--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
I have a servlet on TC 5.5.9 that services a page with applet content that
will reguire three GETs to complete the page 1) basic page, 2) applet jar
and 3)image. The basic page GETs served OK and the browser comes back with
a GET to request the JAR file which is modestly large (150K) so I attempt to
chunk the output (8K chunks). I read the file into a ByteArrayOutputStream
to get the real byte count as usual, then setup the headers before starting
the write, then loop on [write a chunk, flush chunk], but the flush hangs at
the 32K mark every time (actually, the next flush after the previous 32k
mark was hit). I've set the response buffer size to be larger than the byte
array but it still does not help. I know sometimes output sockets need to
be flushed, so I added a response.reset() as the first step, but does not
help either. Looks as though the browser (IE) can't handle the large
chunked stream or the response.flush() is not being translated to output to
the browser. I would have thought setting the content type to
application/x-java-archive would be suffient to tell the browser how to
handle it, but who knows. Any ideas or experiences?
- outputstream.flush() hangs at 32K Daniel