- From: Roland Mainz <Roland.Mainz@informatik.med.uni-giessen.de>
- Date: Mon, 05 Apr 1999 22:01:38 +0200
- To: "jigsaw@w3.org" <jigsaw@w3.org>, W3 Jigsaw Mailinglist <www-jigsaw@w3.org>
Hi ! ---- Roland Mainz wrote: > The following code does NOT work, and the question is: WHY ? > The file downloaded by the browser is different in size from the > original file,and ZIP > files are not recogzized as such by WinZip 7.0 (reg). > (Method d( String ) is only a logging function). > > di.filename ="dk-lite.zip"; > di.contenttype = "application/octet-stream"; > > It doesn't matter if I use a BufferedINputStream instead of the plane > FileInputStream > as input, and it doesn't matter to move getOutputStream behind the > setContentType > method call... > It doesn't matter to use the int-copy-loop below, or the 256-byte copy > loop which is > commented-out here. ...and so on... I found - maybe - a part of the puzzle... Question: Why does org.w3c.jigsaw.servlet.JIgsawHttpServletResponse.flushStream() call reply.setContentLength() ? IMHO corrupts this any previous, servlet-"made" setContentLength call (this assumes that flushStrean() is called after servlet's doGet() is done...). ---- Bye, Roland -- __ . . __ (o.\ \/ /.o) Roland Mainz C programmer \__\/\/__/ Roland.Mainz@informatik.med.uni-giessen.de MPEG specialist /O /==\ O\ gisburn@w-specht.rhein-ruhr.de Sun&&Amiga programmer (;O/ \/ \O;) TEL +49 (0) 2426901568 FAX +49 (0) 2426901569
Received on Monday, 5 April 1999 16:00:41 UTC