Hi Alvaro, Alvaro Lopez Ortega wrote: [...] > I have tried to reproduce the problem but I have not succeed. Both > fcgi and scgi pass the same variable set: > > ========= > $ ./manage.py runfcgi method=threaded host=127.0.0.1 port=3033 > protocol=scgi daemonize=false > > 2008-08-25 14:30:45 : WSGIServer starting up > 2008-08-25 14:30:47 : GET /demo/lalala > environ['PATH_INFO'] /lalala > environ['SCRIPT_NAME'] /demo > ========= > > ========= > $ ./manage.py runfcgi method=threaded host=127.0.0.1 port=3033 > protocol=fcgi daemonize=false > > environ['PATH_INFO'] /lalala > environ['SCRIPT_NAME'] /demo > ========= > > Anyway, that wasn't the main issue, was it?
I wasn't clear in my original mail. I was comparing: ./manage.py runserver with ./manage.py runfcgi ... That is, running the normal "django" non-flup server (HTTP) vs flup server (SCGI/FCGI). But maybe the current behavior is to be expected, and I'm misunderstanding this? My original cherokee setup was: vserver!1!rule!10000!match = directory vserver!1!rule!10000!match!directory = /demo but if I change it to: vserver!1!rule!10000!match = directory vserver!1!rule!10000!match!directory = / it to works without changing django, though, I would like to have several django scgi apps mounted on different places, eg: /demo1 --> scgi_app1 /demo2 --> scgi_app2 ... Humm, but maybe this is not really a cherokee question, and the ball is entirely on the scgi app side? Thanks! Best regards, Rui Lopes _______________________________________________ Cherokee mailing list [email protected] http://lists.octality.com/listinfo/cherokee
