general/3794: Java client gets server status 206 or server status 200 as exception when reading data from server

[ARiel Bardin <ar...@cadviewer.com>]

>Number:         3794
>Category:       general
>Synopsis:       Java client gets server status 206 or server status 200 as exception when reading data from server
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    apache
>State:          open
>Class:          sw-bug
>Submitter-Id:   apache
>Arrival-Date:   Fri Jan 29 10:30:01 PST 1999
>Last-Modified:
>Originator:     ariel@cadviewer.com
>Organization:
apache
>Release:        1.2.6
>Environment:
Apache on Linux (RedHat). Client usually Netscape 4.05. 
>Description:
This sound like the bug reported in the README and FAQ but all attempts to use the 
workarounds recommended have failed. Post to dejanews went unanswered.

I have a strange problem using Apache 1.2.6 on Linux. I have an applet
(JDK1.02) that is started from the server. Once the applet starts in the browser (usually Netscaep 4.05)it
contacts the server using a URLConnection to fetch a
binary file (a CAD drawing). This usually works fine.
In some cases (seems like Netscape 4.05 mostly) the applet gets an
IOException of "server status 206" or "server status 200" in the middle
of reading the file. The log file shows a '200' result but the file size
is too small.
This problem happens infrequently, i.e. the same browser may work a
couple of times then fail.

I set the "BrowserMatch" to force http1.0 response and am at a loss.
The log file shows a request using HTTP/1.0 and the response when
Mozilla is the user agent is HTTP/1.0

The attempts at workaround are as follows:

BrowserMatch Java1.0 force-response-1.0
BrowserMatch JDK/1.0 force-respons-1.0
#BrowserMatch * force_response-1.0
BrowserMatch Mozilla force-response-1.0
BrowserMatch Java force-response-1.0
BrowserMatch "^Java" force-response-1.0
BrowserMatch "^Mozilla" force-response-1.0
BrowserMatch "^JDK" force-response-1.0
BrowserMatchNoCase mozilla downgrade-1.0
BrowserMatchNoCase jdk downgrade-1.0
BrowserMatchNoCase java downgrade-1.0       

The last 3 lines have been added recently.

I suspect it's the "region" bug. Is there a way to disable partial returns via the server?

I have also seen this problem when the browser tries to load the archive containing the Java applet,
i.e. event before the applet starts. It fails to load the archive with "server status 206" then loads the classes one at a time.
>How-To-Repeat:
It's hard to repeat. It happens infrequently. a sample URL is http://www.viewcad.com/demo/0.html
It may happen more often if partial content is in the browser cache. i.e. stop the browser mid way through loading the drawing,
then reload. 
>Fix:
Disable the "Region" response? i.e. have a switch to allways return the full contents?

Thanks for the help. 
>Audit-Trail:
>Unformatted:
[In order for any reply to be added to the PR database, ]
[you need to include <ap...@Apache.Org> in the Cc line ]
[and leave the subject line UNCHANGED.  This is not done]
[automatically because of the potential for mail loops. ]
[If you do not include this Cc, your reply may be ig-   ]
[nored unless you are responding to an explicit request ]
[from a developer.                                      ]
[Reply only with text; DO NOT SEND ATTACHMENTS!         ]