Message-Id: <firstname.lastname@example.org> Date: Fri, 27 Jun 1997 12:59:49 -0400 To: Sven Verdoolaege <email@example.com> From: Jonathan Roy <firstname.lastname@example.org> Subject: Re: FastCGI with: Solaris 2.5.X, Apache, and Perl with sfio97 In-Reply-To: <email@example.com> That's understandable, but keep in mind some new naive person installing FastCGI/etc under Solaris who gets the latest version of things like FCGI.pm will have an unreliable setup which can lock up the FCGI manager. I guess the real solution is for Open Market to get the patches into their development kit, release a new version of it, then release a FCGI.pm .31 with the patches applied... -Jonathan At 06:51 PM 6/27/97 +0200, Sven Verdoolaege wrote: >On Jun 26, Jonathan Roy <firstname.lastname@example.org> wrote: >> >> FCGI-0.30 does NOT have the os_unix.c OS_IsFcgi() patch to include >> locking around the accept() call. This is the problem area which causes the >> manager to die on Solaris. I can only assume this means the patch to >> prevent zillions of dying fcgis at restart time is also missing. I don't >> believe that patch ever appeared on the mailing list... it must have some >> sort of anti-patch filter running. ;) >> >Hmm... I _did_ just copy the one from the fastcgi 2.02b distribution. >I'm not sure I should have patched files. >I'd rather wait for a new library release. > >To the person that will do that, could you _please_ include >the version number in the name of the tarball. > >skimo > > -- Jonathan Roy - email@example.com - Idle Communications, Inc. Idle Communications accepts contract work in Perl/C/C++ programming, CGI development, and general system admin.