Hi Alex, On Tue, Mar 15, 2011 at 2:44 PM, Alexander Burger <[email protected]> wrote: > Hi Edwin, > >> (de app () >> (unless *SesId >> (setq >> *Port% (not *Gate) >> *SesId (pack (in "/dev/urandom" (rd 7)) "~") >> *Sock (port *HPorts '*Port) ) >> (timeout *Timeout) ) ) >> >> from there, *Sesid and *Timeout are per per child globals. that much > > Right. > > In fact, the whole environment (global as well as local) is private for > each child.
a-ha. yes. > > >> i can infer. *Gate is a flag for usage of httpGate? how about *Port%? > > Yes. '*Gate' is an indicator that the communication with the client runs > via 'httpGate'. It is used to construct internal link addresses, base > hrefs etc. > > '*Port%' is related to that. It is a flag which is set only temporarily > in the initial phase of a session, to suppress AJAX transfers (and thus > fall back to plain HTTP transfers) for the _first_ transaction of a > child process. ah. thank you. was wondering why that was. > > The reason is that AJAX transfers fail if the port changed. It seems to > believe that the request is from another server (according to the rule > is that XMLHttpRequests can only access content from the originating > server), despite the IP address is the same, just another port number. > > Cheers, > - Alex > -- > UNSUBSCRIBE: mailto:[email protected]?subject=Unsubscribe > -- UNSUBSCRIBE: mailto:[email protected]?subject=Unsubscribe
