> > > And a second: I downloaded your example from ulim/iebug and > experimented > > > a bit with it. If i save the generatet pdf in a file and put it in > the > > > same directory and call it, it works in all browsers, including IE > 5.5. > > > How do you explan this? There should be no difference between the > > > streams? > > > > As I said, IE does some checks on the stream, if it arrives via HTTP, > > which it doesn't do, when opening a file from the filesystem. > > No, you misunderstood me. I saved the PDF stream in a file, put it in > the webapps/<Context>-directory and served it using tomcat via HTTP, > same context and server as the cocoon output. > Ever tried this? Try it with your bug sample, it takes just some > seconds. How do you explain this difference? AFAI remember the thing that really solved the IE-PDF problems in C1 was setting the Content-Length header. This would fit in the picture you described... But this needs to be double-checked... But AFAI understood setting a Content-Length header in C2 could be a problem because we have streams - no documents. So what about creating the cache object first and then serving the request from cache - so we know about the size of the object. Maybe Carsten can tell if this possible at all? -- Torsten --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
