Re: Should php3_fopen_for_parser change the current working directory?

On Tue, Nov 23, 1999 at 10:14:07PM -0800, David Hedbor wrote: > Sascha Schumann <sascha@schumann.cx> writes: > > > On Tue, Nov 23, 1999 at 07:23:27PM -0500, rubys@us.ibm.com wrote: > > > > > > > > > Recap: > > > > > > Virtually simultaneously, I suggested that this code be removed, David > > > Hedbor suggested it be made virtual (emulated), and Sascha Schumann added > > > this to the todo list. > > > > In fact, this was discussed some months ago already. > > > > > I believe that Sascha and David are on the right long term path. For the > > > short term, I would like to propose that this code be guarded with an IFDEF > > > specifically for Apache and CGI. > > > > You can make it more general: For everything which does not > > use ZTS. > > Well, that is not entirely a good idea. The Roxen module currently > doesn't use ZTS (actually doesn't work with it). It's still a problem > since Roxen itself is threaded. What I did is to have a mutex lock > that prevents running more than one PHP script at any one time. Maybe > it could simply be a configure option, that would be automatically > enabled by ZTS and Roxen (to take two examples)? Ie when ZTS is on, it > would be enabled w/o extra effort from the person compiling the > source. Before we go into great detail on how to disable something broken, we should look into fixing it. Do you plan to make the Roxen module multi-threaded? -- Regards, Sascha Schumann Consultant