Re: [PHP-DEV] Re: TestFest 2010 News Announcement.

2010-06-10 Thread Hannes Magnusson
Why doesn't he have karma yet? IMO web/php/trunk/archive/ should be writable by all developers.. Eric: You need to commit the xinclude entry to the archive.xml file too. ..And php-webmaster@ is for webmaster stuffs :) -Hannes On Wed, Jun 9, 2010 at 16:45, Eric Stewart ericleestew...@gmail.com

Re: [PHP-DEV] [PATCH] New PDO methods for PostgreSQL driver

2010-06-10 Thread Ilia Alshanetsky
Denis, I've cleaned up the patch, fixing whitespace, a few memory leaks and optimized the copy from array code. The code for the copy to/from array/file was just committed. I am still reviewing the inTransaction code, since that is something I think can probably be generic and not database

Re: [PHP-DEV] [PATCH] New PDO methods for PostgreSQL driver

2010-06-10 Thread Ilia Alshanetsky
I've added the transaction code as a generic method inTransaction(), by default it'll just use in_txn internal parameter, but allows the driver to extend this (as was done in PostgreSQL) and provide a detailed status code. On Wed, May 26, 2010 at 1:11 PM, Denis Gasparin

Re: [PHP-DEV] Type hinting

2010-06-10 Thread Christian Schneider
Zeev Suraski wrote: That's exactly what I had in mind. Have E_TYPE off by default and implement it throughout PHP - instead of just type hints. for example: $foo = 0; $foo += (int)'123abc'; // no error $foo += '123abc'; // E_TYPE Out of curiosity: What would you do with the E_TYPE on the

Re: [PHP-DEV] Type hinting

2010-06-10 Thread Zeev Suraski
At 15:12 10/06/2010, Christian Schneider wrote: Zeev Suraski wrote: That's exactly what I had in mind. Have E_TYPE off by default and implement it throughout PHP - instead of just type hints. for example: $foo = 0; $foo += (int)'123abc'; // no error $foo += '123abc'; // E_TYPE Out of

Re: [PHP-DEV] Type hinting

2010-06-10 Thread Ferenc Kovacs
On Thu, Jun 10, 2010 at 2:12 PM, Christian Schneider cschn...@cschneid.comwrote: Zeev Suraski wrote: That's exactly what I had in mind. Have E_TYPE off by default and implement it throughout PHP - instead of just type hints. for example: $foo = 0; $foo += (int)'123abc'; // no error

[PHP-DEV] Gearman support for php-fpm?

2010-06-10 Thread Rasmus Lerdorf
Has anybody thought about adding Gearman support to the fpm sapi? Managing Gearman workers is pretty much identical to managing fastcgi workers, so it doesn't seem like much of a stretch. -Rasmus -- PHP Internals - PHP Runtime Development Mailing List To unsubscribe, visit:

Re: [PHP-DEV] Gearman support for php-fpm?

2010-06-10 Thread Reilly, Patrick
Rasmus, I can work on it if no one has it already. Warm Regards, Patrick On Thu, Jun 10, 2010 at 7:56 AM, Rasmus Lerdorf ras...@lerdorf.com wrote: Has anybody thought about adding Gearman support to the fpm sapi? Managing Gearman workers is pretty much identical to managing fastcgi workers,

Re: [PHP-DEV] Gearman support for php-fpm?

2010-06-10 Thread Brian Moon
The big difference is that different processes do different jobs in Gearman. All PHP workers in fpm just run what ever code is handed to them. How do you handle that? Brian. http://brian.moonspot.net/ On 6/10/10 9:56 AM, Rasmus Lerdorf wrote: Has anybody thought about adding

Re: [PHP-DEV] Gearman support for php-fpm?

2010-06-10 Thread Frederic Hardy
On 06/10/2010 17:20, Brian Moon wrote: The big difference is that different processes do different jobs in Gearman. All PHP workers in fpm just run what ever code is handed to them. How do you handle that? Threading ;) ? On 6/10/10 9:56 AM, Rasmus Lerdorf wrote: Has anybody thought about

Re: [PHP-DEV] Gearman support for php-fpm?

2010-06-10 Thread Rasmus Lerdorf
On 6/10/10 8:30 AM, Frederic Hardy wrote: On 06/10/2010 17:20, Brian Moon wrote: The big difference is that different processes do different jobs in Gearman. All PHP workers in fpm just run what ever code is handed to them. How do you handle that? Threading ;) ? Definitely not. Gearman is

Re: [PHP-DEV] Gearman support for php-fpm?

2010-06-10 Thread Tjerk Anne Meesters
All fpm workers would be called using the same function name (e.g. runphp). This allows you to reuse a gearmand instance for other workers too. The advantages of fpm vs gearman would be mostly performance (it uses ping and requires an underlying layer on top of tcp/ip), followed by dynamic

Re: [PHP-DEV] Gearman support for php-fpm?

2010-06-10 Thread Rasmus Lerdorf
On 6/10/10 8:53 AM, Tjerk Anne Meesters wrote: All fpm workers would be called using the same function name (e.g. runphp). This allows you to reuse a gearmand instance for other workers too. I don't think we would need to require that. Obviously you could build your Gearman worker such that

Re: [PHP-DEV] Gearman support for php-fpm?

2010-06-10 Thread Brian Moon
[gearman] worker1.path = /path/to/worker1.php worker1.pm = dynamic worker1.pm.max_children = 10 worker1.pm.start_servers = 5 worker2.path = /path/to/worker2.php worker2.pm = dynamic worker2.pm.max_children = 4 worker2.pm.start_servers = 2 So, this is a proposed new entry in the ini format? I

Re: [PHP-DEV] Gearman support for php-fpm?

2010-06-10 Thread Rasmus Lerdorf
On 6/10/10 9:31 AM, Brian Moon wrote: [gearman] worker1.path = /path/to/worker1.php worker1.pm = dynamic worker1.pm.max_children = 10 worker1.pm.start_servers = 5 worker2.path = /path/to/worker2.php worker2.pm = dynamic worker2.pm.max_children = 4 worker2.pm.start_servers = 2 So, this

Re: [PHP-DEV] Proper return after throwing exception

2010-06-10 Thread Alexey Zakhlestin
On 10.06.2010, at 23:41, Pieter de Zwart wrote: Hey guys, Another n00b question: What is the proper way to return from a function after throwing an exception? My code looks like: zend_throw_exception(amqp_connection_exception_class_entry, blah blah blah, 0 TSRMLS_CC); return;

[PHP-DEV] Proper return after throwing exception

2010-06-10 Thread Pieter de Zwart
Hey guys, Another n00b question: What is the proper way to return from a function after throwing an exception? My code looks like: zend_throw_exception(amqp_connection_exception_class_entry, blah blah blah, 0 TSRMLS_CC); return; Should I instead RETURN_FALSE from there? Thanks, Pieter --