Re: [PHP4BETA] What is ext/lcg/lcg.c?
| From: | Zeev Suraski | Date: | Sun, 29 Aug 1999 22:14:21 +0000 |
| Subject: | Re: [PHP4BETA] What is ext/lcg/lcg.c? | ||
| References: | 1 | Groups: | php.version4 |
| Request: | Send a blank email to php-version4+get-3855@lists.php.net to get a copy of this message | ||
At 22:45 29/08/99 , Stig Bakken wrote:
On Sun, 29 Aug 1999, Zeev Suraski wrote: At 22:30 29/08/99 , Sascha Schumann wrote:Not always, but when it comes to architecture, I think we should. It will take years before Apache 2.0 goes from vapourware to an actual shelf product, and more years before it becomes popular. We should not establish our architecture around it. Much like we can exploit Apache 2 specific hooks, we can easily (much more easily as a matter of fact) feature a pooling API for multithreaded servers such as IIS. I didn't want to do that since it would not be compatible with the multiprocess model. I don't believe in having SAPI cover everything and have the module query it as to which capabilities are available and which aren't. Things are complex enough as it is. Zeev -- Zeev Suraski <zeev@zend.com> http://www.zend.com/ For a PGP public key, finger bourbon@netvision.net.ilhonest. Do youWell, I'm not sure what we're waiting for exactly, to beplatformhave an idea on how to implement this shared pool of objects in aWe can't really rely on Apache though. We must use features available in all major servers. Yes, but we cannot always restrict ourselves to use only the least common subset of functionality. If one extension requires something that SAPI's backend can't provide, we just have to disable parts of that extension.independent (read: works under both UNIX and Win32) way?Well, I'm not sure what you mean with "objects."As Rasmus noted, we will need some flexible mechanism to provide modules with additional capabilities (i.e. on process/thread startup/shutdown events already exists, I needed on server startup/shutdown events). These fit well into the existing zend_module_entry structure.Stig mentioned portability problems wrt server capabilities, i.e.#define SAPI_PRE_FORK_START_HOOK #define SAPI_PROCESS_START_HOOKwhich allow modules to adapt their behaviour to possible configurations.I don't know what new features Apache 2 will introduce in this area, but we should take that into account. Maybe someone has pointers to papers about the new design of Apache.