[PHP-DEV] Re: /ext/session session.c

[Yasuo Ohgaki]

Yasuo Ohgaki wrote:

> There is yet another problem related to external session save handler 
> modules...
> 
> It seems session_module_name() causes race and segfaults.
> A user reported bug for mm save handler also. (I've noticed
> the problem with pgsql save handler before the bug report, though)
> 
> Bug #14780
> 


I cannot reproduce segfault after cvs update.

I guess somebody is fixed :)

--

Yasuo Ohgaki


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14809 Updated: global session vars are being assigned by value, not reference

[yohgaki]

ID: 14809
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Bogus
Bug Type: Session related
Operating System: linux 2.4.7-10
PHP Version: 4.1.1
New Comment:

I guess you are using session register within a function, right?
If so, please ask support question to php-general.
If not, reopen this bug report with short/complete script.

Previous Comments:


[2002-01-02 16:51:39] [EMAIL PROTECTED]

hi
This seems to be a problem with the way session variables register in
4.1.1
I have defined a session variable called $session thus:
session_start();
session_register("session");
global $session;
$session['profileName'] = "name";

This seems to work fine.  On the next page I do:

session_start();
global $session;
echo "profileName=".$session['profileName']; (displays name)
$session['profileName']="new name";

however, this doesn't seem to actually change the session variable,
since on the next page if i echo $session['profileName'] it still shows
'name' instead of 'new name'

I can modify the session var value using
$_SESSION[session][profileName]="new name"; and then it works fine on
the next page.

The latter format worked with 4.0.5, i just installed 4.1.1 and now
apparently I have to modify all my scripts..

Here is my config:
'./configure' '--with-mysql' '--with-apxs=/usr/local/apache/bin/apxs'
'--enable-track-vars=yes'

register_globals is on.
apache version is 1.3.20






Edit this bug report at http://bugs.php.net/?id=14809&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14397 Updated: Wrong results when reading a float

[lobbin]

ID: 14397
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Feedback
Bug Type: Sybase (dblib) related
Operating System: Linux
PHP Version: 4.0.6
New Comment:

Can you reproduce this error with PHP 4.1.1?

Previous Comments:


[2001-12-10 02:36:19] [EMAIL PROTECTED]

I found that the wrong float results happen with float(8) but NOT with
float(16) !!!



[2001-12-10 02:05:59] [EMAIL PROTECTED]

I have successfully written a float number into a sybase system - that I
can display correctly with isql.
But when reading with php I got back 0 oder some strange numbers as
6.784567 instead of teh correct float number.





Edit this bug report at http://bugs.php.net/?id=14397&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14354 Updated: sybase_query returns 1 regardless of delete success

[lobbin]

ID: 14354
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Feedback
Bug Type: Sybase (dblib) related
Operating System: Linux Mandrake 8
PHP Version: 4.0.6
New Comment:

Does this problem still exist on 4.1.1?

Previous Comments:


[2001-12-12 06:41:44] [EMAIL PROTECTED]

oh, ok, then it is not bogus.



[2001-12-12 06:24:21] [EMAIL PROTECTED]

I'm not sure if I can agree with you on that. If I don't use the @ sign,
sybase_query displays a warning of the type

Warning: Sybase message: Children still exist in 'table1'. Cannot delete
parent 'table2'. (severity 16) in /path/to/my/script.php on line 1432

Doesn't that mean that sybase_query didn't return ok? 

Besides, according to the documentation, the function
sybase_affected_rows is only implemented for the ct library, not dblib.
Unfortunately, we haven't been able to get sybase-ct to work for us,
even though we tried all the hints we could get.




[2001-12-11 13:58:59] [EMAIL PROTECTED]

Actually, this is not a bug. The query still executes ok, but the result
is different then you want. You can check with the sybase_affected_rows
function if it is deleted or not:
www.php.net/sybase_affected_rows

Derick



[2001-12-11 07:36:16] [EMAIL PROTECTED]

Hello? Anybody home? I kinda need this problem to be solved, please.



[2001-12-05 18:53:13] [EMAIL PROTECTED]

When trying to delete a record that Sybase will not allow to be deleted
because of referential integrity constraints, sybase_query returns 1.
When the record can be deleted successfully, the return code is also
1.

Example:
// Delete record with referential integrity constraint
$strSQL = "delete from MyTable where MyId=3";
$iResult = sybase_query($strSQL, $iLink );
echo "Query Result: $iResult";  // Displays 1

// Delete record without referential integrity constraint
$strSQL = "delete from MyTable where MyId=4";
$iResult = sybase_query($strSQL, $iLink );
echo "Query Result: $iResult";  // Displays 1, too

We use Sybase ASE 11.9.2 on a Linux Mandrake 8 Box.
We have tried several times to run sybase-ct instead of the dblib, but
could never make it work.

Our current configure line:

./configure \
--with-apxs=/usr/sbin/apxs \
--with-pear \
--with-gd=yes \
--with-tiff-dir=/usr/local \
--with-png-dir=/usr/local \
--with-jpeg-dir=/usr/local \
--with-ttf=/usr/src/freetype-1.3.1 \
--with-zlib \
--with-xml \
--with-ftp \
--with-mcal=../libmcal \
--with-imap=/usr/local \
--with-sybase=/opt/sybase-11.9.2 \
--with-mysql=/usr \
--with-mcrypt=/usr/local \
--with-mhash \
--with-pdflib \
--with-zlib-dir=/usr/local \
--with-fdftk=../fdftk \
--with-config-file-path=/etc/httpd/conf \
--enable-versioning \
--enable-track-vars

We use both PHP3 and PHP4 as apache modules; it's a test server, we need
it that way.

Please contact me asap if you need more information.
Michael.






Edit this bug report at http://bugs.php.net/?id=14354&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #11188 Updated: translate national characters in freeeTDS-driver

[lobbin]

ID: 11188
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Feedback
Bug Type: Sybase (dblib) related
Operating System: freeBSD
PHP Version: 4.0.5
New Comment:

Can you
  - try this with a newer version of freetds
  - try this with PHP 4.1.1

and see if the problem still exists...

Previous Comments:


[2001-06-30 15:04:57] [EMAIL PROTECTED]

This is not related to the mssql extension. You are using the sybase
extension.



[2001-05-30 04:33:07] [EMAIL PROTECTED]

I have: FreeBSD-FreeTDS-MSSQL 7.0
This line
mssql_query("SELECT * FROM table WHERE field='national characters'")
dont normal work

function ascci2unicode() in FreeTDS-driver is not good!!!







Edit this bug report at http://bugs.php.net/?id=11188&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #11171 Updated: Faulty MSSQL-queries locks PHP

[lobbin]

ID: 11171
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Feedback
Bug Type: Sybase (dblib) related
Operating System: FreeBSD
PHP Version: 4.0.5
New Comment:

Can you try this with PHP 4.1.1?

Previous Comments:


[2001-07-02 07:08:58] [EMAIL PROTECTED]

But the Sybase-library is used for MSSQL anyhow, so what's the
difference? It still doesn't work.



[2001-06-30 15:11:33] [EMAIL PROTECTED]

This is not related to the mssql extension. You are using the sybase
extension.



[2001-05-29 11:15:03] [EMAIL PROTECTED]

Bug #7881, as posted by [EMAIL PROTECTED] and bug #9244 is still present
in PHP4.0.5.
Setup: FreeBSD 3.4-STABLE, Apache 1.3.20, PHP4.0.5, newest
SybaseOpenClient for FreeBSD (not sure of version, older doesn't work
either), SQLServer 7.0SP3 on WinNT.
Sending a valid query, using mssql_query() works fine, sending an
invalid query (for a
table that does not exist for example), hangs the script. PHP does not
timeout, the browser
does eventually.
For more details, see bug #7881.

This is very important for me to get fixed, would be happy for some
information on why this happens?

Thanks!
Peter





Edit this bug report at http://bugs.php.net/?id=11171&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #10347 Updated: Core Dump when selecting TIMESTAMP with FreeTDS and SQL Server

[lobbin]

ID: 10347
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Feedback
Bug Type: Sybase (dblib) related
Operating System: Linux Kernel 2.2.16
PHP Version: 4.0.4pl1
New Comment:

Can you:
   - try a newer version of freetds
   - try this with PHP 4.1.1

and see if you still have this problem?

Previous Comments:


[2001-06-30 15:07:30] [EMAIL PROTECTED]

This is not related to the mssql extension. You are using the sybase
extension.



[2001-04-16 12:37:21] [EMAIL PROTECTED]

When I try and select a TIMESTAMP column from SQL Server, php core
dumps.   I am using PHP 4.0.4pl1, with FreeTDS 0.51 and SQL Server 7.0
SP1 on kernel 2.2.16, glibc-2.2

Something as simple as:
SELECT dateJoined FROM members;

will cause a core dump

The bad line in add_string_to_string is :
memcpy(result->value.str.val+op1->value.str.len, op2->value.str.val,
op2->value.str.len);

GDB values:
(gdb) print result->value.str.val+op1->value.str.len
$1 = 0x81d3be2 ""
(gdb) print op2->value.str.val
$2 = 0x81d08e4 ""
(gdb) print  op2->value.str.len
$3 = -1

(gdb) bt
#0  0x40791ace in memcpy () from /lib/libc.so.6
#1  0x80eda9e in add_string_to_string (result=0xbfffdaa0,
op1=0xbfffdaa0, op2=0x81c6c2c)
at zend_operators.c:1004
#2  0x8125efe in execute (op_array=0x81c69b4) at ./zend_execute.c:1389
#3  0x812679a in execute (op_array=0x81b7d6c) at ./zend_execute.c:1559
#4  0x80f0ef8 in zend_execute_scripts (type=8, file_count=3) at
zend.c:729
#5  0x8061b18 in php_execute_script (primary_file=0xb6b0) at
main.c:1221
#6  0x805fcf8 in main (argc=2, argv=0xb764) at cgi_main.c:738
#7  0x40727f31 in __libc_start_main (main=0x805f434 , argc=2,
ubp_av=0xb764, 
init=0x805d784 <_init>, fini=0x81309dc <_fini>, rtld_fini=0x4000e274
<_dl_fini>, 
stack_end=0xb75c) at ../sysdeps/generic/libc-start.c:129





Edit this bug report at http://bugs.php.net/?id=10347&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #9402 Updated: never ending cycle

[lobbin]

ID: 9402
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Feedback
Bug Type: Sybase (dblib) related
Operating System: RedHat 7.0
PHP Version: 4.0.4pl1
New Comment:

-> Feedback.

Previous Comments:


[2002-01-07 02:28:11] [EMAIL PROTECTED]

Does this problem occur on PHP 4.1.1?



[2001-06-30 15:06:51] [EMAIL PROTECTED]

This is not related to the mssql extension. You are using the sybase
extension.



[2001-02-22 09:12:18] [EMAIL PROTECTED]

When I call mssql_query and sql server error occured (for example I add
a column to table, but didn't redefine view), then php script doesn't
stop, but instead of that, it will wait forever. I have sybase-ct built
in php (can compile stand-alone module, but can use, don't know why). I
don't  think that in earlier versions (4.x) was this problem, but I am
not sure.

Thanx for help

ojo.






Edit this bug report at http://bugs.php.net/?id=9402&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #9402 Updated: never ending cycle

[lobbin]

ID: 9402
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Status: Open
Bug Type: Sybase (dblib) related
Operating System: RedHat 7.0
PHP Version: 4.0.4pl1
New Comment:

Does this problem occur on PHP 4.1.1?

Previous Comments:


[2001-06-30 15:06:51] [EMAIL PROTECTED]

This is not related to the mssql extension. You are using the sybase
extension.



[2001-02-22 09:12:18] [EMAIL PROTECTED]

When I call mssql_query and sql server error occured (for example I add
a column to table, but didn't redefine view), then php script doesn't
stop, but instead of that, it will wait forever. I have sybase-ct built
in php (can compile stand-alone module, but can use, don't know why). I
don't  think that in earlier versions (4.x) was this problem, but I am
not sure.

Thanx for help

ojo.






Edit this bug report at http://bugs.php.net/?id=9402&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #9244 Updated: sybase-ct does not report MSSQL-server errors

[lobbin]

ID: 9244
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Feedback
Bug Type: Sybase (dblib) related
Operating System: Linux Mandrake 7.1
PHP Version: 4.0.4pl1
New Comment:

Do you still have this problem PHP 4.1.1?

Previous Comments:


[2001-06-30 15:06:07] [EMAIL PROTECTED]

This is not related to the mssql extension. You are using the sybase
extension.



[2001-02-13 15:14:02] [EMAIL PROTECTED]

Bug #7881, as posted by [EMAIL PROTECTED] is still present in
PHP4.0.4pl1. My configuration is: Linux Mandrake 7.1, Apache 1.3.17,
PHP4.0.4pl1, sybase-common-11.9.2.-3.1386.rpm,
sybase-openclient-11.1.1-3.i386.rpm, SQLServer 7.0SP3 on Win2K.
Sending a valid query, using mssql_query() works fine, sending an
invalid query (for a table that does not exist for example), hangs the
script. PHP does not timeout, the browser does eventually.
For more details, see bug #7881.

Thanks,

Arjan





Edit this bug report at http://bugs.php.net/?id=9244&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14567 Updated: exit signal segmentation fault(11)

[lobbin]

ID: 14567
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Old Status: Feedback
Status: Closed
Bug Type: Reproducible crash
Operating System: linux  2.2.16
PHP Version: 4.1.0
New Comment:

No feedback. Closing.

Previous Comments:


[2001-12-17 15:56:17] [EMAIL PROTECTED]

Please provide a short reproducing example, and provide a backtrace. You
can read how to do that via bugs.php.net/how-to-report.php

Derick



[2001-12-17 15:54:19] [EMAIL PROTECTED]

I am not certain of the source of this one, if I load a page there seems
to be about a 50% chance of a segmentation fault with  4.1.0, but works
fine with 4.0.6 using the same configuration.

 './configure' '--with-apxs=/usr/sbin/apxs' '--with-mysql' '--with-gd'
'--with-jpeg' '--with-ttf' '--with-pdflib' '--with-png-dir=/usr/lib'
'--with-tiff-dir=/usr/lib' '--with-jpeg-dir=/usr/lib' '--with-imap'
'--enable-track-vars' '--with-jpeg-dir=/usr/lib/' '--with-zlib'






Edit this bug report at http://bugs.php.net/?id=14567&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14459 Updated: output error

[lobbin]

ID: 14459
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Old Status: Feedback
Status: Closed
Bug Type: MSSQL related
Operating System: Win 2000
PHP Version: 4.1.0
New Comment:

No feedback. Closing.

Previous Comments:


[2001-12-17 16:38:17] [EMAIL PROTECTED]

Can you post a short reproducing script? (With as less as possible
database functions used).

Derick



[2001-12-12 10:52:59] [EMAIL PROTECTED]

all my pages still work fine, i still can query the database but i get
this error at the bottom..

I am using Win 2000, MS-SQL server 7.0, and Apache 1.3.20 


Warning: MS SQL error: Invalid parameter in DB-LIBRARY 
function reference. (severity 11) in Unknown on line 0






Edit this bug report at http://bugs.php.net/?id=14459&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14214 Updated: getcwd() and `pwd` return wrong directory

[lobbin]

ID: 14214
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Old Status: Feedback
Status: Closed
Bug Type: Filesystem function related
Operating System: Linux
PHP Version: 4.0.6
New Comment:

Closed it should be.

Previous Comments:


[2001-12-17 07:14:32] [EMAIL PROTECTED]

No feedback. Closing.



[2001-11-26 02:44:36] [EMAIL PROTECTED]

What is your configure line?

Derick



[2001-11-24 18:47:19] [EMAIL PROTECTED]

I've written a small shell script in PHP, which should return the
current working directory:

-
#!/usr/bin/php -q

-

I've stored the script as /scripts/test.php and on the shell, I'm in
/scripts/Testing.  When I run the script with »../test.php«, it returns
»/scripts« two times.  I'd expect it to return »/scripts/Testing« two
times.

However, HTTP_ENV_VARS['PWD'] and HTTP_SERVER_VARS['PWD'] both contain
the correct (ie. /scripts/Testing) path.





Edit this bug report at http://bugs.php.net/?id=14214&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #12345 Updated: Error in yp_match return value

[lobbin]

ID: 12345
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Old Status: Feedback
Status: Closed
Bug Type: YP/NIS related
Operating System: Solaris 2.7
PHP Version: 4.0.6
New Comment:

No feedback. Closed.

Previous Comments:


[2001-12-17 10:48:35] [EMAIL PROTECTED]

Changes were made between version 4.0.5 and 4.0.6 to display this
warning.

However, the function still returns FALSE, and should work as before if
warning is suppressed via @yp_match() or error_reporting()

R.



[2001-07-24 13:20:27] [EMAIL PROTECTED]

After upgrading to PHP v4.0.6 from v4.0.5 this problem seems to have
started. When using yp_match() the documentation states:

"yp_match() returns the value associated with the passed key out of the
specified map or FALSE"

However, FALSE is *NOT* returned if the key does not exist. Instead, I
get the warning:

"Warning: no such key in map in [my script] on line [my line]"

As a quick fix, I surpress the error with @yp_match(). Has the function
changed or is the documentation wrong?

My config is as follows:
./configure  
--with-apxs=/afs/rcf/apps/apache/sun4x_57/1.3.20/bin/apxs 
--with-config-file-path=/afs/rcf/apps/apache/sun4x_57/1.3.20/conf
--prefix=/afs/rcf/apps/apache/sun4x_57/1.3.20 
--with-pgsql=/afs/rcf/apps/pgsql 
--with-imap=/afs/rcf/admin/local 
--with-ldap=/afs/rcf/admin/local 
--with-xml 
--without-mysql 
--enable-track-vars 
--enable-magic-quotes 
--enable-yp 
--enable-debug 
--disable-short-tags





Edit this bug report at http://bugs.php.net/?id=12345&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #11875 Updated: bad retiurn in yp_first

[lobbin]

ID: 11875
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Old Status: Feedback
Status: Closed
Bug Type: YP/NIS related
Operating System: Solaris 5.5.1
PHP Version: 4.0.6
New Comment:

No feedback. Closed.

Previous Comments:


[2001-12-17 10:53:29] [EMAIL PROTECTED]

Does this work under PHP 4.1.0?


R.



[2001-07-04 07:55:35] [EMAIL PROTECTED]

Hello,

After upgrading PHP server version 4.0.6
, we have notice a problem in the yp_first function
(it was working fine with PHP Version  4.0.5



it returns a value in $key but nothing in $entry[$key]

Ex:

First entry is hedic and the value is  (empty string)

With the ypmatch command in unix shell, it works: 
# ypmatch hedic hosts.byname
x.y.z.w   hedic

Thanks in advance for your help.







Edit this bug report at http://bugs.php.net/?id=11875&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14495 Updated: PHP Windows Binary packaged with old version of GD

[lobbin]

ID: 14495
Updated by: lobbin
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Closed
Bug Type: GD related
Operating System: Windows
PHP Version: 4.1.0
New Comment:

PHP 4.1.1, is now shipped with both gd1 and gd2.

Previous Comments:


[2001-12-13 16:49:20] [EMAIL PROTECTED]

The Windows binary ZIP for PHP 4.0.6 shipped with the GD extension built
against at least GD 2.0.  For whatever reason the 4.1.0 binary contains
the extension built against GD 1.6.2.  This means a whole lot of image
functions that were available to Windows servers in 4.0.6 are suddenly
gone.

Unfortunately I don't have a Windows build environment to put together
my own copy of the extension.





Edit this bug report at http://bugs.php.net/?id=14495&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14898: $HTTP_POST_FILES['uploadedfile']['name'] error

[a1593]

From: [EMAIL PROTECTED]
Operating system: linux slackware8
PHP version:  4.1.1
PHP Bug Type: HTTP related
Bug description:  $HTTP_POST_FILES['uploadedfile']['name'] error

/*
This is a bug fix for rfc1867.c
try to upload file from win2k with name C:\DISK0\»\³\.txt
the original code get name as '.ext' which should be '»\³\.txt'. 
the following modify is necessary to for client using charset=BIG5.
thanks PHP.
lai [EMAIL PROTECTED]
*/
---code modified--
//s = strrchr(filenamebuf, '\\');
for (s=filenamebuf+strlen(filenamebuf)-1;s>filenamebuf;s--) {
 if (*s=='\\' && *(s-1)>0) break;
}
--code modified end --- 
 
-- 
Edit bug report at: http://bugs.php.net/?id=14898&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14886 Updated: Access Violation in error handler with MySQL (ISAPI)

[jlim]

ID: 14886
User updated by: [EMAIL PROTECTED]
Reported By: [EMAIL PROTECTED]
Old Summary: Access Violation in error handler with PEAR DB (ISAPI)
Status: Open
Bug Type: IIS related
Operating System: Win2000
PHP Version: 4.1.1
New Comment:

I have simplified the crash to the following script (note that $obj is
null). The bug does not happen when $testmysql = false.

test start";
if ($testmysql) {
$conn = mysql_connect('localhost','root');
}
$obj->run();

print "test end";

?>

I had to run "ab -n2 -c10 " several times before problems
started to occur on a W2k IIS server receiving no other visitors and had
just been restarted.

This is ApacheBench, Version 1.3c <$Revision: 1.45 $> apache-1.3
Copyright (c) 1996 Adam Twiss, Zeus Technology Ltd,
http://www.zeustech.net/
Copyright (c) 1998-2000 The Apache Group, http://www.apache.org/

Server Software:Microsoft-IIS/5.0
Server Hostname:jaguar
Server Port:80

Document Path:  /php/err.php
Document Length:145 bytes

Concurrency Level:  10
Time taken for tests:   184.950 seconds
Complete requests:  2
Failed requests:556
   (Connect: 0, Length: 556, Exceptions: 0)
Non-2xx responses:  555
Total transferred:  5504927 bytes
HTML transferred:   2847858 bytes
Requests per second:108.14
Transfer rate:  29.76 kb/s received

Connnection Times (ms)
  min   avg   max
Connect:0 022
Processing: 490   466
Total:  490   488


Previous Comments:


[2002-01-06 06:15:53] [EMAIL PROTECTED]

I ran ApacheBench on a PEAR DB script (see below), and IIS (ISAPI mode)
will die eventually. It is so bad that often "iisreset" is unable to
restart IIS. 

The ab command I used:

ab -n1 -c10 

with the following results:

This is ApacheBench, Version 1.3c <$Revision: 1.45 $> apache-1.3
Copyright (c) 1996 Adam Twiss, Zeus Technology Ltd,
http://www.zeustech.net/
Copyright (c) 1998-2000 The Apache Group, http://www.apache.org/

Server Software:Microsoft-IIS/5.0
Server Hostname:jaguar
Server Port:80

Document Path:  /lens/adodb/peartest.php
Document Length:142 bytes

Concurrency Level:  10
Time taken for tests:   269.158 seconds
Complete requests:  1
Failed requests:5272
   (Connect: 0, Length: 5272, Exceptions: 0)
Non-2xx responses:  1868
Total transferred:  1891906 bytes
HTML transferred:   769177 bytes
Requests per second:37.15
Transfer rate:  7.03 kb/s received

Connnection Times (ms)
  min   avg   max
Connect:0 0 5
Processing: 2   268  2771
Total:  2   268  2776


The interesting thing is that the crash only happens when an invalid SQL
statement is entered in the script below(notice the column "badcolumn"
which does not exist). If a valid SQL statement is entered ("badcolumn"
is removed), no crash occurs, and ApacheBench runs fine. 

So I guess it is some problem with PHP's error-handler or MySQL
extension not being thread safe. No dll extensions were installed.
Standard pre-compiled PHP downloaded from php.net was used.

John Lim

=== THE SCRIPT ===

query('select
badcolumn,productid,productname,unitsinstock,unitprice from
products');
while (DB_OK === $rs->fetchInto($fields)) {
$id=$fields[0];
$name=$fields[1];
$unitsinstock=$fields[2];
$unitprice=$fields[3];

print "$id, $name, $unitsinstock, $unitprice";
}

$rs->free();

?>





Edit this bug report at http://bugs.php.net/?id=14886&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14780 Updated: Segmentation fault with "mm",PHP works fine with "files"

[yohgaki]

ID: 14780
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Critical
Status: Closed
Bug Type: Session related
Operating System: RH 7.2
PHP Version: 4.1.1
Assigned To: yohgaki
New Comment:

I haven't changed any, but it stopped segfaulting after cvs udpate.
Fixed in 4.2.0-dev.
Please reopen this report if you still have problem with 4.2.0-dev.

Previous Comments:


[2002-01-06 22:11:10] [EMAIL PROTECTED]

Changed Status to Critical.

There are many problems with external session save handlers.



[2002-01-06 22:01:57] [EMAIL PROTECTED]

Found the cause and work around. (Use of session_module_name() is
causing race) I need to reconsider how to make a real fix for this
segfault :)





[2002-01-06 06:19:55] [EMAIL PROTECTED]

Hi,

Here is my configure statement. Notice there's no pgsql:

./configure --with-mm --enable-inline-optimization --with-zlib
--with-oci8=/u01/app/or
acle/product/8.1.7 --disable-onstatement --enable-ftp --with-apxs 
--enable-shmop --en
able-sysvsem --with-gd


John Lim



[2002-01-02 17:08:20] [EMAIL PROTECTED]

I noticed this problem before your report :) It's not a mm, but pgsql
save handler though. Root of the problem is the same, I guess. 
I'm not 100% sure, yet, but multiple close calls may be the cause.





[2001-12-31 02:12:09] [EMAIL PROTECTED]

A simple script that reproduces the problem with "mm". Works fine for a
while until "mm" breaks under stress:

$abc1\n\n";
?>





The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/?id=14780


Edit this bug report at http://bugs.php.net/?id=14780&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] /ext/session session.c

[Yasuo Ohgaki]

There is yet another problem related to external session save handler 
modules...

It seems session_module_name() causes race and segfaults.
A user reported bug for mm save handler also. (I've noticed
the problem with pgsql save handler before the bug report, though)

Bug #14780

Here is again,

1) Simplest fix would be hard coded solution
2) Use semaphore to lock during session_module_name()
3) Others?

If I fix with 2), I probably make session.c does not compile
under Windows since I don't know Windows. I appreciate if someone
knows both UNIX and Windows could fix this bug.

Is there better solution for this?

FYI, 4.1.x requires session_module_name() to set external
session save handler. (i.e. Cannot set external save handler
in php.ini) ini_set() may work, but I haven't tried it yet.

-- 
Yasuo Ohgaki


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14829 Updated: Session ID is not added if output is compressed with ob_gzhandler

[yohgaki]

ID: 14829
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Assigned
Old Bug Type: Session related
Bug Type: Documentation problem
Operating System: Windows ME
PHP Version: 4.1.0
Old Assigned To: 
Assigned To: yohgaki
New Comment:

Output control would be better type, but change this to Doc problem,
since I'm going to change doc.

BTW, why you use  ob_implicit_flush() with ob_gzhandler?
Anyway, use zlib.compression instead of ob_gzhandler, then output
compression and trans sid would work. Or register ob_gzhandler *before*
starting session. (i.e. The order of output handler regsiteration is
*wrong* and it simply does not work) 



Previous Comments:


[2002-01-03 12:40:03] [EMAIL PROTECTED]

This is the prepend for all of my files:

// Start session
$s_permission = false;
@session_start();
session_register("s_permission");
$fallback = session_name()."=".session_id();

// Start output buffering
ob_start("ob_gzhandler");
ob_implicit_flush(0);



This is the footer:

ob_end_flush();



The script output works, but the session id is not added to any URLs (if
cookies are set off for sessions). When I remove the output buffering
things turn out fine and the SESSION-ID is added to the URLs. I've tried
to add  on each URL, but that didn't help either. I invented a
variable called $fallback, and add this to each URL, but this can't be a
perfect solution, as without output compression this is not necessary.

Thanx
Daniel





Edit this bug report at http://bugs.php.net/?id=14829&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #13579 Updated: variables deglobalized after session_register()

[yohgaki]

ID: 13579
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Assigned
Old Bug Type: Session related
Bug Type: Documentation problem
Operating System: Win2k
PHP Version: 4.0.6
Old Assigned To: 
Assigned To: yohgaki
New Comment:

I'll update doc so that this kind of report will not be submitted to bug
database :)

Previous Comments:


[2001-12-19 23:08:16] [EMAIL PROTECTED]

This is half user's fault
Can we do something to this issue?

To reporter: Could you update PHP version when you test and confirm
problem exists in newer versions? Read "Reference Explained" section
carefully for the reason why this is problem in your script.

http://www.php.net/manual/en/language.references.php



[2001-10-06 16:31:21] [EMAIL PROTECTED]

the following script will print out
unset
set
the first time it is run although it should actually print 'set' twice;
somehow, the variable needs to be globalized again after the session has
been registered.

normal php4 setup with everything compiled-in.
php -v >> 4.0.6devel
-
function test()
{
  global $glb_var;

  session_save_path("e:\\temp");
  session_start();
  session_register('glb_var');

  test2();

  print "glb_var='$glb_var'";

  global $glb_var;
  print "glb_var='$glb_var'";
}

function test2()
{
  global $glb_var;

  $glb_var = 'set';
}

$glb_var = 'unset';
test();





Edit this bug report at http://bugs.php.net/?id=13579&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #13472 Updated: input type=hidden should be in a fieldset if there is one

[yohgaki]

ID: 13472
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Analyzed
Bug Type: Session related
Operating System: Any
PHP Version: 4.0.6


Previous Comments:


[2001-12-05 13:22:53] [EMAIL PROTECTED]

hum... not a bug ? PHP is not rewriting html code well, so I'd call it a
bug :-)

Anyway... any chance to get it fixed soon ? 
That shouldnt be /that/ hard to do, since you just have to write the
input after the first fieldset if there is one, or jst after the form is
there isnt any...



[2001-12-05 10:22:58] [EMAIL PROTECTED]

Not a bug. Making it a feature request.



[2001-09-28 04:12:48] [EMAIL PROTECTED]

When enabling trans-id, php rewrites the forms to add an input
type=hidden. the problem is that, in xhtml strict, inputs have to be in
a fieldset element. when a fieldset element is already here, php still
writes the input type=hidden before the fieldset, causing the page to be
invalid html.
reproducing is easy :


http://www.w3.org/1999/xhtml"; xml:lang="en" lang="en">

PHP Bug with forms and sessions




its valid html, but php will, with transid, make it invalid
by putting this input between the form and the fieldset...






Edit this bug report at http://bugs.php.net/?id=13472&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14880 Updated: _SESSION will NOT automatically register it as a session variable

[yohgaki]

ID: 14880
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Feedback
Bug Type: Session related
Operating System: Debian 3.0 (Woody)
PHP Version: 4.1.0
New Comment:

It works for me with 4.2.0-dev.
What is your configure line and related ini setting?

Previous Comments:


[2002-01-06 07:16:18] [EMAIL PROTECTED]

";
}else{
  echo $_SESSION["foo"];
}
?>



[2002-01-05 21:52:19] [EMAIL PROTECTED]

";
  $_SESSION["iVar"] = 8;
}
function func2(){
  echo "\$_SESSION[iVar] ".$_SESSION[iVar]."";
}
session_start();

if ($_REQUEST["action"] != "page2"){
  func1();
}

func2();

if (session_is_registered("iVar") == FALSE){
  echo "iVar is not registered";
}
echo "page1";
echo "page2";
?>





Edit this bug report at http://bugs.php.net/?id=14880&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14834 Updated: SegFault when passing HTTP_SESSION_VARS

[yohgaki]

ID: 14834
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Analyzed
Bug Type: Session related
Operating System: Debian 3.0 (Woody)
PHP Version: 4.1.0
New Comment:

Confirmed with 4.2.0-dev (2002/1/7). It does segfault with my linux
also.

Previous Comments:


[2002-01-03 18:07:36] [EMAIL PROTECTED]

The following script always produces a segmentation fault on verison
4.1.0:  


Test is complete

The problem goes away if you remove the session array as a parameter to
the foo() function (which is not expecting any parameters anyway).  The
problem also goes away if you remove the "global" declaration inside the
foo() function.  No errors or warning are generated when the function
call occurs.  

Granted, this code snippet looks very odd (why would anyone want to pass
HTTP_SESSION_VARS, for example), but was the quickest (least lines of
code) way to set up the conditions that will always force the error. 
The problem also occurs if you try to pass the new _SESSION array and
still reference the older HTTP_SESSION_VARS in the function.

Reigster_globals is off.  Error_level is set to E_ALL.  Apache version
is 1.3.22-5.








Edit this bug report at http://bugs.php.net/?id=14834&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14497 Updated: PHP causes segfault when session handler=user

[yohgaki]

ID: 14497
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Analyzed
Status: Assigned
Bug Type: Session related
Operating System: FreeBSD 4.4-Stable
PHP Version: 4.1.0, 4-200112131200
Assigned To: yohgaki
New Comment:

I've not committed the fix for this bug yet, but you can work around the
segfault. 

Return string when there is no data to read or failed to write. (i.e.
return '';) User session save handler expect string data, if you return
other than string, it segfualts.


Previous Comments:


[2001-12-21 03:36:49] [EMAIL PROTECTED]

Assigned to myself. By the I updated this bug report, I know the fix,
but I forgot what is was now :(  
I'll work on this after I finish things have to do




[2001-12-19 23:00:15] [EMAIL PROTECTED]

Is this fixed?
Anyone mind if I fix this and commit?
--
Yasuo Ohgaki




[2001-12-14 16:00:04] [EMAIL PROTECTED]

I had already tried out your user handlers (as you can see from the bug
report).  Your handlers weren't causing the crash but were helping in
making the crash happen. (I would guess that the initialization of the
internal data structures from your handlers allowed the invalid "return
false;" pointer to be fubar'd in such a way to cause a segfault.)

Read the bug report, it's all there, including on how I was reproducing
the crash.

Your session handlers have a few problems when there is concurrent
access for the same session id.  (It *DOES* happen, especially with
AvantGo clients, trust me on this one)  You also do not check for
expiration in your session_read.  Since garbage collection doesn't
happen on every single access, there's a possibility that stale data
would get loaded.

Also, since your session handlers aren't mentioned anywhere on the PHP
website under the session documentation, as well as not stressing the
fact that returning false will cause data corruption, it still doesn't
really address the issue.  

Personally I don't think the doing something in a script language should
cause a low-level crash. I believe there was another recent bug dealing
with the xslt extension that explained this issue well: "But PHP
generating nice corefiles is not ok."

At most I think PHP should return an error when the data isn't what was
expected, not segfault, or core, or crash.




[2001-12-14 15:29:01] [EMAIL PROTECTED]

Could you take a look at my user session handlers using PostgreSQL.
You'll see what kind of values should be returned.
Please report the result.

http://www.zend.com/codex.php?id=456&single=1



[2001-12-14 12:57:35] [EMAIL PROTECTED]

After reading other similar problems, looking at PHP/Zend source code
and tracking down all the different crashes, I believe I found the
problem.  I went back to the internal session code and for the past 2
hours or so things are working ok.  (No faults)

My session_read function had a "return false;" when there was no data
available.  Obviously, the lower-level functions treat this "false" as
data and try and manipulate it as such.  (Perhaps even trying to
unserialize it?)

Regardless, switching to "return '';" eliminated the crashes for me.

Whether or not this is some other bug popping up (freeing the memory for
a false incorrectly) or just a documentation issue I really don't know. 
At the very least, since many of the tutorials on user-defined session
handlers have "return false;" (I just checked PHPBuilder and the code
there has it), if there's a way to fix it in the code then great!  If
not, maybe put it in a FAQ somewhere or put it up in the documentation
under the set_handler function.

Comments?




The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/?id=14497


Edit this bug report at http://bugs.php.net/?id=14497&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14566 Updated: Session variables become read only

[yohgaki]

ID: 14566
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Status: Open
Bug Type: Session related
Operating System: Solaris 7
PHP Version: 4.1.0
New Comment:

This comment is work around for the bug.
Do not use session_register(), but use only $HTTP_SESSION_VARS or
$_SESSION to register/unregister session vars. 

To register,
$_SESSION['var'] = 1234;

To unregister
unset($_SESSION['var']);

Then your script should work.

Previous Comments:


[2001-12-18 12:55:34] [EMAIL PROTECTED]

register_globals is on.

I have been trying to narrow everything down to a simple test case, but
have not yet been able to.

It now appears that something acts differently in 4.1.0 that changes the
reference of the global variable to something different than the
HTTP_SESSION_VARS[] variable which did not do so in 4.0.6.

A workaround is to set HTTP_SESSION_VARS[varname] =& GLOBALS[varname]
before the end of the script.



[2001-12-17 15:55:08] [EMAIL PROTECTED]

Can you post your script, and tell use the settings of
register_globals?

Derick



[2001-12-17 15:47:43] [EMAIL PROTECTED]

In PHP 4.1.0 it seems that session variables can only be set once and
will not be updated from subsequent pages (when using a user
save_handler at least).  The same code works fine in PHP 4.0.6.

It seems that the write handler registered with
session_set_save_handler() gets the same session data that was read with
the read handler except that newly set session variables get added.





Edit this bug report at http://bugs.php.net/?id=14566&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #13320 Updated: shmget() fails if running two copies of php4 simultaniously as different users

[yohgaki]

ID: 13320
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Analyzed
Status: Assigned
Bug Type: Session related
Operating System: Linux/Debian/unstable
PHP Version: 4.0.6
Assigned To: yohgaki


Previous Comments:


[2001-12-19 22:42:19] [EMAIL PROTECTED]

This is expected. MM module will have problem you've mentioned.
This behavior will not be changed at least anytime soon...

--
Yasuo Ohgaki



[2001-09-15 13:05:38] [EMAIL PROTECTED]

Even with the session handler set to 'files', the mod_mm MINIT function
gets called, putting a session_mm.sem in /tmp with permissions set to
600. This stops anyone else from running php (such as from the command
line) at the same time.

I see three possible fixes (but I don't know if these would work,
otherwise I'd send you a patch):
1. Don't initialize the shared memory segment if the session save
handler is set to anything but 'mm'
2. If the shared memory system fails, set a flag instead of erroring
out, and have the rest of the shared memory system check this flag and
let them error out when they try to use the shared memory system
3. (almost a kludge) Have the shared memory system look at either the
$TMPDIR environment varible for where it should stick the segment, or
have some other environment varible / configuration option that someone
can set to put the session_mm.sem in someplace other than /tmp





Edit this bug report at http://bugs.php.net/?id=13320&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #13320 Updated: shmget() fails if running two copies of php4 simultaniously as different users

[yohgaki]

ID: 13320
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Status: Analyzed
Bug Type: Session related
Operating System: Linux/Debian/unstable
PHP Version: 4.0.6
Old Assigned To: 
Assigned To: yohgaki
New Comment:

I'll work on this one ;)
This report is actually a dup of 13268, though.

Previous Comments:


[2001-12-19 22:42:19] [EMAIL PROTECTED]

This is expected. MM module will have problem you've mentioned.
This behavior will not be changed at least anytime soon...

--
Yasuo Ohgaki



[2001-09-15 13:05:38] [EMAIL PROTECTED]

Even with the session handler set to 'files', the mod_mm MINIT function
gets called, putting a session_mm.sem in /tmp with permissions set to
600. This stops anyone else from running php (such as from the command
line) at the same time.

I see three possible fixes (but I don't know if these would work,
otherwise I'd send you a patch):
1. Don't initialize the shared memory segment if the session save
handler is set to anything but 'mm'
2. If the shared memory system fails, set a flag instead of erroring
out, and have the rest of the shared memory system check this flag and
let them error out when they try to use the shared memory system
3. (almost a kludge) Have the shared memory system look at either the
$TMPDIR environment varible for where it should stick the segment, or
have some other environment varible / configuration option that someone
can set to put the session_mm.sem in someplace other than /tmp





Edit this bug report at http://bugs.php.net/?id=13320&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #13268 Updated: static path in php and mm

[yohgaki]

ID: 13268
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Analyzed
Status: Assigned
Bug Type: Session related
Operating System: linux
PHP Version: 4.0CVS-2001-09-12
Old Assigned To: 
Assigned To: yohgaki
New Comment:

Assign to myself

Previous Comments:


[2001-12-19 22:37:49] [EMAIL PROTECTED]

Will this be changed?
Any one want me to fix it by adding addtional ini entry?



[2001-09-12 13:57:13] [EMAIL PROTECTED]

php (from 4.0.6 and the latest cvs code), has PS_MM_PATH 
statically defined in ext/session/mod_mm.c.

This, coupled with hardcoding of only an extension in 
libmm.so, allows for a very easy DOS on systems that use 
session management for at least cgi and external binary 
placements for php, and quite possibly module installation.

Is it produceable? sure.. touch /tmp/session_mm.sem (if 
using php and mm)







Edit this bug report at http://bugs.php.net/?id=13268&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #8545 Updated: Session variables not carried forward in Netscape Client

[yohgaki]

ID: 8545
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Status: Analyzed
Bug Type: Session related
Operating System: Linux RH 6.2
PHP Version: 4.0.6RC3
New Comment:

Is this bug outstanding?
To reporter: please update PHP Version. (i.e. Try  with 4.2.0-dev)

Previous Comments:


[2001-06-14 23:12:11] [EMAIL PROTECTED]

Seems like the session_start() doesn set the cookie when
Apache is run in Inetd mode. Reproduced with PHP 4.0.6RC3




[2001-02-28 04:03:15] [EMAIL PROTECTED]

hi,
observation is same even with the latest one.

rgds
Bhaskar



[2001-02-27 18:17:25] [EMAIL PROTECTED]

Does this happen with latest CVS snapshot from http://snaps.php.net/
??

--Jani




[2001-01-16 08:03:21] [EMAIL PROTECTED]

Hi,
an update on this.
This problem is observed if the server type in apache web server is
INETD. If we run apache as standalone it works fine.

can u check this and revert why sessions are not working in netscape if
apache runs as inetd.

rgds
Bhaskar



[2001-01-10 06:59:28] [EMAIL PROTECTED]

can i have the design team views asap.

thanks in advance
Bhaskar



The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/?id=8545


Edit this bug report at http://bugs.php.net/?id=8545&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14780 Updated: Segmentation fault with "mm",PHP works fine with "files"

[yohgaki]

ID: 14780
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Assigned
Status: Critical
Bug Type: Session related
Operating System: RH 7.2
PHP Version: 4.1.1
Assigned To: yohgaki
New Comment:

Changed Status to Critical.

There are many problems with external session save handlers.

Previous Comments:


[2002-01-06 22:01:57] [EMAIL PROTECTED]

Found the cause and work around. (Use of session_module_name() is
causing race) I need to reconsider how to make a real fix for this
segfault :)





[2002-01-06 06:19:55] [EMAIL PROTECTED]

Hi,

Here is my configure statement. Notice there's no pgsql:

./configure --with-mm --enable-inline-optimization --with-zlib
--with-oci8=/u01/app/or
acle/product/8.1.7 --disable-onstatement --enable-ftp --with-apxs 
--enable-shmop --en
able-sysvsem --with-gd


John Lim



[2002-01-02 17:08:20] [EMAIL PROTECTED]

I noticed this problem before your report :) It's not a mm, but pgsql
save handler though. Root of the problem is the same, I guess. 
I'm not 100% sure, yet, but multiple close calls may be the cause.





[2001-12-31 02:12:09] [EMAIL PROTECTED]

A simple script that reproduces the problem with "mm". Works fine for a
while until "mm" breaks under stress:

$abc1\n\n";
?>





[2001-12-31 01:55:44] [EMAIL PROTECTED]

When stress testing my web server with

ab -n5000 -c10 

after about 1000-1500 requests the child processes start to fail.

In my apache error_log I see many entries like the following:

[Fri Dec 28 19:50:06 2001] [notice] child pid 14215 exit signal
Segmentation fault (11)

If I repeat the "ab" test several times, eventually Apache will hang and
I have to restart the web server. Here are the results:

This is ApacheBench, Version 1.3
Copyright (c) 1996 Adam Twiss, Zeus Technology Ltd,
http://www.zeustech.net/
Copyright (c) 1998-1999 The Apache Group, http://www.apache.org/

Benchmarking semut (be patient)...
Server Software:Apache/1.3.20
Server Hostname:semut
Server Port:80

Document Path:  /juris/case.php
fbba51299f
Document Length:85383 bytes

Concurrency Level:  10
Time taken for tests:   1638.945 seconds
Complete requests:  5000
Failed requests:3446
   (Connect: 0, Length: 3446, Exceptions: 0)
Total transferred:  429091301 bytes
HTML transferred:   427029653 bytes
Requests per second:3.05
Transfer rate:  261.81 kb/s received

Connnection Times (ms)
  min   avg   max
Connect:0 9   197
Processing:  1692  3263 11353
Total:   1692  3272 11550

When I switch from "mm" to "files" everything works fine. I have a
suspicion that the bug is related to the fact that I am running on an
SMP processor.









Edit this bug report at http://bugs.php.net/?id=14780&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14780 Updated: Segmentation fault with "mm",PHP works fine with "files"

[yohgaki]

ID: 14780
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Status: Assigned
Bug Type: Session related
Operating System: RH 7.2
PHP Version: 4.1.1
Assigned To: yohgaki
New Comment:

Found the cause and work around. (Use of session_module_name() is
causing race) I need to reconsider how to make a real fix for this
segfault :)



Previous Comments:


[2002-01-06 06:19:55] [EMAIL PROTECTED]

Hi,

Here is my configure statement. Notice there's no pgsql:

./configure --with-mm --enable-inline-optimization --with-zlib
--with-oci8=/u01/app/or
acle/product/8.1.7 --disable-onstatement --enable-ftp --with-apxs 
--enable-shmop --en
able-sysvsem --with-gd


John Lim



[2002-01-02 17:08:20] [EMAIL PROTECTED]

I noticed this problem before your report :) It's not a mm, but pgsql
save handler though. Root of the problem is the same, I guess. 
I'm not 100% sure, yet, but multiple close calls may be the cause.





[2001-12-31 02:12:09] [EMAIL PROTECTED]

A simple script that reproduces the problem with "mm". Works fine for a
while until "mm" breaks under stress:

$abc1\n\n";
?>





[2001-12-31 01:55:44] [EMAIL PROTECTED]

When stress testing my web server with

ab -n5000 -c10 

after about 1000-1500 requests the child processes start to fail.

In my apache error_log I see many entries like the following:

[Fri Dec 28 19:50:06 2001] [notice] child pid 14215 exit signal
Segmentation fault (11)

If I repeat the "ab" test several times, eventually Apache will hang and
I have to restart the web server. Here are the results:

This is ApacheBench, Version 1.3
Copyright (c) 1996 Adam Twiss, Zeus Technology Ltd,
http://www.zeustech.net/
Copyright (c) 1998-1999 The Apache Group, http://www.apache.org/

Benchmarking semut (be patient)...
Server Software:Apache/1.3.20
Server Hostname:semut
Server Port:80

Document Path:  /juris/case.php
fbba51299f
Document Length:85383 bytes

Concurrency Level:  10
Time taken for tests:   1638.945 seconds
Complete requests:  5000
Failed requests:3446
   (Connect: 0, Length: 3446, Exceptions: 0)
Total transferred:  429091301 bytes
HTML transferred:   427029653 bytes
Requests per second:3.05
Transfer rate:  261.81 kb/s received

Connnection Times (ms)
  min   avg   max
Connect:0 9   197
Processing:  1692  3263 11353
Total:   1692  3272 11550

When I switch from "mm" to "files" everything works fine. I have a
suspicion that the bug is related to the fact that I am running on an
SMP processor.









Edit this bug report at http://bugs.php.net/?id=14780&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14897: "unable to fork" - cause & solution (well, at least for me)

[louis]

From: [EMAIL PROTECTED]
Operating system: Windows 2000
PHP version:  4.1.1
PHP Bug Type: Program Execution
Bug description:  "unable to fork" - cause & solution (well, at least for me)

in 4.0.6, popen() would just call CreateProcess() with the specified
command.

this was ok.

in 4.1.0 popen() now prepends "cmd.exe /c " (or "command.com /c " on
win9x).

the problem with this, is that many web admins will disable access for the
web apps to access cmd.exe (eg, MS's IISLockDown does this), and hence even
though the web app may have access to c:\utils\myapp.exe, a command like
exec("c:\utils\myapp.exe") now fails on 4.1.0 but works on on 4.0.6.

also, the hardcoding of cmd.exe should not be done.  it should get it from
the enviroment.

summary:

have commands that prepend the result of an enviroment lookup on "%comspec%
/c " to the front of a command

AND

have commands that dont, so that I can call exec("c:\utils\myapp.exe"); 
without having to enable access to cmd.exe

OR

any interpretation of these issues, that allow for a sensible amount of
flexibilty.

Louis Solomon
www.SteelBytes.com


-- 
Edit bug report at: http://bugs.php.net/?id=14897&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #13665 Updated: session save handler - write not working

[yohgaki]

ID: 13665
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Bogus
Bug Type: Session related
Operating System: Windows 2000
PHP Version: 4.0.6
New Comment:

Please ask support questions to [EMAIL PROTECTED]

Previous Comments:


[2001-10-14 11:43:54] [EMAIL PROTECTED]

Note that because session_write() is called upon request shutdown, you
will never see any output from that function.  If you're expecting to
see debugging output from that function, you're out of luck.  That best
way to output debugging messages from your custom write function is to
write to a file instead of standard out.

In other words, 'echo' and 'print' don't work in a session write
handler.



[2001-10-14 03:30:56] [EMAIL PROTECTED]

Hello,

I have register globals turned off and have a 'user' defined session
handler that stores the sessions in a db.

It appears as though the 'session write' function isn't getting
called.

I have read alot of posts from other having this trouble but no
solutions. There is no mention of anyone working on the problem.

How can I help?

Douglas.





Edit this bug report at http://bugs.php.net/?id=13665&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14896: Multi threaded support

[alan_k]

From: [EMAIL PROTECTED]
Operating system: All
PHP version:  4.0CVS-2002-01-06
PHP Bug Type: Feature/Change Request
Bug description:  Multi threaded support

Couldnt find it in the database so I thought I better add it.

$id = thread_create(function_or_object_method);
thread_stop($id);

* especially so that global vars etc. are available in the other threads..
- without resorting to shm + pcntl

similar to perl?/pike/python etc.

This would be useful for PHP-GTK apps. (albeit not essential) 
- although Its one of those be-carefull features :)
-- 
Edit bug report at: http://bugs.php.net/?id=14896&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #13236 Updated: Cycles in object references are not restored correctly

[yohgaki]

ID: 13236
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Suspended
Bug Type: Session related
Operating System: Linux
PHP Version: 4.0.6
New Comment:

This may not be fixed. Even if it is fixed, it will not be done in near
future.

Previous Comments:


[2001-12-19 22:26:52] [EMAIL PROTECTED]

May be we should change Status to 
Suspended 
or change Type to
Feature request?

--
Yasuo



[2001-10-23 14:05:17] [EMAIL PROTECTED]

== [EMAIL PROTECTED] ==
The references the currently open Bug id #13236.  I have confirmed this
bug.

I tried to add my comments to that bug, but apparently only the creator
of the bug can do
that.

I don't have time to look at the PHP source code, but generally the way
serialization of
objects is done (taken from C++ foundation classes I've worked with) is
on write build a
hash table of object pointers/addresses as key, and a unique id (auto
increment) as value.
 Write the id and class on first instance, and then serialize only the
id on subsequent
references to same object.

Then on read serialization (unserialization), build the same hash table
with the key and
value reversed roles, i.e. the key is the id, and the value is the
pointer to the
object.

AFAIK, this method will support cyclical references.  From the number of
bugs I've seen
reported about storing objects and/or references in sessions, I am
guessing that PHP is
not using this general algorithm.

I've seen other object and/or referenence bugs in sessions in past
releases of PHP
(starting with PHP4.0.3pl1), and haven't had time to go back and test to
see what has been
fixed, what has been broken, and what has not been fixed.  It gets quite
confusing and
arduous to track.

I add this suggestion here if it may be of assistance, as a general way
to kill all these
bugs.  I'd really like to see the serialization of objects and
references work correctly
in a general way, as it is a very powerful tool, vs. using other
bandaids such long
if-else constructs or redundant arrays to achieve the equivalent of
persistant object
references.

Please take my input constructively.
=



[2001-09-10 13:01:17] [EMAIL PROTECTED]

If you use session_register on an object that contains references to
other objects, that each contain a reference to the first object, these
objects will not be restored correctly.

This is a huge limitation in PHP!!!






Edit this bug report at http://bugs.php.net/?id=13236&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #13554 Updated: Storing complex object as session variable makes apache go beserk

[yohgaki]

ID: 13554
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Feedback
Bug Type: Session related
Operating System: Redhat Linux
PHP Version: 4.0.6
New Comment:

Does this happen with PHP 4.1.1?

Previous Comments:


[2001-10-05 03:46:35] [EMAIL PROTECTED]

It should also be noted that I dont see this problem with much smaller /
less complex objects.





[2001-10-05 02:54:51] [EMAIL PROTECTED]

Before this, I've started the session, and have included the file
defining the Test class.

  session_register("testCache");

  if(!isset($testCache))
  {
 $testCache = new Test($testid);
 echo "Loading non cached...";
  }

It should be noted that an instantiated Test object is really pretty
darn hefty and contains arrays of objects in its attributes with many
methods.  Total footprint is probably around 150k or so.

The first time this code is executed, testCache is empty
and is created within the body of the if statement

The second time (reload the page for instance) it causes the apache
process servicing the request to gobble up insane amounts of memory.  As
far as I can tell, apache will never respond back.  On the reload, the
session variable should be set







Edit this bug report at http://bugs.php.net/?id=13554&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14628 Updated: Session functions block on lock

[yohgaki]

ID: 14628
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Feedback
Bug Type: Session related
Operating System: Linux
PHP Version: 4.1.0
New Comment:

I guess your httpd is crashing. Please check your apache log. Do you get
core file?



Previous Comments:


[2001-12-22 06:12:34] [EMAIL PROTECTED]

Hi

No Frames involved.  Thanks 

Chris




[2001-12-21 20:27:46] [EMAIL PROTECTED]

Do you have frames with your php script?
If yes, write session_write_close() to see if it helps.
I don't have time to check mod_files.c now, just trying to find work
around ;)

PHP version 4.0.6 => 4.1.0



[2001-12-21 11:11:31] [EMAIL PROTECTED]

Hi

Have installed 4.1, same problem as highlighted below:
[root@fweb2 /root]# strace -p 6739
flock(4, LOCK_EX


[root@fweb2 /root]# lsof | grep sessions
php6739 fonetastic4u   REG  72,10   163294275
/www/external/sessions/sess_104bb2ffcc67af16af4095cbd92a8f64
php6949 fonetastic4u   REG  72,10   163294275
/www/external/sessions/sess_104bb2ffcc67af16af4095cbd92a8f64



[2001-12-20 19:00:02] [EMAIL PROTECTED]

Could you try 4.1.0 and/or 4.2.0-dev see if it's fixed?
http://www.php.net/downloads.php (4.1.0)
http://snaps.php.net/ (4.2.0-dev)

Please report the result and update PHP version.

--
Yasuo Ohgaki




[2001-12-20 12:40:04] [EMAIL PROTECTED]

When running php with fastcgi, php gets stuck on call to open session
file. Strace output gives following;

open("/www/external/sessions/sess_7cd8fc3921075f6f1483f052619e44e8",O_RDWR)
= 4
flock(4, LOCK_EX






Edit this bug report at http://bugs.php.net/?id=14628&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14798 Updated: session.gc_maxlifetime does not work (Reopen Bug ID #3793)

[yohgaki]

ID: 14798
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Status: Open
Bug Type: Session related
Operating System: Win 2k
PHP Version: 4.1.0
New Comment:

Any windows users can confirm this?

BTW, gc_maxlifetime is in sec. and *explained* explicitly in php.ini
IIRC.

Previous Comments:


[2002-01-02 08:37:15] [EMAIL PROTECTED]

Befor going into the bug-report an importent question:
session.gc_maxlifetime documented as a lifetime messured in *seconds*
with default = 1440 
   1440s = 24min. hmm 24min ?? 
Expectiong the gc_maxlifetime to be rather long 24 min. seams a short
time AND 24 relates more to 24h! So is it realy *seconds* ?!
 ---
The Bug:
As reported in Bug ID #3793 (from [EMAIL PROTECTED]) following still
happens (taken from [EMAIL PROTECTED], 2000-12-07 and veryfied by
[EMAIL PROTECTED], 2001-11-25):
1) "session.gc_maxlifetime" does not work - I set this to 60 sec, but I
can read values from the session even when time expired. (I start the
session and then I wait
more than 60 sec before calling other script)

2) When I set "session.gc_probability = 100"  in php.ini, ALL other
session files are deleted (only one session can be used at the time - if
2 clients are connected to
server, the second client deletes session of the first client, etc.).
.

My Comment:
To (1): This may be intended when using cookies! Because if
session.cookie_lifetime is =0 (until browser is restarted) finding a
valid SID in the cookie may prevent the gc from destroying the session
data. (The doc leves this open).

To (2): For testing I've set "session.gc_probability = 50" but the
effect is the same. As soon as the gc runs, all other session-files are
deleted. gc_maxlifetime has no influance. 

[EMAIL PROTECTED] wrote that it may have to do with 'atime' and I would
think so too. I would consider to use 'mtime' (maybe as fallback). Even
the PHP-manual makes restriktions to 'atime': "Some Unix filesystems can
be mounted with atime updates disabled to increase the performance."


[Session]
session.gc_probability = 100
session.gc_maxlifetime = 60
session.save_handler = files
session.save_path = c:\tmp\
session.use_cookies = 1
session.name = PHPSESSID
session.auto_start = 0
session.cookie_lifetime = 0
session.cookie_path = /
session.cookie_domain =
session.serialize_handler = php
session.referer_check =
session.entropy_length = 0
session.entropy_file =
;session.entropy_length = 16
;session.entropy_file = /dev/urandom
session.cache_limiter = nocache
session.cache_expire = 180
session.use_trans_sid = 1






Edit this bug report at http://bugs.php.net/?id=14798&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14612 Updated: session save handler module uses php_session_register_module() does not work.

[yohgaki]

ID: 14612
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Critical
Status: Closed
Bug Type: Session related
Operating System: Linux
PHP Version: 4.0CVS-2001-12-19
New Comment:

Fixed in 4.2.0-dev

Previous Comments:


[2001-12-20 03:42:11] [EMAIL PROTECTED]

Move the stuff in MINIT to GINIT, and it shold work again.
I was looking into that, but forgot it afterwards.

Derick



[2001-12-19 23:50:37] [EMAIL PROTECTED]

My description is not correct.
There is  a way to use external session save handler module.
i.e. session_module_name()



[2001-12-19 22:01:10] [EMAIL PROTECTED]

Addiotinal info:

Since session module is initilized at first, there is no way to register
external session module. This is critical, I think.

Changed status = Critical.



[2001-12-19 21:46:07] [EMAIL PROTECTED]

Ok. I found the reason and can fix it now.

It try to find session module *before* registering.
I guess it is due to module initilization order has been changed. 
(I could use mm module at leat with 4.0.6, IIRC).

Questin is:
Should it be fixed in session module or
module initilization order should be fixed...





[2001-12-19 21:10:21] [EMAIL PROTECTED]

Starting httpd: Unknown(0) : Fatal error - Cannot find save handler mm
/usr/sbin/apachectl startssl: httpd started

'./configure' \
'--with-apxs' \
'--without-mysql' \
'--without-pear' \
'--enable-debug' \
'--with-mm' \
"$@"

php.ini
session.save_handler=mm

files and user modules works, since they are statically defined in
ps_modules.





Edit this bug report at http://bugs.php.net/?id=14612&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

Re: [PHP-DEV] Re: [PHP-CVS] cvs: php4 /ext/session session.c

[Yasuo Ohgaki]

Mlwmohawk wrote:

> On Wednesday 02 January 2002 04:37 pm, Yasuo Ohgaki wrote:
> 
>>Yasuo Ohgaki wrote:
>>
>>BTW, if I fix the issue with 3), I have to change msession a line
>>so that it does not register it as session save handler. (handler
>>is registered by #ifdef)
>>
>>I have to move pgsql session save handler to php4 source tree,
>>since save handlers must be compiled with session module with
>>option 3).
>>
>>Comments are welcome :)
>>
> 
> A lot of mail comes through this list. Sorry if I cause a rehash. What I 
> don't understand is why there is the need to remove 
> php_session_register_module().


I was not going to remove php_session_register_module()
I'll leave current session.c as is, as a result of discussion
with Derick.

Disableing error message and my fix for crash with external
save handler is enough for now, I think. I can close some
bug reports with these changes :)

PS: When someone implemented smart ini parser/module loader,
the error message should be enabled again.

-- 
Yasuo Ohgaki


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14528 Updated: Conflicting types for 'gdioctx'

[grichards]

ID: 14528
User updated by: [EMAIL PROTECTED]
Reported By: [EMAIL PROTECTED]
Old Status: Closed
Status: Open
Bug Type: Compile Failure
Operating System: Red Hat 7.2
PHP Version: 4.0.6
New Comment:

I just got around to upgrading to 4.1.0, no difference, still can't
compile --with-gd



Previous Comments:


[2002-01-05 16:51:37] [EMAIL PROTECTED]

No feedback. Closing.



[2001-12-14 21:32:37] [EMAIL PROTECTED]

Update to PHP 4.1.0..this should be fixed in it. If not, please 
try the latest CVS snapshot: http://snaps.php.net/

(I'm pretty sure that this is fixed..just can't remember when :)




[2001-12-14 19:27:07] [EMAIL PROTECTED]

Hi. Thanks for the reply!

There is only /usr/include/gd.h on my system:

[root@server1 php-4.0.6]# locate gd.h
/usr/include/gd.h
/usr/local/src/gd-2.0.1/gd.h
/usr/src/php-4.0.6/ext/gd/php_gd.h


Any other ideas?



[2001-12-14 19:20:57] [EMAIL PROTECTED]

This is often due to the fact that you have multiple versions of the gd
header files on your system.  Do a 'locate gd.h' and make sure you only
have 1.  If you have more than 1, get rid of the old bogus ones.



[2001-12-14 18:52:33] [EMAIL PROTECTED]

I'm getting the same bug a lot of other people have posted. I haven't
seen a solution in all the reports that work. I've installed gd 2.0.1.

My configure:

./configure --with-apxs=/usr/local/apache/bin/apxs --with-openssl
--enable-ftp --with-gd=shared,/usr -with-mysql=/usr --with-pdflib
--with-jpeg-dir=/usr/local --with-tiff-dir=/usr/local

results of make:

gd.c:95: conflicting types for `gdIOCtx'
/usr/local/include/gd_io.h:18: previous declaration of `gdIOCtx'
make[3]: *** [gd.slo] Error 1
make[3]: Leaving directory `/usr/src/php-4.0.6/ext/gd'

What does this mean? What can I do?

Gabe





Edit this bug report at http://bugs.php.net/?id=14528&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #13341 Updated: Seffault in outputbuffers

[yohgaki]

ID: 13341
Updated by: yohgaki
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Analyzed
Bug Type: Reproducible crash
Operating System: linux
PHP Version: 4.0.6
Old Assigned To: 
Assigned To: yohgaki
New Comment:

Assign to me so that I don't forget about this.
Zeev may want to fix this, though :)

Previous Comments:


[2002-01-06 16:54:47] [EMAIL PROTECTED]

also true for 4.1.1

Greetz,




[2002-01-06 09:26:14] [EMAIL PROTECTED]

yup still segfaults with 4.1.0



[2002-01-06 07:41:19] [EMAIL PROTECTED]

Does this problem still occur with 4.1.1 and/or the latest CVS?



[2001-09-17 06:20:22] [EMAIL PROTECTED]

Fixed typo;

Btw where did the bug reporters stats go? It looks good to be on top ;)



[2001-09-17 06:14:58] [EMAIL PROTECTED]

Hiya,

Using the same var from the function and changing will sefault, copying
it first to another var will work... but should not happen, see
examples

I segfaulted :( :( :(

Greetz,

Wico de Leeuw





Edit this bug report at http://bugs.php.net/?id=13341&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Re: [PHP-CVS] cvs: php4 /ext/session session.c

[Yasuo Ohgaki]

[EMAIL PROTECTED] wrote:

> On Thu, 3 Jan 2002, Yasuo Ohgaki wrote:
> 
> 
>>>Which is perfectly fine as it is IMO. What is the change here that you
>>>suggest exactly?
>>>
>>
>>- Do not use php_session_register_module()
>>- Add #ifdef to ps_modules struct for external save
>>   handler modules
>>
> 
> You want to add the pgsql module here too then?


Depends, since there is no real users and there is no
way to download other than check out from CVS :)

> 
> 
>>Just a few lines of changes, but it requries to compile save handler
>>modules with session module.
>>
>>
> 
> I'm against removing the php_session_register_module() call, for the
> reason that I need it for SRM. Sascha implemented it because if this
> (maybe partly) reason. If this call is removed, or deprecated, there is no
> way for an extension to register a session module, without modifying the
> PHP Source, which is a bad idea IMO.


php_session_register_module() should be there.
I'm proposing temp work around for known problem, since real fix would
take much longer time to implement.

> 
> 
>>>You still failed to tell me why this is all needed. If it's only the
>>>runtime undefined symbol stuff, then I think it's nto worth the hassle to
>>>'fix' it. It's not broken now. And I think it's not that big a problem at
>>>all.
>>>
>>That's why I just get rid of error message for now.
>>
>>Should I leave it as is?
>>
> 
> That's my opinion here.


Ok. Then I just leave the session.c as is now. (No error message for
bogus session save handler name. All external save handlers are bogus
when module is initilized :)

I've changed session.c so that it won't segfault under certain
condition when external session save handler is used. These
changes should be enough for now.

-- 
Yasuo Ohgaki


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14895: Error when trying to load the php4 module

[snowgod]

From: [EMAIL PROTECTED]
Operating system: FreeBSD 4.3-RELEASE
PHP version:  4.1.1
PHP Bug Type: OpenSSL related
Bug description:  Error when trying to load the php4 module

Here is my configuration when compiling php

./configure  --with-mysql --enable-track-vars --with-snmp --with-imap
--with-apxs=/usr/local/apache/bin/apxs --enable-ucd-snmp-hack
--with-imap-ssl --with-openssl=/usr --enable-sockets --enable-ftp
--with-bz2 --with-zlib

My apache version is 1.3.20 (I tried with 1.3.22 as well and had the same
issue.

The error I get when attempting to start the apache server (compiles fine)
is:

Cannot load /usr/local/libexec/apache/libphp4.so into server:
/usr/local/libexec/apache/libphp4.so: Undefined symbol "ssl_onceonlyinit"

libssl exists:
(root@warped) [/usr/lib]$ ls -ld /usr/lib/libssl.so
lrwxrwxrwx  1 root  wheel  11 Sep  7 17:59 /usr/lib/libssl.so ->
libssl.so.2
(root@warped) [/usr/lib]$ ls -ld /usr/lib/libssl.so.2
-r--r--r--  1 root  wheel  176348 Apr 21  2001 /usr/lib/libssl.so.2

I can post a list of apache modules if nessicary, but that doesn't seem to
matter, I do use mod_ssl though.

-- 
Edit bug report at: http://bugs.php.net/?id=14895&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14894 Updated: socket_read should return string, not int

[junk-php]

ID: 14894
User updated by: [EMAIL PROTECTED]
Reported By: [EMAIL PROTECTED]
Status: Open
Bug Type: Documentation problem
Operating System: Linux
PHP Version: 4.1.0
New Comment:

Examples in mainline socket text should also be updated, they use the
bogus version. 

Previous Comments:


[2002-01-06 18:07:39] [EMAIL PROTECTED]

http://www.php.net/manual/en/function.socket-read.php

Claims to return an int. The prototype should be string socket_read. 

- AZ






Edit this bug report at http://bugs.php.net/?id=14894&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Re: new bug viewing/editing form

[Jim Winstead]

Manuel Lemos <[EMAIL PROTECTED]> wrote:
> Jim Winstead wrote:
>> > - Voting does not work on Opera 5.05 for Linux. When you submit the
>> > vote, it fails with the message "missing parameter score".
>> 
>> it's a bug in opera. http://bugs.php.net/~jimw/bugs/opera-broken.php
>> 
>> but because of the order of the fields on the new page, this bug doesn't
>> get triggered there.
> 
> What do you mean? I just noticed that it still causes the problem above.

i mean there is a bug in opera that causes voting to fail on the
existing bug viewing/voting/editing page (where the voting stuff is in a
box that is floated to the right of the bug header). in my testing, the
bug in opera is no longer triggered by the new version of this page
(where the voting form is below the bug header).

regardless of whether the new version triggers this bug in opera or not,
it remains a bug in opera. i'm not going to spend any more of my time
figuring out how to work around it.

>> i really, really do not want to force people to register and deal with
>> all that. if abuse of the voting system becomes a problem, we'll figure
>> out a way to deal with it at that time. (the voting does track the ip of
>> submissions, so that is one easy filter that can be applied.)
> 
> hummm... I was able to vote on the same bug several times in a row and
> it counted. Are you sure that IP filter is working?

there is no filter. the ip is only being tracked at this time.

> Anyway, that is easy to work around, especially because the source of
> the page is available, and who intends to fool the system will always
> find a easy way to do it.

of course. that's why i am sticking with the path of least resistance.

if the existing bug-voting system proves to not be useful (because of
people stuffing the ballot box, or not voting, or whatever), we can
always change it (slightly or dramatically). but until that happens, i
see very little reason to change it.

>> i have my doubts that the voting will really prove all that useful in
>> cajoling developers to work on specific bugs. i suspect it will be more
> 
> I think it will help the QA time to concentrate on bugs that seem more
> urgent, if the vote system can work reliably.

my doubts come from a long time of watching developers of php work on
what they want to work on, without a great deal of concern for what it
is that users want. i don't see any reason to believe requiring people
to register to vote on bugs will change that situation.

jim

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14894: socket_read should return string, not int

[junk-php]

From: [EMAIL PROTECTED]
Operating system: Linux
PHP version:  4.1.0
PHP Bug Type: Documentation problem
Bug description:  socket_read should return string, not int

http://www.php.net/manual/en/function.socket-read.php

Claims to return an int. The prototype should be string socket_read. 

- AZ

-- 
Edit bug report at: http://bugs.php.net/?id=14894&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

RE: [PHP-DEV] Re: new bug viewing/editing form

[Melvyn Sopacua]

Lukas Smith said at 23:37 6-1-2002:

>Anyways I guess the important thing is to think about the audience you
>are trying to get to report bugs.

That wouldn't necessarily be the same people, who you need for a vote.
The audience for a vote, would be to get an indication of the affected
sites (would have been wonderful to have that available during the whole
exit() discussion) and to get an insight on how important your usergroup
perceives the issue.

For instance obsoleting mysql_db_query from the next release, would be a
good test :-)

This would also make descisions for a maintenance release have some more
measurable criteria.

That people also provide feedback, is a plus. And this can now be done,
because not only the original reporter of a bug can post feedback, but
anybody. That is however a seperate plus to the new bug system, besides
the voting improvements.


Met vriendelijke groeten / With kind regards,

IDG.nl
Melvyn Sopacua
Webmaster


XML-error: undefined entity "peace" at line 1.



-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14893: RE starting with (.*) might break

[japhy]

From: [EMAIL PROTECTED]
Operating system: SunOS
PHP version:  4.1.1
PHP Bug Type: PCRE related
Bug description:  RE starting with (.*) might break

Here's the problem:



It fails, but it really shouldn't.  You can fool the engine into not having
the bug:



The bug is thus:  a regex that starts with .* can logically be made to
start with an implicit anchor to the beginning of the string.  However,
this optimization can break the success of a regex if the .* is captured
(as above) and used later (the back-reference \1).  I've contacted the
author of the PCRE package.
-- 
Edit bug report at: http://bugs.php.net/?id=14893&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

Re: [PHP-DEV] Re: new bug viewing/editing form

[Melvyn Sopacua]

Hi,

>Bugzilla implements that kind of authentication to avoid reports from
>faked people that do not intend to provide feedback later.

Which is what you want, for a good voting system.
You want users, who took the time to search through the active database,
to cast their vote, so developers get some idea of the impact of the
bug.
Registration is a 2-step (do + confirm) boundery that takes away a lot
of the potential voters, for reasons ranging from lack of anti-paranoia
medication to time constraints and Jolt overdose.

> > Floods can be easily detected, and bugs which outstand tremendously can
> > be investigated.
>
>The problem to avoid here are not exactly floods, but the occasional
>user that in the hope of having his reports addressed with greater
>priority would submit say like 12 votes. How would you tell if those
>votes would be from the same user or not?

You wouldn't and an IP/cookie based session, would take care of most of
that. If somebody actually takes to the time to use multiple IP addresses,
and browsers to cast a not too suspiscious amount of votes over a lenghtly
period of time, let's just reward this frustration, before the poor guy
jumps off some building.

No matter what procedural importance you award to the voting tool, it will
be one of the factors, which stimulates a (group of) developers to
work on a bug.

Met vriendelijke groeten / With kind regards,

IDG.nl
Melvyn Sopacua
Webmaster


XML-error: undefined entity "peace" at line 1.



-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

RE: [PHP-DEV] Re: new bug viewing/editing form

[Lukas Smith]

I have seen this used (doesn't even Yahoo use this?)
Anyways I guess the important thing is to think about the audience you
are trying to get to report bugs.

What is important to get anything fixed? The initial Report and then
Feedback!
What slows down fixes: Bug reports without Feedback (as you are likely
to waste a lot of time)
So how to best get it this worked out?
1) Get people that actually care to have this bug fixed.
2) Have a method to contact the person for additional feedback.
3) Make the people think before they report if this really is a bug
4) Atleast some level of competence

The down side of course is that registering might turn off some people.
But those do not fall in 1), so we can maybe live without their reports
to begin with? So the downside is actually its upside!
It ensures that you get the type of person that fit 1) and enables 2).

Since you have to provide your email address you also might take more
care as to your quality. This is just a mind thing and obviously you can
register using any email address that you normally do not use at all.
But I think requiring registration still helps a bit with 3).

So for 1) - 3) registration will help.

For 4) ... well registration is not much of a test :-)
But 2,5 out of 4 is quite good :-) 

Obviously the people that fit 1) potentially might also have an interest
in manipulating the votes.

This also helps stop multiple votes (does not make it impossible though)

Best regards,
Lukas Smith
[EMAIL PROTECTED]
___
 DybNet Internet Solutions GbR
 Alt Moabit 89
 10559 Berlin
 Germany
 Tel. : +49 30 83 22 50 00
 Fax : +49 30 83 22 50 07
 www.dybnet.de [EMAIL PROTECTED]
___

> -Original Message-
> From: Manuel Lemos [mailto:[EMAIL PROTECTED]]
> Sent: Sunday, January 06, 2002 11:20 PM
> To: [EMAIL PROTECTED]
> Subject: Re: [PHP-DEV] Re: new bug viewing/editing form
> 
> Hello,
> 
> Melvyn Sopacua wrote:
> >
> > Manuel Lemos said at 22:18 6-1-2002:
> >
> > >- Assuming that the number of votes may influence in the priority
that
> > >developers will give to fix each bug, it seems easy to mislead
> > >developers because somebody that realizes that may submit a bunch
of
> > >votes just to make it outstand in the pending bug queue. I suggest
that
> > >you adopt an authentication scheme like bugzilla, that requires
> > >submitters to subscribe confirming the subscriptions by e-mail.
This
> > >way, multiple votes from the same subscriber would only count as
one.
> >
> > If you go down that road, instead of assuming mature users, and
kicking
> > out the occasional kid, you'll end up with a lot of maintenance in
the
> > the subscriber database, deleting numerous hotmail accounts.
> 
> I think that bug reporters and people that know and understand the
point
> of the bug database are mature users.
> 
> Bugzilla implements that kind of authentication to avoid reports from
> faked people that do not intend to provide feedback later.
> 
> 
> 
> > Floods can be easily detected, and bugs which outstand tremendously
can
> > be investigated.
> 
> The problem to avoid here are not exactly floods, but the occasional
> user that in the hope of having his reports addressed with greater
> priority would submit say like 12 votes. How would you tell if those
> votes would be from the same user or not?
> 
> 
> 
> > There's also a much easier authentication for votes, which separates
> > humans from bots, described here:
> > http://www.webtechniques.com/archives/2001/12/perl/
> 
> Interesting. This would be better than it is now, despite it would not
> avoid the problem.
> 
> Regards,
> Manuel Lemos
> 
> --
> PHP Development Mailing List 
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
> To contact the list administrators, e-mail:
[EMAIL PROTECTED]



-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Re: new bug viewing/editing form

[Manuel Lemos]

Hello,

Jim Winstead wrote:
> > - Voting does not work on Opera 5.05 for Linux. When you submit the
> > vote, it fails with the message "missing parameter score".
> 
> it's a bug in opera. http://bugs.php.net/~jimw/bugs/opera-broken.php
> 
> but because of the order of the fields on the new page, this bug doesn't
> get triggered there.

What do you mean? I just noticed that it still causes the problem above.

 
> > - Assuming that the number of votes may influence in the priority that
> > developers will give to fix each bug, it seems easy to mislead
> > developers because somebody that realizes that may submit a bunch of
> > votes just to make it outstand in the pending bug queue. I suggest that
> > you adopt an authentication scheme like bugzilla, that requires
> > submitters to subscribe confirming the subscriptions by e-mail. This
> > way, multiple votes from the same subscriber would only count as one.
> 
> i really, really do not want to force people to register and deal with
> all that. if abuse of the voting system becomes a problem, we'll figure
> out a way to deal with it at that time. (the voting does track the ip of
> submissions, so that is one easy filter that can be applied.)

hummm... I was able to vote on the same bug several times in a row and
it counted. Are you sure that IP filter is working?

Anyway, that is easy to work around, especially because the source of
the page is available, and who intends to fool the system will always
find a easy way to do it.


 
> i have my doubts that the voting will really prove all that useful in
> cajoling developers to work on specific bugs. i suspect it will be more

I think it will help the QA time to concentrate on bugs that seem more
urgent, if the vote system can work reliably.


Regards,
Manuel Lemos

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

Re: [PHP-DEV] Re: new bug viewing/editing form

[Manuel Lemos]

Hello,

Melvyn Sopacua wrote:
> 
> Manuel Lemos said at 22:18 6-1-2002:
> 
> >- Assuming that the number of votes may influence in the priority that
> >developers will give to fix each bug, it seems easy to mislead
> >developers because somebody that realizes that may submit a bunch of
> >votes just to make it outstand in the pending bug queue. I suggest that
> >you adopt an authentication scheme like bugzilla, that requires
> >submitters to subscribe confirming the subscriptions by e-mail. This
> >way, multiple votes from the same subscriber would only count as one.
> 
> If you go down that road, instead of assuming mature users, and kicking
> out the occasional kid, you'll end up with a lot of maintenance in the
> the subscriber database, deleting numerous hotmail accounts.

I think that bug reporters and people that know and understand the point
of the bug database are mature users.

Bugzilla implements that kind of authentication to avoid reports from
faked people that do not intend to provide feedback later.


 
> Floods can be easily detected, and bugs which outstand tremendously can
> be investigated.

The problem to avoid here are not exactly floods, but the occasional
user that in the hope of having his reports addressed with greater
priority would submit say like 12 votes. How would you tell if those
votes would be from the same user or not?


 
> There's also a much easier authentication for votes, which separates
> humans from bots, described here:
> http://www.webtechniques.com/archives/2001/12/perl/

Interesting. This would be better than it is now, despite it would not
avoid the problem.

Regards,
Manuel Lemos

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Re: new bug viewing/editing form

[Jim Winstead]

Manuel Lemos <[EMAIL PROTECTED]> wrote:
> Jim Winstead wrote:
>> as threatened after i implemented bug voting, i've redesign the bug
> 
> Congratulations! Voting is an excellent idea that will help developers
> to sort out the priorities of what bugs should be fixed first. Just a
> couple of comments:

the voting has been there for a week or so, actually. it is just hard to
notice on the current page because wide comments cause it to get pushed
way off to the right.

> - Voting does not work on Opera 5.05 for Linux. When you submit the
> vote, it fails with the message "missing parameter score".

it's a bug in opera. http://bugs.php.net/~jimw/bugs/opera-broken.php

but because of the order of the fields on the new page, this bug doesn't
get triggered there.

> - Assuming that the number of votes may influence in the priority that
> developers will give to fix each bug, it seems easy to mislead
> developers because somebody that realizes that may submit a bunch of
> votes just to make it outstand in the pending bug queue. I suggest that
> you adopt an authentication scheme like bugzilla, that requires
> submitters to subscribe confirming the subscriptions by e-mail. This
> way, multiple votes from the same subscriber would only count as one.

i really, really do not want to force people to register and deal with
all that. if abuse of the voting system becomes a problem, we'll figure
out a way to deal with it at that time. (the voting does track the ip of
submissions, so that is one easy filter that can be applied.)

i have my doubts that the voting will really prove all that useful in
cajoling developers to work on specific bugs. i suspect it will be more
useful in getting users to notice bugs that have already been reported.
(and it is a safety valve for preventing 'me too!' comments now that
anyone will be able to add a comment to a bug.)

jim

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #13341 Updated: Seffault in outputbuffers

[wico]

ID: 13341
User updated by: [EMAIL PROTECTED]
Reported By: [EMAIL PROTECTED]
Status: Open
Bug Type: Reproducible crash
Operating System: linux
PHP Version: 4.0.6
New Comment:

also true for 4.1.1

Greetz,


Previous Comments:


[2002-01-06 09:26:14] [EMAIL PROTECTED]

yup still segfaults with 4.1.0



[2002-01-06 07:41:19] [EMAIL PROTECTED]

Does this problem still occur with 4.1.1 and/or the latest CVS?



[2001-09-17 06:20:22] [EMAIL PROTECTED]

Fixed typo;

Btw where did the bug reporters stats go? It looks good to be on top ;)



[2001-09-17 06:14:58] [EMAIL PROTECTED]

Hiya,

Using the same var from the function and changing will sefault, copying
it first to another var will work... but should not happen, see
examples

I segfaulted :( :( :(

Greetz,

Wico de Leeuw





Edit this bug report at http://bugs.php.net/?id=13341&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

Re: [PHP-DEV] Re: new bug viewing/editing form

[Melvyn Sopacua]

Manuel Lemos said at 22:18 6-1-2002:

>- Assuming that the number of votes may influence in the priority that
>developers will give to fix each bug, it seems easy to mislead
>developers because somebody that realizes that may submit a bunch of
>votes just to make it outstand in the pending bug queue. I suggest that
>you adopt an authentication scheme like bugzilla, that requires
>submitters to subscribe confirming the subscriptions by e-mail. This
>way, multiple votes from the same subscriber would only count as one.

If you go down that road, instead of assuming mature users, and kicking
out the occasional kid, you'll end up with a lot of maintenance in the
the subscriber database, deleting numerous hotmail accounts.

Floods can be easily detected, and bugs which outstand tremendously can
be investigated.

There's also a much easier authentication for votes, which separates
humans from bots, described here:
http://www.webtechniques.com/archives/2001/12/perl/


Met vriendelijke groeten / With kind regards,

IDG.nl
Melvyn Sopacua
Webmaster


XML-error: undefined entity "peace" at line 1.



-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Re: new bug viewing/editing form

[Manuel Lemos]

Hello,

Jim Winstead wrote:
> 
> as threatened after i implemented bug voting, i've redesign the bug

Congratulations! Voting is an excellent idea that will help developers
to sort out the priorities of what bugs should be fixed first. Just a
couple of comments:

- Voting does not work on Opera 5.05 for Linux. When you submit the
vote, it fails with the message "missing parameter score".

- Assuming that the number of votes may influence in the priority that
developers will give to fix each bug, it seems easy to mislead
developers because somebody that realizes that may submit a bunch of
votes just to make it outstand in the pending bug queue. I suggest that
you adopt an authentication scheme like bugzilla, that requires
submitters to subscribe confirming the subscriptions by e-mail. This
way, multiple votes from the same subscriber would only count as one.

Keep up the good work,
Manuel Lemos

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

Re: [PHP-DEV] new bug viewing/editing form

[Markus Fischer]

On Sun, Jan 06, 2002 at 12:13:21PM -0800, Jim Winstead wrote : 
>  http://bugs.php.net/~jimw/bugs/bug.php?id=14841

When I click on 'Edit submission' my CVS password is exposed
in clear text ...

-- 
Please always Cc to me when replying to me on the lists.

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

RE: [PHP-DEV] new bug viewing/editing form

[Chris Newbill]

Just a nit-picky thing but maybe you should have the year with all of
the dates.

-Chris

-Original Message-
From: Jim Winstead [mailto:[EMAIL PROTECTED]] 
Sent: Sunday, January 06, 2002 1:13 PM
To: [EMAIL PROTECTED]
Subject: [PHP-DEV] new bug viewing/editing form

as threatened after i implemented bug voting, i've redesign the bug
viewing/editing form. thanks to some of the fine folks from the php-qa
list, it's already been tested against most common browsers, and enough
people complained about the netscape 4 rendering that i applied some
tweaks that have made it more tolerable.

unless i hear some loud screams, i'll go ahead and commit this in a day
or two.

(the rest of this mail is the same as what i posted to php-qa a few days
ago, for the benefit of php-dev folks.)

you can see the new look here:

 http://bugs.php.net/~jimw/bugs/bug.php?id=14841

(that's just a test bug i opened which you can feel free to add comments
to and otherwise mess with. don't worry, this version is hardcoded to
just send the mails to me for now.)

screenshots from mozilla 0.9.7 on linux, so you have some idea of how it
is supposed to look:

  viewing a bug: http://bugs.php.net/~jimw/view.png

adding a comment: http://bugs.php.net/~jimw/add-comment.png

(oh, did i forget to mention anyone can add comments to a bug? :)

  changing the bug (developer, logged in):
http://bugs.php.net/~jimw/developer-logged-in.png
  changing the bug (developer, not logged in):
http://bugs.php.net/~jimw/developer-not-logged-in.png

  changing the bug (submitter, logged in):
http://bugs.php.net/~jimw/user-logged-in.png
  changing the bug (submitter, not logged in):
http://bugs.php.net/~jimw/user-not-logged-in.png

  the success page from any submission (modulo different text):
http://bugs.php.net/~jimw/updated.png

i also tweaked the submission form just slightly to change how error
messages were displayed. you can see a sample here:

http://bugs.php.net/~jimw/incomplete-bug.png

let me know what you think, or if what you see doesn't roughly
approximate what it is that i see. (i know that opera's fieldsets don't
look as cool as mozilla's, so the voting form looks less pretty on
opera.)

jim

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

Re: [PHP-DEV] PHP Extensions: Creating objects

[Markus Fischer]

On Sun, Jan 06, 2002 at 02:04:33PM -0600, Christopher Brown-Floyd wrote : 
> [...]. I've read the PHP docs [...]

Take a look at the Zend API docs at zend.com/apidoc , more
specific at the chapters creating variables / objects .

> and looked at the code
> for other extensions, but I haven't seen any examples of this being done.

You must have missed extensions like domxml then. There are a
few others too using objects.

-- 
Please always Cc to me when replying to me on the lists.

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] PHP Extensions: Creating objects

[Christopher Brown-Floyd]

I want convert my template script into an extension for PHP, but I'm lost on
how to do it so that it results in a user-visible object and not a mass of
functions. An example of the end-result can be found at:
http://www.ultratemplate.com/. I've read the PHP docs and looked at the code
for other extensions, but I haven't seen any examples of this being done.
Can someone enlighten me?

TIA,
Chris



-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

RE: [PHP-DEV] new bug viewing/editing form

[Richard Heyes]

> as threatened after i implemented bug voting, i've redesign the bug
> viewing/editing form. thanks to some of the fine folks from the php-qa
> list, it's already been tested against most common browsers, and enough
> people complained about the netscape 4 rendering that i applied some
> tweaks that have made it more tolerable.
> 
> unless i hear some loud screams, i'll go ahead and commit this in a day
> or two.
> 
> (the rest of this mail is the same as what i posted to php-qa a few days
> ago, for the benefit of php-dev folks.)
> 
> you can see the new look here:

The link colours could do with being more distinct.

-- 
Richard Heyes
"If you have any trouble sounding condescending,
find a Unix user to show you how it's done." - Scott Adams 

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] new bug viewing/editing form

[Jim Winstead]

as threatened after i implemented bug voting, i've redesign the bug
viewing/editing form. thanks to some of the fine folks from the php-qa
list, it's already been tested against most common browsers, and enough
people complained about the netscape 4 rendering that i applied some
tweaks that have made it more tolerable.

unless i hear some loud screams, i'll go ahead and commit this in a day
or two.

(the rest of this mail is the same as what i posted to php-qa a few days
ago, for the benefit of php-dev folks.)

you can see the new look here:

 http://bugs.php.net/~jimw/bugs/bug.php?id=14841

(that's just a test bug i opened which you can feel free to add comments
to and otherwise mess with. don't worry, this version is hardcoded to
just send the mails to me for now.)

screenshots from mozilla 0.9.7 on linux, so you have some idea of how it
is supposed to look:

  viewing a bug: http://bugs.php.net/~jimw/view.png

adding a comment: http://bugs.php.net/~jimw/add-comment.png

(oh, did i forget to mention anyone can add comments to a bug? :)

  changing the bug (developer, logged in):
http://bugs.php.net/~jimw/developer-logged-in.png
  changing the bug (developer, not logged in):
http://bugs.php.net/~jimw/developer-not-logged-in.png

  changing the bug (submitter, logged in):
http://bugs.php.net/~jimw/user-logged-in.png
  changing the bug (submitter, not logged in):
http://bugs.php.net/~jimw/user-not-logged-in.png

  the success page from any submission (modulo different text):
http://bugs.php.net/~jimw/updated.png

i also tweaked the submission form just slightly to change how error
messages were displayed. you can see a sample here:

http://bugs.php.net/~jimw/incomplete-bug.png

let me know what you think, or if what you see doesn't roughly
approximate what it is that i see. (i know that opera's fieldsets don't
look as cool as mozilla's, so the voting form looks less pretty on
opera.)

jim

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] CVS Account Request: flo

[florian geiges]

for translating doc's

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14892 Updated: Download doesn't work

[mfischer]

ID: 14892
Updated by: mfischer
Reported By: [EMAIL PROTECTED]
Old Summary: Fix your shit
Status: Bogus
Bug Type: Unknown/Other Function
Operating System: 
PHP Version: 4.1.1
New Comment:

Try a mirror if you have troubles accessing php.net directly.

Previous Comments:


[2002-01-06 13:41:32] [EMAIL PROTECTED]

I wish you happy new year, and now get lost.

Derick



[2002-01-06 13:40:36] [EMAIL PROTECTED]

yopur download links stop trnxfr in the middle of the items lets get
this resolved so that i can user your site





Edit this bug report at http://bugs.php.net/?id=14892&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14892 Updated: Fix your shit

[derick]

ID: 14892
Updated by: derick
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Bogus
Bug Type: Unknown/Other Function
Operating System: 
PHP Version: 4.1.1
New Comment:

I wish you happy new year, and now get lost.

Derick

Previous Comments:


[2002-01-06 13:40:36] [EMAIL PROTECTED]

yopur download links stop trnxfr in the middle of the items lets get
this resolved so that i can user your site





Edit this bug report at http://bugs.php.net/?id=14892&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14892: Fix your shit

[Usuck]

From: [EMAIL PROTECTED]
Operating system: 
PHP version:  4.1.1
PHP Bug Type: Unknown/Other Function
Bug description:  Fix your shit

yopur download links stop trnxfr in the middle of the items lets get this
resolved so that i can user your site
-- 
Edit bug report at: http://bugs.php.net/?id=14892&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #9266 Updated: Unable to load 14 of php's extensions

[irc-html]

ID: 9266
Updated by: irc-html
Reported By: [EMAIL PROTECTED]
Status: Analyzed
Old Bug Type: Documentation problem
Bug Type: Feature/Change Request
Operating System: WinMe, Win98, Win2000
PHP Version: 4.0.4pl1
New Comment:

The extensions are all (mostly) listed in a table at
http://www.php.net/manual/en/install.windows.php#install.windows.extensions
which also lists if the extension is dependent upon a specific library
(another dll).

Most of the extension documentation also lists 'You need xxx to use this
extension', or 'In order to use this extension you need to install xxx',
etc.  I think it is clearly documented in the manual.  A possibility
would be to improve the error message, perhaps 'php_xxx.dll:  This
extension requires xxx library installed', something simliar would be
helpful.

Changing to feature request.

Previous Comments:


[2002-01-06 13:18:40] [EMAIL PROTECTED]

so better documentation is needed for extensions that rely on external
libraries.



[2001-02-14 20:07:11] [EMAIL PROTECTED]

Yes, a more detailed error message, or/and two more lines in the manual
could have saved me some time. And maybe it would for others too.

Thank you.



[2001-02-14 16:49:32] [EMAIL PROTECTED]

all of theese extensions depend on other DLLs 
being installed on your system
like, for example, sybase_ct client DLLs
(this is the native sybase call interface, _not_ the ODBC stuff)

maybe we could improve the error message?
(moved to feature request)



[2001-02-14 16:49:10] [EMAIL PROTECTED]

all of theese extensions depend on other DLLs 
being installed on your system
like, for example, sybase_ct client DLLs
(this is the native sybase call interface, _not_ the ODBC stuff)

maybe we could improve the error message?
(moved to feature request)



[2001-02-14 16:41:08] [EMAIL PROTECTED]

I spent almost 2 day trying to figure out how to load the sybase
extension under windows. I tried everything I could, but 14 of the
extensions would not load under WinMe, Win98, Win2000, with any
combination of apache, pws, iis. The extension_dir is just set all fine,
and the rest of the extensions load and function well.

I only wanted to get some data from a sybase db. I could actually solve
it with ODBC, but I still thought it's not the real way.

I discussed this problem on the local php mailing list, everyone knew
it, noone knew why. I searched through the php manual and the rest of
the web just to find a fraction of information that could be helpful
either way, but there was no.

The extensions that made me 2 horribe days:

php_curl.dll
php_fdf.dll
php_gettext.dll
php_ifx.dll
php_interbase.dll
php_ldap.dll
php_mssql65.dll
php_mssql70.dll
php_oci8.dll
php_openssl.dll
php_oracle.dll
php_sablot.dll
php_sybase_ct.dll
php_yaz.dll

A sample message:
PHP Warning:  Unable to load dynamic library
'c:\php\extensions\php_sybase_ct.dll'

My question is if it is possible to load these extensions some way, or
should I give up trying, becouse they will never load.

Thank you a lot.

I wish you nice developing.

Best regards:

Mike Myers





Edit this bug report at http://bugs.php.net/?id=9266&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #9558 Updated: array hanling

[jan]

ID: 9558
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Feedback
Bug Type: Feature/Change Request
Operating System: All OS
PHP Version: 4.0.4pl1
New Comment:

what about unlink($array['key']); $first=$array[0]; and
$last=$array[count($array)-1]. ?

Previous Comments:


[2001-03-05 06:32:06] [EMAIL PROTECTED]

There are no normal delete from array (by the key) there are no
possibility to get last, first element without altering the array itself
(or to examine the element), in general in PHP the array is the list
replacement (2 in 1)(for ex. in C++ we have to separate types/classes:
array and the list)  but there are no list functionality. If this will
cause performance loses then new type creation sugested.





Edit this bug report at http://bugs.php.net/?id=9558&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #9266 Updated: Unable to load 14 of php's extensions

[jan]

ID: 9266
Updated by: jan
Reported By: [EMAIL PROTECTED]
Status: Analyzed
Old Bug Type: Feature/Change Request
Bug Type: Documentation problem
Operating System: WinMe, Win98, Win2000
PHP Version: 4.0.4pl1
New Comment:

so better documentation is needed for extensions that rely on external
libraries.

Previous Comments:


[2001-02-14 20:07:11] [EMAIL PROTECTED]

Yes, a more detailed error message, or/and two more lines in the manual
could have saved me some time. And maybe it would for others too.

Thank you.



[2001-02-14 16:49:32] [EMAIL PROTECTED]

all of theese extensions depend on other DLLs 
being installed on your system
like, for example, sybase_ct client DLLs
(this is the native sybase call interface, _not_ the ODBC stuff)

maybe we could improve the error message?
(moved to feature request)



[2001-02-14 16:49:10] [EMAIL PROTECTED]

all of theese extensions depend on other DLLs 
being installed on your system
like, for example, sybase_ct client DLLs
(this is the native sybase call interface, _not_ the ODBC stuff)

maybe we could improve the error message?
(moved to feature request)



[2001-02-14 16:41:08] [EMAIL PROTECTED]

I spent almost 2 day trying to figure out how to load the sybase
extension under windows. I tried everything I could, but 14 of the
extensions would not load under WinMe, Win98, Win2000, with any
combination of apache, pws, iis. The extension_dir is just set all fine,
and the rest of the extensions load and function well.

I only wanted to get some data from a sybase db. I could actually solve
it with ODBC, but I still thought it's not the real way.

I discussed this problem on the local php mailing list, everyone knew
it, noone knew why. I searched through the php manual and the rest of
the web just to find a fraction of information that could be helpful
either way, but there was no.

The extensions that made me 2 horribe days:

php_curl.dll
php_fdf.dll
php_gettext.dll
php_ifx.dll
php_interbase.dll
php_ldap.dll
php_mssql65.dll
php_mssql70.dll
php_oci8.dll
php_openssl.dll
php_oracle.dll
php_sablot.dll
php_sybase_ct.dll
php_yaz.dll

A sample message:
PHP Warning:  Unable to load dynamic library
'c:\php\extensions\php_sybase_ct.dll'

My question is if it is possible to load these extensions some way, or
should I give up trying, becouse they will never load.

Thank you a lot.

I wish you nice developing.

Best regards:

Mike Myers





Edit this bug report at http://bugs.php.net/?id=9266&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #9250 Updated: Nested Functions are BAD, they make for inefficient debugging

[jan]

ID: 9250
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Bogus
Bug Type: Feature/Change Request
Operating System: Any
PHP Version: 4.0.4pl1
New Comment:

brackets are used for other controll structures as well, so disabling
nested functions do not solve the problem. ->bogus

Previous Comments:


[2001-02-14 04:10:46] [EMAIL PROTECTED]

If you have a missing close-bracket "}" somewhere in your code, the PHP
parser tells you that the error is at the last line.  If you have a
multi-hundred or thousand line file, this makes life a bitch.

Rasmus tells me that the reason for this behavior is because functions
can be nested, which means that:

function a 
{
  function b
  {
  } 
}
is OK. This is a silly feature.  Why do you have it, for scoping
reasons?  It means that you can't effectively tell where syntax errors
occur.

Suggestion: Either make nested functions turned off by default but you
can turn them on with a PHP global variable, like
$PHP_BIZARRO_NESTED_FUNCTIONS = 1;

or

have the parser syntax set up so that it marks where a possible syntax
error occurred (IOW it sees a possible nested function, but it could
also be a syntax error), and if it reaches the end of the file with a
missing "}" or two, that it displays the location where it parsed the
nested function.





Edit this bug report at http://bugs.php.net/?id=9250&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #9195 Updated: Default class function arguments

[jan]

ID: 9195
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Analyzed
Bug Type: Feature/Change Request
Operating System: Linux
PHP Version: 4.0.2
New Comment:

please take part in the discussion about the upcoming ZendEngine2 on
[EMAIL PROTECTED] 

Previous Comments:


[2001-02-09 13:00:53] [EMAIL PROTECTED]

I'm requesting that the default function arguments for a class method be
able to reference the class variable members in it's definition. 

Here is an example of it's use:

class searchclass {

var $hits;
var $results;

function save($filename, $start = 0, $end = $this->hits){ /* <- this
returns Parse error: */
if($fp = @fopen($filename, "w")){
for($i = $start; $i < $end; $i++){
fwrite($fp, implode("|", $this->results[$i]));
}
}else{
return false;
}
}

function search(){
/* ... */
}

}





Edit this bug report at http://bugs.php.net/?id=9195&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #8978 Updated: Add a 'readonly' possibility to the session module

[jan]

ID: 8978
Updated by: jan
Reported By: Maxim Derkachev <[EMAIL PROTECTED]>
Old Status: Analyzed
Status: Closed
Bug Type: Feature/Change Request
Operating System: 
PHP Version: 4.0.4pl1
New Comment:

seems fixed then. reopen if I am wrong

Previous Comments:


[2001-05-16 04:17:12] Maxim Derkachev <[EMAIL PROTECTED]>

Forgot to include the batteries :)
After the patch above is applied, one could do:
session_start(SESS_READ_ONLY);
to start a readonly session. 
Functions that supposed to write the session data (core session
functions, not actual savehandler functions) will be disabled.
On the other page, if session_start() is called without the 
SESS_READ_ONLY flag, one could get the normal fully functional session,
which will write the data. That would allow to use session in framed
pages, when one frame is allowed to change the session data and another
frames only read the data, and in many other cases. E.g. for me the
feature has become inevitable when I needed to write a support chat,
which should read session variables, but should not change them and, the
most important, it should not save them, because a client could browse
other parts of the site  (and this could affect the sesson vars) while
he is chatting with the support. Without the readonly possibility, the
new session variables could be easily rewrited by the chat script with
outdated values.



[2001-05-16 04:02:23] Maxim Derkachev <[EMAIL PROTECTED]>

just made a patch against the current sources (session.c and
php_session.h).

*** php_session.h.orig  Tue May 15 15:16:50 2001
--- php_session.h   Tue May 15 15:23:26 2001
***
*** 96,100 
--- 96,103 
zend_bool define_sid;
zend_bool use_cookies;
+   int readonly;
  } php_ps_globals;
+
+ #define SESS_READONLY 1

  extern zend_module_entry session_module_entry;
*** session.c.orig  Tue May 15 15:16:04 2001
--- session.c   Wed May 16 11:54:31 2001
***
*** 526,529 
--- 526,533 
PLS_FETCH();

+   if (PS(readonly)) {
+   return;
+   }
+
if (!PG(register_globals)) {
if (!PS(http_session_vars)) {
***
*** 899,902 
--- 903,911 
zend_bool retval = SUCCESS;

+   if (PS(readonly)) {
+   php_error(E_WARNING, "Trying to destroy readonly
session");
+   return FAILURE;
+   }
+
if (PS(nr_open_sessions) == 0) {
php_error(E_WARNING, "Trying to destroy uninitialized
session");
***
*** 1265,1270 
--- 1274,1297 
  PHP_FUNCTION(session_start)
  {
+   pval **flag;
PSLS_FETCH();

+   if (ZEND_NUM_ARGS() > 1)
+   WRONG_PARAM_COUNT;
+
+   if (ZEND_NUM_ARGS() == 0 ) {
+   PS(readonly) = 0;
+   }
+   if (ZEND_NUM_ARGS() == 1 && zend_get_parameters_ex(1, &flag) !=
FAILURE) {
+   convert_to_long_ex(flag);
+   if (((int) ((*flag)->value.lval)) == SESS_READONLY) {
+   PS(readonly) = 1;
+   }
+   else {
+   PS(readonly) = 0;
+   }
+   }
+
+
php_session_start(PSLS_C);

***
*** 1314,1317 
--- 1341,1347 
PSLS_FETCH();

+   if (PS(readonly))
+   return;
+
if (PS(nr_open_sessions) == 0)
RETURN_FALSE;
***
*** 1353,1356 
--- 1383,1388 
PSLS_FETCH();

+   REGISTER_LONG_CONSTANT("SESS_READ_ONLY", SESS_READONLY,
CONST_CS);
+
php_rinit_session_globals(PSLS_C);

***
*** 1404,1407 
--- 1436,1440 
PS(module_number) = module_number;
REGISTER_INI_ENTRIES();
+   REGISTER_LONG_CONSTANT("SESS_READ_ONLY", SESS_READONLY,
CONST_CS);
return SUCCESS;
  }








[2001-01-29 06:21:31] Maxim Derkachev <[EMAIL PROTECTED]>

Just faced the fact that the possibility to call session 'readonly' 
should be added. 
For example, when somebody calls a framed pages where all 
frames are php scripts that needs session variables. But in this 
case only one of them should be allowed to write session state, 
because every frame would write session state in an unpredictable order,

and variables registered/changed in one frame could be overwritten 
by other frames, and that would definitely break an application. 
I suggest session_start could take an optional READONLY flag to 
disable write of the session data during the page shutdown.
The idea is similar to call page_close() on only one frame in a framed
page in PHPLib-based applications.





Edi

[PHP-DEV] Bug #8974 Updated: Blank lines in includes cause session errors

[jan]

ID: 8974
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Bogus
Bug Type: Feature/Change Request
Operating System: All
PHP Version: 4.0.4pl1
New Comment:

technically it is an user error. to solve it either eliminate any
whitespace or call ob_start() before session_start(). the same thing
applies for header(). no bug. bogus

Previous Comments:


[2001-01-29 00:42:42] [EMAIL PROTECTED]

A PHP script with includes followed by commands which set headers (ie
session_start) will generate errors if include file(s) contain blank
lines. For example:



where functions.inc contains leading or trailing blank lines will result
in an error like:

Warning: Cannot send session cache limiter - headers already sent

This problem appears to be very common (according to a search on
google), and has no doubt frustrated many PHP users.

Unfortunately this behaviour is technically 'correct', although it
should probably be fixed anyway.






Edit this bug report at http://bugs.php.net/?id=8974&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #8863 Updated: function to get constant values by variable name

[jan]

ID: 8863
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Closed
Bug Type: Feature/Change Request
Operating System: Any
PHP Version: 4.0.4pl1
New Comment:

and close it

Previous Comments:


[2002-01-06 13:00:36] [EMAIL PROTECTED]

see get_defined_constants



[2001-01-23 12:21:28] [EMAIL PROTECTED]

I believe PHP lacks a function to get constant values by
name where the constant's name is stored in a string variable.

What I mean is something like

function get_constant($const_name)
{
eval ("\$myvar = " . $const_name . ";");
return $myvar;
}

but I really think that using eval() here is a waste.

Am I missing something?

andrew





Edit this bug report at http://bugs.php.net/?id=8863&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #8863 Updated: function to get constant values by variable name

[jan]

ID: 8863
Updated by: jan
Reported By: [EMAIL PROTECTED]
Status: Open
Bug Type: Feature/Change Request
Operating System: Any
PHP Version: 4.0.4pl1
New Comment:

see get_defined_constants

Previous Comments:


[2001-01-23 12:21:28] [EMAIL PROTECTED]

I believe PHP lacks a function to get constant values by
name where the constant's name is stored in a string variable.

What I mean is something like

function get_constant($const_name)
{
eval ("\$myvar = " . $const_name . ";");
return $myvar;
}

but I really think that using eval() here is a waste.

Am I missing something?

andrew





Edit this bug report at http://bugs.php.net/?id=8863&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #8681 Updated: Function arguments in 'global'

[jan]

ID: 8681
Updated by: jan
Reported By: [EMAIL PROTECTED]
Status: Open
Bug Type: Feature/Change Request
Operating System: Linux/FreeBSD
PHP Version: 4.0.4
New Comment:

so what is this anyway?

Previous Comments:


[2001-02-22 10:52:15] [EMAIL PROTECTED]

Not a bug.



[2001-01-12 13:11:42] [EMAIL PROTECTED]

function foo ($bar)
{
  global $bar;
  # ...
}

Opposite to php3, php4 doesn't warn you. This is now documented.





Edit this bug report at http://bugs.php.net/?id=8681&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #8599 Updated: a kind Exception handling

[jan]

ID: 8599
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Closed
Bug Type: Feature/Change Request
Operating System: 
PHP Version: 4.0.4
New Comment:

We'll have exception handling in PHP5, please follow the development of
the ZendEngine2 on [EMAIL PROTECTED] closing.

Previous Comments:


[2001-01-10 01:58:07] [EMAIL PROTECTED]

http://php.net/manual/function.set-error-handler.php
set_error_handler lets you register your own error handling function

however try...catch is also a much-needed feature



[2001-01-08 13:48:53] [EMAIL PROTECTED]

something like java's exception handling
or more kind of register_shutdown_function i.e
register_mysql_errorhandling_function

thnx for the attention





Edit this bug report at http://bugs.php.net/?id=8599&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] CVS Account Request: ott

[Andreas Otterstein]

translation PHP manual into german

-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #8214 Updated: Request for Signal Handling functions

[jan]

ID: 8214
Updated by: jan
Reported By: [EMAIL PROTECTED]
Status: Open
Bug Type: Feature/Change Request
Operating System: *nix - CGI
PHP Version: 4.0 Latest CVS (12/12/2000)
New Comment:

doesen't pcnt do the job?

Previous Comments:


[2000-12-12 16:49:52] [EMAIL PROTECTED]

I would rather see:
posix_register_sig_func(char function, mixed signal)

And handle both "9" and "KILL" as valid signals.



[2000-12-12 16:35:41] [EMAIL PROTECTED]

A few additions to the posix module to allow registering functions when
certain signals are recieved would be cool. somthing like

posix_register_sig_func(char function,int signal);

where signal is HUP etc :)

would allow some long running systems in CGI mode with a graceous
shutdown. 


James





Edit this bug report at http://bugs.php.net/?id=8214&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #8197 Updated: lack of socket() functions

[jan]

ID: 8197
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Closed
Bug Type: Feature/Change Request
Operating System: Windows
PHP Version: 4.0.3pl1
New Comment:

they should be now part of the windows distribution. closing

Previous Comments:


[2000-12-11 16:58:28] [EMAIL PROTECTED]

Please include a DLL (or built-in) for the socket() functions in a Win32
build.





Edit this bug report at http://bugs.php.net/?id=8197&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #8058 Updated: Missing PDF functionality

[jan]

ID: 8058
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Summary: 
Old Status: Open
Status: Closed
Bug Type: Feature/Change Request
Operating System: linux 2.2.x
PHP Version: 4.0.3pl1
New Comment:

closing, much has been added since the request came in. reopen it if
something is still missing.

Previous Comments:


[2000-11-30 15:19:09] [EMAIL PROTECTED]

I'd like to see some new function into PDF support like:

- URL links functions (PDFLib do that)
- Form support so we can interface with FDF.

Thanks





Edit this bug report at http://bugs.php.net/?id=8058&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #7974 Updated: N/A

[jan]

ID: 7974
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Closed
Bug Type: Feature/Change Request
Operating System: N/A
PHP Version: 4.0.3pl1
New Comment:

please take part in the discussion on [EMAIL PROTECTED] about the
upcoming features in ZE2. PHP4 is not likely to see it.

Previous Comments:


[2000-11-25 20:02:41] [EMAIL PROTECTED]

Add multiple inheritance to Classes.





Edit this bug report at http://bugs.php.net/?id=7974&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #6768 Updated: list() construct reference assignment

[jan]

ID: 6768
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Duplicate
Bug Type: Feature/Change Request
Operating System: Linux 2.2.14
PHP Version: 4.0.2
New Comment:

dupe of #7930 and marking it "dublicate"

Previous Comments:


[2000-09-15 05:28:18] [EMAIL PROTECTED]

I'm requesting that that the list() construct which is an assignment
construct, assign variable references. Like such..







Edit this bug report at http://bugs.php.net/?id=6768&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #7899 Updated: Class selfdestruct

[jan]

ID: 7899
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Closed
Bug Type: Feature/Change Request
Operating System: any
PHP Version: 4.0 Latest CVS (21/11/2000)
New Comment:

we'll have destructors and more improved OO with ZE2 (PHP5). please
propose anything to [EMAIL PROTECTED] closing

Previous Comments:


[2000-11-21 04:41:15] [EMAIL PROTECTED]

Hiya 

I'd like to see a destructor for classes from within the class 

example:

$file = new file("somefile.ext);
// instert some long script here
$file->del_file();
at this point $file should be FALSE of NULL (set from within the
class)

some ideas:
unset($this);
or $this->Destruct();

Greetz,

Wico





Edit this bug report at http://bugs.php.net/?id=7899&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #7480 Updated: next(), end(), prev(), reset() etc. don't return references

[jan]

ID: 7480
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Bogus
Status: Duplicate
Bug Type: Feature/Change Request
Operating System: RedHat Linux 7.0
PHP Version: 4.0.1pl2
New Comment:

actually they are dublicates, silly me.

Previous Comments:


[2002-01-06 12:01:10] [EMAIL PROTECTED]

dupes are bogus..



[2002-01-06 11:47:32] [EMAIL PROTECTED]

dupe of 7578. 



[2000-10-26 09:26:40] [EMAIL PROTECTED]

next(), end(), prev(), reset() etc. don't return references

meaning this example code won't work as it's thought to do:







Edit this bug report at http://bugs.php.net/?id=7480&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #7480 Updated: next(), end(), prev(), reset() etc. don't return references

[jan]

ID: 7480
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Closed
Status: Bogus
Bug Type: Feature/Change Request
Operating System: RedHat Linux 7.0
PHP Version: 4.0.1pl2
New Comment:

dupes are bogus..

Previous Comments:


[2002-01-06 11:47:32] [EMAIL PROTECTED]

dupe of 7578. 



[2000-10-26 09:26:40] [EMAIL PROTECTED]

next(), end(), prev(), reset() etc. don't return references

meaning this example code won't work as it's thought to do:







Edit this bug report at http://bugs.php.net/?id=7480&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #7766 Updated: not OO until automatic class destructor is added

[jan]

ID: 7766
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Closed
Bug Type: Feature/Change Request
Operating System: FreeBSD Apache
PHP Version: 4.0.3pl1
New Comment:

we'll have destructors with ZE2 (PHP5).  closing

Previous Comments:


[2000-11-12 04:49:46] [EMAIL PROTECTED]

Moved to feature requests




[2000-11-12 04:33:56] [EMAIL PROTECTED]

Then why add classes?  Are you implying to businesses that if they
invest in your features, you do not intend to complete them?



[2000-11-12 04:21:58] [EMAIL PROTECTED]

Then why add classes?  Are you implying to businesses that if they
invest in your features, you do not intend to complete them?



[2000-11-12 04:00:29] [EMAIL PROTECTED]

Nobody ever claimed that PHP was an OO language.



[2000-11-12 03:24:44] [EMAIL PROTECTED]

The lack of an automatic class destructor makes it impossible to make
code completely OO.

For example, we have a Table class which outputs an HTML  with
error checking.  It is impossible to detect the case where
Table::EndTable() is not called ( not issued) because class
destructors do not exist in the PHP4 language.

PLEASE CORRECT THIS GLARING DEFICIENCY ASAP!

I am a prolific commercial programming of $million products (Corel
Painter, Cool Page, etc).  PHP4 is a an incredible tool.  We are
considering purchasing the new compiler technology.  So please fix this
silly omission.  Please.  Please.





Edit this bug report at http://bugs.php.net/?id=7766&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #7480 Updated: next(), end(), prev(), reset() etc. don't return references

[jan]

ID: 7480
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Closed
Bug Type: Feature/Change Request
Operating System: RedHat Linux 7.0
PHP Version: 4.0.1pl2
New Comment:

dupe of 7578. 

Previous Comments:


[2000-10-26 09:26:40] [EMAIL PROTECTED]

next(), end(), prev(), reset() etc. don't return references

meaning this example code won't work as it's thought to do:







Edit this bug report at http://bugs.php.net/?id=7480&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14891: Additional libraries directories missing during compilation

[bp]

From: [EMAIL PROTECTED]
Operating system: FreeBSD 4.2-RELEASE
PHP version:  4.1.1
PHP Bug Type: Compile Failure
Bug description:  Additional libraries directories missing during compilation

./configure --with-pgsql=/usr/local/some-non-std-path/pgsql
produced a config_vars.mk files ommiting the postgres
library directory:

$ grep /usr/local/some-non-std-path/pgsql config_vars.mk
$

config_vars.mk had to be hand tooled to compile php.

This trouble is also exhibited with msql.
-- 
Edit bug report at: http://bugs.php.net/?id=14891&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

Re: [PHP-DEV] Care to test cli sapi?

[Edin Kadribasic]

> > > Looks pretty cool!  Finally someone actually took the time to do it :)
> > >
> > > Do you have CVS access?
> >
> > Yes. Should I commit it?
> 
> Yes
> 
> Derick

Well its in the CVS now. You can go ahead and find all the bugs :)

Edin


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #7276 Updated: CVSup/rsync/easier patch access desired

[jan]

ID: 7276
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Closed
Bug Type: Feature/Change Request
Operating System: n/a
PHP Version: 4.0.3
New Comment:

no bug, see http://www.php.net/anoncvs.php

Previous Comments:


[2000-10-17 07:59:52] [EMAIL PROTECTED]

Hello,

a few days ago I downloaded 4.0.3 and now seem to have to download
4.0.3pl1. I'd rather not download everything and  then also re-configure
everything. I'd rather download a patch kit and apply that, keeping my
existing configuration, and run, or I'd like to cvsup a source tree and
start from there. Unfortunately you don't seem to support cvsup or the
like, and also don't offer a plain FTP service, less so one with patch
kits, or at least I couldn't find it.

A comment is appreciated since your CVS server also doesn't exactly
scream (may also be a network problem, of course).


Best Regards,
--Toni++





Edit this bug report at http://bugs.php.net/?id=7276&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #13034 Updated: Apache segfaults when acessing certain scripts

[dk]

ID: 13034
User updated by: [EMAIL PROTECTED]
Reported By: [EMAIL PROTECTED]
Old Status: Feedback
Status: Open
Bug Type: Reproducible crash
Operating System: RedHat Linux 7.0.
PHP Version: 4.0.6
New Comment:

Well that problem occured some time ago.
As far as I can say it doesn't occur with php-4.1.1, but it occured only
under some circumstances:
Accessing Arrays or Objects in a recursive way.

Well with 4.1.1 there are other problems now:
When I use the builtin sessionmanagement with register_globals off it
doesn't work at all (I followed the release notes) - but that is another
story and I don't have the time right now to do debugging on that.

best regards

Daniel [datenPUNK] Khan

Previous Comments:


[2002-01-06 07:39:33] [EMAIL PROTECTED]

Does this problem still occur with 4.1.1 and/or the latest CVS?



[2001-08-30 17:10:27] [EMAIL PROTECTED]

yes - i couldn't get the crash over apache - then i tried via CGI and it
crashed.
it crashes with php 4.0.6 and the dev snap of 29th of august.




[2001-08-30 17:00:23] [EMAIL PROTECTED]

I can not reproduce this.




[2001-08-30 16:59:57] [EMAIL PROTECTED]

You didn't say with what version this backtrace was generated with, also
is it with that script you said
you could NOT get the crash? Please be clear about these things.





[2001-08-30 13:28:19] [EMAIL PROTECTED]

hello,

please read the whole thread - i have included the script AND i have
also tried it with the latest development snapshot. i really have
included all i have :)
for me it looks like a bug in the echo function (vfprintf.c). the script
doesn't crash without the echo funtions. it also crashes with "print".

regards

daniel



The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/?id=13034


Edit this bug report at http://bugs.php.net/?id=13034&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #6701 Updated: Use of http_proxy for fopen("http:// ...

[jan]

ID: 6701
Updated by: jan
Reported By: [EMAIL PROTECTED]
Status: Open
Bug Type: Feature/Change Request
Operating System: Customized RedHat 6.2
PHP Version: 4.0.0
New Comment:

although the request should remain open, this specific problem can be
solved using PEAR::Cache.

Previous Comments:


[2000-09-12 23:39:12] [EMAIL PROTECTED]

It would be nice if fopen("http://www.mywebsite.com";, "r"); used the
value in http_proxy as a proxy server instead of always fetching the
data directly.  Using CURL is an option, but its a lot more complex.  If
this functionality isn't too difficult to achieve, it would be much
appreciated.

For example, I have a PHP program that HTMLizes RFCs and displays them. 
It grabs the rfc###.txt directly from http://www.ietf.org/ but there's
no reason for it to refetch them every time if they can be cached by our
proxy server.

Note: blanking $http_proxy should disable the proxy.





Edit this bug report at http://bugs.php.net/?id=6701&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #14890: is_uploaded_file does not work correctly

[fjperreault]

From: [EMAIL PROTECTED]
Operating system: Unix/Linux
PHP version:  4.1.1
PHP Bug Type: Feature/Change Request
Bug description:  is_uploaded_file does not work correctly

This function can return a false positive if you set a MAX_FILE_SIZE limit
in the HTML form sending the file and then proceeed to send a file bigger
than this max.  If a call is made to _is_uploaded_file to see if the file
made it to the server, the function will return TRUE even though the file
doesn't exist on the server and the filename is set to "none".  In this
case, the function _is_uploaded_file should return FALSE, not TRUE because
it has not been uploaded.
-- 
Edit bug report at: http://bugs.php.net/?id=14890&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #6380 Updated: Session handlers should be allowed in objects

[jan]

ID: 6380
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Closed
Bug Type: Feature/Change Request
Operating System: 
PHP Version: 4.0 Latest CVS (27/08/2000)
New Comment:

this is possible using
session_set_save_handler(array('class'=>'method'),...);
or do I misunderstand the problem here?

Previous Comments:


[2000-08-27 14:32:53] [EMAIL PROTECTED]

It would be nice (for encapsulation purposes) to allow user-defined
session handlers to be implemented as class methods, either through the
established array($this, "foo") syntax or via a session_set_object()
function.





Edit this bug report at http://bugs.php.net/?id=6380&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #13341 Updated: Seffault in outputbuffers

[wico]

ID: 13341
User updated by: [EMAIL PROTECTED]
Reported By: [EMAIL PROTECTED]
Old Status: Feedback
Status: Open
Bug Type: Reproducible crash
Operating System: linux
PHP Version: 4.0.6
New Comment:

yup still segfaults with 4.1.0

Previous Comments:


[2002-01-06 07:41:19] [EMAIL PROTECTED]

Does this problem still occur with 4.1.1 and/or the latest CVS?



[2001-09-17 06:20:22] [EMAIL PROTECTED]

Fixed typo;

Btw where did the bug reporters stats go? It looks good to be on top ;)



[2001-09-17 06:14:58] [EMAIL PROTECTED]

Hiya,

Using the same var from the function and changing will sefault, copying
it first to another var will work... but should not happen, see
examples

I segfaulted :( :( :(

Greetz,

Wico de Leeuw





Edit this bug report at http://bugs.php.net/?id=13341&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

[PHP-DEV] Bug #6136 Updated: Exception Handling

[jan]

ID: 6136
Updated by: jan
Reported By: [EMAIL PROTECTED]
Old Status: Open
Status: Closed
Bug Type: Feature/Change Request
Operating System: linux
PHP Version: 4.0 Latest CVS (13/08/2000)
New Comment:

this will come with ZendEngine2 (PHP5). closing.

Previous Comments:


[2000-08-13 10:21:44] [EMAIL PROTECTED]

Error handling: please please please add exception handling (some sort
of "try ... catch ... finally" as implemented in Java, C++, Delphi, ...)
to PHP.

It is simply annoying to check function results for errors and to do
error handling in if-statements.

Especially for the new MySQL-transactions such an exception-handling
would be great.





Edit this bug report at http://bugs.php.net/?id=6136&edit=1


-- 
PHP Development Mailing List 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]