong Not enough info: http://bugs.php.net/fix.php?id=46593r=notenoughinfo Submitted twice: http://bugs.php.net/fix.php?id=46593r=submittedtwice register_globals: http://bugs.php.net/fix.php?id=46593r=globals PHP 4 support discontinued: http://bugs.php.net/fix.php?id=46593r=php4 Daylight Savings:http://bugs.php.net/fix.php?id=46593r=dst IIS Stability: http://bugs.php.net/fix.php?id=46593r=isapi Install GNU Sed: http://bugs.php.net/fix.php?id=46593r=gnused Floating point limitations: http://bugs.php.net/fix.php?id=46593r=float No Zend Extensions: http://bugs.php.net/fix.php?id=46593r=nozend MySQL Configuration Error: http://bugs.php.net/fix.php?id=46593r=mysqlcfg
#46593 [Opn-Csd]: $object instanceof __CLASS__ is not allowed
ID: 46593 User updated by: jtaal at eljakim dot nl Reported By: jtaal at eljakim dot nl -Status: Open +Status: Closed Bug Type: Feature/Change Request Operating System: Irrelevant PHP Version: 5.2.6 New Comment: I've found the solution: $otherObject instanceof self (maybe someone can find this issue using a searchengine) Previous Comments: [2008-11-17 13:17:22] jtaal at eljakim dot nl Description: The PHP parser currently cannot handle __CLASS__ directly after instanceof. I think this should be possible. Reproduce code: --- class myclass { function mymethod($otherObject) { if ($otherObject instanceof __CLASS__) { $otherObject-mymethod(null); } return 'works!'; } } $a = new myclass(); print $a-mymethod($a); Expected result: works! Actual result: -- Parse error: syntax error, unexpected T_CLASS_C, expecting T_STRING or T_VARIABLE or '$' in ... on line 4 -- Edit this bug report at http://bugs.php.net/?id=46593edit=1
#45801 [Com]: PHPUnit testcase crashes (segfault)
ID: 45801 Comment by: jtaal at eljakim dot nl Reported By: jtaal at eljakim dot nl Status: Feedback Bug Type: Reproducible crash Operating System: * PHP Version: 5.2.6 Assigned To: fb-req-jani New Comment: Sebastian Bergman managed to narrow it down: http://bugs.php.net/bug.php?id=45805 please close this bug Previous Comments: [2008-08-13 01:04:25] [EMAIL PROTECTED] Please try using this CVS snapshot: http://snaps.php.net/php5.2-latest.tar.gz For Windows (zip): http://snaps.php.net/win32/php5.2-win32-latest.zip For Windows (installer): http://snaps.php.net/win32/php5.2-win32-installer-latest.msi [2008-08-12 16:22:27] jtaal at eljakim dot nl Description: I'm getting a segmentation fault when running a very simple PHPUnit testcase. I'm using the latest pear package from pear.phpunit.de. The testcase is issuing an E_STRICT. I think deep inside PHPUnit something happens that causes this to happen. However, since PHPUnit is just a bunch of PHP classes, I dare say this is a bug inside the PHP interpreter. I've been able to run and crash this on both Linux (debian lenny) and Windows XP SP3 (both using PHP 5.2.6). Reproduce code: --- ?php // MyTest.php require_once 'PHPUnit/Framework.php'; class MyObject { public function getX() { return NULL; } } class MyTest extends PHPUnit_Framework_TestCase { public function test1() { $obj = new MyObject; $source = $obj-getX(); } } Actual result: -- $ phpunit MyTest.php Strict Standards: Only variables should be assigned by reference in /home/jtaal/public_html/MyTest.php on line 18 PHPUnit 3.2.21 by Sebastian Bergmann. Segmentation fault -- Edit this bug report at http://bugs.php.net/?id=45801edit=1
#45801 [NEW]: PHPUnit testcase crashes (segfault)
From: jtaal at eljakim dot nl Operating system: Linux Windows PHP version: 5.2.6 PHP Bug Type: Reproducible crash Bug description: PHPUnit testcase crashes (segfault) Description: I'm getting a segmentation fault when running a very simple PHPUnit testcase. I'm using the latest pear package from pear.phpunit.de. The testcase is issuing an E_STRICT. I think deep inside PHPUnit something happens that causes this to happen. However, since PHPUnit is just a bunch of PHP classes, I dare say this is a bug inside the PHP interpreter. I've been able to run and crash this on both Linux (debian lenny) and Windows XP SP3 (both using PHP 5.2.6). Reproduce code: --- ?php // MyTest.php require_once 'PHPUnit/Framework.php'; class MyObject { public function getX() { return NULL; } } class MyTest extends PHPUnit_Framework_TestCase { public function test1() { $obj = new MyObject; $source = $obj-getX(); } } Actual result: -- $ phpunit MyTest.php Strict Standards: Only variables should be assigned by reference in /home/jtaal/public_html/MyTest.php on line 18 PHPUnit 3.2.21 by Sebastian Bergmann. Segmentation fault -- Edit bug report at http://bugs.php.net/?id=45801edit=1 -- Try a CVS snapshot (PHP 5.2): http://bugs.php.net/fix.php?id=45801r=trysnapshot52 Try a CVS snapshot (PHP 5.3): http://bugs.php.net/fix.php?id=45801r=trysnapshot53 Try a CVS snapshot (PHP 6.0): http://bugs.php.net/fix.php?id=45801r=trysnapshot60 Fixed in CVS: http://bugs.php.net/fix.php?id=45801r=fixedcvs Fixed in release: http://bugs.php.net/fix.php?id=45801r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=45801r=needtrace Need Reproduce Script:http://bugs.php.net/fix.php?id=45801r=needscript Try newer version:http://bugs.php.net/fix.php?id=45801r=oldversion Not developer issue: http://bugs.php.net/fix.php?id=45801r=support Expected behavior:http://bugs.php.net/fix.php?id=45801r=notwrong Not enough info: http://bugs.php.net/fix.php?id=45801r=notenoughinfo Submitted twice: http://bugs.php.net/fix.php?id=45801r=submittedtwice register_globals: http://bugs.php.net/fix.php?id=45801r=globals PHP 4 support discontinued: http://bugs.php.net/fix.php?id=45801r=php4 Daylight Savings: http://bugs.php.net/fix.php?id=45801r=dst IIS Stability:http://bugs.php.net/fix.php?id=45801r=isapi Install GNU Sed: http://bugs.php.net/fix.php?id=45801r=gnused Floating point limitations: http://bugs.php.net/fix.php?id=45801r=float No Zend Extensions: http://bugs.php.net/fix.php?id=45801r=nozend MySQL Configuration Error:http://bugs.php.net/fix.php?id=45801r=mysqlcfg
#45773 [NEW]: is_static_method
From: jtaal at eljakim dot nl Operating system: Irrelevant PHP version: 5.2.6 PHP Bug Type: Feature/Change Request Bug description: is_static_method Description: I'd like to be able to find out whether a method is static, without using ReflectionMethod. It keeps code cleaner that doesn't need object oriented reflection. It could also be somewhat quicker (avoiding to instantiate a Reflection object). -- Edit bug report at http://bugs.php.net/?id=45773edit=1 -- Try a CVS snapshot (PHP 5.2): http://bugs.php.net/fix.php?id=45773r=trysnapshot52 Try a CVS snapshot (PHP 5.3): http://bugs.php.net/fix.php?id=45773r=trysnapshot53 Try a CVS snapshot (PHP 6.0): http://bugs.php.net/fix.php?id=45773r=trysnapshot60 Fixed in CVS: http://bugs.php.net/fix.php?id=45773r=fixedcvs Fixed in release: http://bugs.php.net/fix.php?id=45773r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=45773r=needtrace Need Reproduce Script:http://bugs.php.net/fix.php?id=45773r=needscript Try newer version:http://bugs.php.net/fix.php?id=45773r=oldversion Not developer issue: http://bugs.php.net/fix.php?id=45773r=support Expected behavior:http://bugs.php.net/fix.php?id=45773r=notwrong Not enough info: http://bugs.php.net/fix.php?id=45773r=notenoughinfo Submitted twice: http://bugs.php.net/fix.php?id=45773r=submittedtwice register_globals: http://bugs.php.net/fix.php?id=45773r=globals PHP 4 support discontinued: http://bugs.php.net/fix.php?id=45773r=php4 Daylight Savings: http://bugs.php.net/fix.php?id=45773r=dst IIS Stability:http://bugs.php.net/fix.php?id=45773r=isapi Install GNU Sed: http://bugs.php.net/fix.php?id=45773r=gnused Floating point limitations: http://bugs.php.net/fix.php?id=45773r=float No Zend Extensions: http://bugs.php.net/fix.php?id=45773r=nozend MySQL Configuration Error:http://bugs.php.net/fix.php?id=45773r=mysqlcfg
#45768 [NEW]: parse_ini_file and the hash character
From: jtaal at eljakim dot nl Operating system: debian etch PHP version: 5.2.6 PHP Bug Type: Feature/Change Request Bug description: parse_ini_file and the hash character Description: On linux commenting lines in ini like files is oftenly done with # characters instead of ; characters (my.cnf for example) Maybe there should be a setting where you can set (an array of) the comment start characters? Reproduce code: --- array parse_ini_file ( string $filename [[, bool $process_sections ], array $comment_delimiters ]) -- Edit bug report at http://bugs.php.net/?id=45768edit=1 -- Try a CVS snapshot (PHP 5.2): http://bugs.php.net/fix.php?id=45768r=trysnapshot52 Try a CVS snapshot (PHP 5.3): http://bugs.php.net/fix.php?id=45768r=trysnapshot53 Try a CVS snapshot (PHP 6.0): http://bugs.php.net/fix.php?id=45768r=trysnapshot60 Fixed in CVS: http://bugs.php.net/fix.php?id=45768r=fixedcvs Fixed in release: http://bugs.php.net/fix.php?id=45768r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=45768r=needtrace Need Reproduce Script:http://bugs.php.net/fix.php?id=45768r=needscript Try newer version:http://bugs.php.net/fix.php?id=45768r=oldversion Not developer issue: http://bugs.php.net/fix.php?id=45768r=support Expected behavior:http://bugs.php.net/fix.php?id=45768r=notwrong Not enough info: http://bugs.php.net/fix.php?id=45768r=notenoughinfo Submitted twice: http://bugs.php.net/fix.php?id=45768r=submittedtwice register_globals: http://bugs.php.net/fix.php?id=45768r=globals PHP 4 support discontinued: http://bugs.php.net/fix.php?id=45768r=php4 Daylight Savings: http://bugs.php.net/fix.php?id=45768r=dst IIS Stability:http://bugs.php.net/fix.php?id=45768r=isapi Install GNU Sed: http://bugs.php.net/fix.php?id=45768r=gnused Floating point limitations: http://bugs.php.net/fix.php?id=45768r=float No Zend Extensions: http://bugs.php.net/fix.php?id=45768r=nozend MySQL Configuration Error:http://bugs.php.net/fix.php?id=45768r=mysqlcfg
#33870 [NEW]: relative include from symbolic linked file
From: jtaal at eljakm dot nl Operating system: linux PHP version: 4.3.11 PHP Bug Type: Unknown/Other Function Bug description: relative include from symbolic linked file Description: On my development server I have a directory: /var/www/app containing the file: menu.inc dbsettings.inc and the symbolic links db_mysql.inc - /var/export/db_mysql.inc subdir - /var/export/subdir In the directory /var/export: db_mysql.inc In the directory /var/export/subdir: include.inc The file db_mysql.inc has the following line: require_once(dbsettings.inc); The file include.inc has the following line: require_once(../menu.inc); What happens is that the include_path is ., so the file ../menu.inc cannot be found by include.inc I tried to set the include_path to /var/www/app. This worked for the file db_mysql.inc, but it didn't work for menu.inc. I tried to set the include path to .:/var/www/app, but then the file dbsettings.inc couldn't be found. I also tried /var/www/app:/var/www/app/subdir and lots of other combinations. On a release server the two directories are merged so there will be no symlinks. (The problem I have does not occur on this server) I googled lots of hours, but didn't find any thing useful. I expected PHP to open the symlinked files as if they were actually there. Can I configure PHP to do so? Reproduce code: --- /var/export/db_mysql.inc: require_once('dbsettings.inc'); /var/export/subdir/include.inc: require_once('../menu.inc'); /var/www/app/menu.inc: $menu = array( ... ); /var/www/app/dbsettings.inc: $host = 'localhost'; $user = '...'; ... symlinks: /var/www/app/db_mysql.inc - /var/export/db_mysql.inc /var/www/app/subdir - /var/export/subdir Expected result: I expected PHP to open the symlinked files as if they were actually there. Actual result: -- Warning: main(../menu.inc): failed to open stream: No such file or directory in /var/export/subdir/include.inc on line 4 Fatal error: main(): Failed opening required '../menu.inc' (include_path='/var/www/app') in /var/export/subdir/include.inc on line 4 -- Edit bug report at http://bugs.php.net/?id=33870edit=1 -- Try a CVS snapshot (php4): http://bugs.php.net/fix.php?id=33870r=trysnapshot4 Try a CVS snapshot (php5.0): http://bugs.php.net/fix.php?id=33870r=trysnapshot50 Try a CVS snapshot (php5.1): http://bugs.php.net/fix.php?id=33870r=trysnapshot51 Fixed in CVS:http://bugs.php.net/fix.php?id=33870r=fixedcvs Fixed in release:http://bugs.php.net/fix.php?id=33870r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=33870r=needtrace Need Reproduce Script: http://bugs.php.net/fix.php?id=33870r=needscript Try newer version: http://bugs.php.net/fix.php?id=33870r=oldversion Not developer issue: http://bugs.php.net/fix.php?id=33870r=support Expected behavior: http://bugs.php.net/fix.php?id=33870r=notwrong Not enough info: http://bugs.php.net/fix.php?id=33870r=notenoughinfo Submitted twice: http://bugs.php.net/fix.php?id=33870r=submittedtwice register_globals:http://bugs.php.net/fix.php?id=33870r=globals PHP 3 support discontinued: http://bugs.php.net/fix.php?id=33870r=php3 Daylight Savings:http://bugs.php.net/fix.php?id=33870r=dst IIS Stability: http://bugs.php.net/fix.php?id=33870r=isapi Install GNU Sed: http://bugs.php.net/fix.php?id=33870r=gnused Floating point limitations: http://bugs.php.net/fix.php?id=33870r=float No Zend Extensions: http://bugs.php.net/fix.php?id=33870r=nozend MySQL Configuration Error: http://bugs.php.net/fix.php?id=33870r=mysqlcfg
#32262 [NEW]: make require accept wildcards
From: jtaal at eljakim dot nl Operating system: Linux PHP version: 4.3.10 PHP Bug Type: Feature/Change Request Bug description: make require accept wildcards Description: I have a series of classes in PHP 4.3, there is no autoload feature there so I named all my files class.$className.inc.php. I created the class.inc.php file which require_onces all class.*.inc.php files. This works fine for files which are inside the same directory as class.*.inc.php BUT, when I include class.inc.php from another directory the `ls $wildcard` comes up empty, thus nothing is included. Wouldn't it be a good feature to have a require_once_wildcard which includes files. A problem that may rise is the includeSearchPath (in which directory should the wildcard be executed?). A solution is to execute only in the directory where the file resides which calls require_once_wildcard. My source for require_once_wildcard is in the Reproduce code field. Reproduce code: --- function require_once_wildcard($wildcard) { $ls = `ls $wildcard`; $ls = explode(\n, $ls); array_pop($ls); // strip last empty line print_r($ls); foreach ($ls as $inc) { require_once($inc); } } -- Edit bug report at http://bugs.php.net/?id=32262edit=1 -- Try a CVS snapshot (php4): http://bugs.php.net/fix.php?id=32262r=trysnapshot4 Try a CVS snapshot (php5.0): http://bugs.php.net/fix.php?id=32262r=trysnapshot50 Try a CVS snapshot (php5.1): http://bugs.php.net/fix.php?id=32262r=trysnapshot51 Fixed in CVS:http://bugs.php.net/fix.php?id=32262r=fixedcvs Fixed in release:http://bugs.php.net/fix.php?id=32262r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=32262r=needtrace Need Reproduce Script: http://bugs.php.net/fix.php?id=32262r=needscript Try newer version: http://bugs.php.net/fix.php?id=32262r=oldversion Not developer issue: http://bugs.php.net/fix.php?id=32262r=support Expected behavior: http://bugs.php.net/fix.php?id=32262r=notwrong Not enough info: http://bugs.php.net/fix.php?id=32262r=notenoughinfo Submitted twice: http://bugs.php.net/fix.php?id=32262r=submittedtwice register_globals:http://bugs.php.net/fix.php?id=32262r=globals PHP 3 support discontinued: http://bugs.php.net/fix.php?id=32262r=php3 Daylight Savings:http://bugs.php.net/fix.php?id=32262r=dst IIS Stability: http://bugs.php.net/fix.php?id=32262r=isapi Install GNU Sed: http://bugs.php.net/fix.php?id=32262r=gnused Floating point limitations: http://bugs.php.net/fix.php?id=32262r=float No Zend Extensions: http://bugs.php.net/fix.php?id=32262r=nozend MySQL Configuration Error: http://bugs.php.net/fix.php?id=32262r=mysqlcfg
#32262 [WFx]: make require accept wildcards
ID: 32262 User updated by: jtaal at eljakim dot nl Reported By: jtaal at eljakim dot nl Status: Wont fix Bug Type: Feature/Change Request Operating System: * PHP Version: 4.3.* New Comment: will there be a 4.4 branch? Previous Comments: [2005-03-10 12:19:58] [EMAIL PROTECTED] Sorry no new features in the 4.3 branch. [2005-03-10 12:13:31] jtaal at eljakim dot nl Description: I have a series of classes in PHP 4.3, there is no autoload feature there so I named all my files class.$className.inc.php. I created the class.inc.php file which require_onces all class.*.inc.php files. This works fine for files which are inside the same directory as class.*.inc.php BUT, when I include class.inc.php from another directory the `ls $wildcard` comes up empty, thus nothing is included. Wouldn't it be a good feature to have a require_once_wildcard which includes files. A problem that may rise is the includeSearchPath (in which directory should the wildcard be executed?). A solution is to execute only in the directory where the file resides which calls require_once_wildcard. My source for require_once_wildcard is in the Reproduce code field. Reproduce code: --- function require_once_wildcard($wildcard) { $ls = `ls $wildcard`; $ls = explode(\n, $ls); array_pop($ls); // strip last empty line print_r($ls); foreach ($ls as $inc) { require_once($inc); } } -- Edit this bug report at http://bugs.php.net/?id=32262edit=1
#20289 [NEW]: Delete $HTTP_GET_VARS on ob_start
From: [EMAIL PROTECTED] Operating system: win32 PHP version: 4.2.2 PHP Bug Type: Unknown/Other Function Bug description: Delete $HTTP_GET_VARS on ob_start I have certain reasons to believe that the function ob_start deletes some data from $HTTP_GET_VARS. At least that's what I think it is if I place a var_dump after ob_start sometimes data is missing. This time I had something like this: ?mode=managesubmode=del_useruser_id=8 But when The var_dump was executed it only contained mode. If I commented the call to ob_start then submode and user_id weren't ommited Mostly strange is that when I change one letter like submode=del_users It suddenly works again... Strange -- Edit bug report at http://bugs.php.net/?id=20289edit=1 -- Try a CVS snapshot: http://bugs.php.net/fix.php?id=20289r=trysnapshot Fixed in CVS: http://bugs.php.net/fix.php?id=20289r=fixedcvs Fixed in release: http://bugs.php.net/fix.php?id=20289r=alreadyfixed Need backtrace: http://bugs.php.net/fix.php?id=20289r=needtrace Try newer version: http://bugs.php.net/fix.php?id=20289r=oldversion Not developer issue:http://bugs.php.net/fix.php?id=20289r=support Expected behavior: http://bugs.php.net/fix.php?id=20289r=notwrong Not enough info:http://bugs.php.net/fix.php?id=20289r=notenoughinfo Submitted twice:http://bugs.php.net/fix.php?id=20289r=submittedtwice register_globals: http://bugs.php.net/fix.php?id=20289r=globals PHP 3 support discontinued: http://bugs.php.net/fix.php?id=20289r=php3 Daylight Savings: http://bugs.php.net/fix.php?id=20289r=dst IIS Stability: http://bugs.php.net/fix.php?id=20289r=isapi
#20289 [Com]: Delete $HTTP_GET_VARS on ob_start
ID: 20289 Comment by: [EMAIL PROTECTED] Reported By: [EMAIL PROTECTED] Status: Open Bug Type: Unknown/Other Function Operating System: win32 PHP Version: 4.2.2 New Comment: The question I asked becomes more specific. I discovered that when I use a switch: switch($submode) and in the switch exists a case del_user: then data is deleted from $_GET or $HTTP_GET_VARS, but when I change it to case del_user_tmp: or something different... It doesnt work anymore Previous Comments: [2002-11-06 12:04:39] [EMAIL PROTECTED] I have certain reasons to believe that the function ob_start deletes some data from $HTTP_GET_VARS. At least that's what I think it is if I place a var_dump after ob_start sometimes data is missing. This time I had something like this: ?mode=managesubmode=del_useruser_id=8 But when The var_dump was executed it only contained mode. If I commented the call to ob_start then submode and user_id weren't ommited Mostly strange is that when I change one letter like submode=del_users It suddenly works again... Strange -- Edit this bug report at http://bugs.php.net/?id=20289edit=1
#20289 [Com]: Delete $HTTP_GET_VARS on ob_start
ID: 20289 Comment by: [EMAIL PROTECTED] Reported By: [EMAIL PROTECTED] Status: Open Bug Type: Unknown/Other Function Operating System: win32 PHP Version: 4.2.2 New Comment: I get this directly from the function phpinfo() _ENV[QUERY_STRING] mode=manage_sections _ENV[REQUEST_URI] /leden/bijwerken/bijwerken.php?mode=manage_sectionssubmode=del_section_ddel_name=CAO strange It seems that the query string isn't correct Previous Comments: [2002-11-06 15:09:07] [EMAIL PROTECTED] The question I asked becomes more specific. I discovered that when I use a switch: switch($submode) and in the switch exists a case del_user: then data is deleted from $_GET or $HTTP_GET_VARS, but when I change it to case del_user_tmp: or something different... It doesnt work anymore [2002-11-06 12:04:39] [EMAIL PROTECTED] I have certain reasons to believe that the function ob_start deletes some data from $HTTP_GET_VARS. At least that's what I think it is if I place a var_dump after ob_start sometimes data is missing. This time I had something like this: ?mode=managesubmode=del_useruser_id=8 But when The var_dump was executed it only contained mode. If I commented the call to ob_start then submode and user_id weren't ommited Mostly strange is that when I change one letter like submode=del_users It suddenly works again... Strange -- Edit this bug report at http://bugs.php.net/?id=20289edit=1