Re: HTTP_Header and browser abortion

On Mon, 18 Oct 2004 15:40:31 +0200 (MEST), Andre Bischof <a.bischof@gmx.net> wrote: > Hello, > > I'm trying to keep a connection to a browser over a long period (the server > generates very big files in the meantime, and should afterwards display the > download links to the files). > > But after some time (amount seems to differ) the connection is closed. Maybe > it's not the browser who closes the connection, but s.th. in between, like a > proxy? I tried different browsers (IE, Mozilla, Opera), but the connections > closes every time. > > If I use w3m on the server itself, everythings works (in the worst case it > took 13 min, but then the download links are displayed). > > At the moment I use only: Connection: Keep-Alive > > But I also tried (different combinations): > > Expires: 0 > Cache-Control: must-revalidate, post-check=0, pre-check=0 > Cache-Control: no-cache, must-revalidate > Pragma: public > Pragma: no-cache > Expires: Fri, 01 Jan 1999 00:00:00 GMT" > Last-Modified: ".gmdate("D, d M Y H:i:s")."GMT" > Transfer-Encoding: quoted-printable > Connection: default > > Could anyone tell me the correct headers to maintain a connection or point > me to some documentation to that topic? I could not find any documention or > examples within HTTP_Header on the pear website, or did I search wrong? > > Best regards, > André Bischof > I would suggest saving the request to a DB and having a process outside of the webserver generate the files (like a cron job). This would mean that you're not dependant on the client's browser settings. -- paperCrane --Justin Patrin--