because b0rken MSIE still insists on fetching from its cache even though
it's been explicitly told not to.
+Revision 625.12 2005/09/18 14:30:05 ajc
+* Added a "random number" parameter to the wholist ajax refresh request,
+ because b0rken MSIE still insists on fetching from its cache even though
+ it's been explicitly told not to.
+
Revision 625.11 2005/09/18 04:04:32 ajc
* Switched the wholist updater to Prototype's AJAX updater, instead of
using Rico's wrapper around it. It's actually simpler and doesn't seem
Revision 625.11 2005/09/18 04:04:32 ajc
* Switched the wholist updater to Prototype's AJAX updater, instead of
using Rico's wrapper around it. It's actually simpler and doesn't seem
1998-12-03 Nathan Bryant <bryant@cs.usm.maine.edu>
* webserver.c: warning fix
1998-12-03 Nathan Bryant <bryant@cs.usm.maine.edu>
* webserver.c: warning fix
available network interfaces. Normally this will be the case, but if
you are running multiple Citadel systems on one host, it can be useful.
available network interfaces. Normally this will be the case, but if
you are running multiple Citadel systems on one host, it can be useful.
- -> http_port: the TCP port on which you wish your WebCit server to run.
- This can be any port number at all; there is no standard. Naturally,
- you'll want to create a link to this port on your system's regular web
- pages (presumably on an Apache server running on port 80). Or, if you
- are installing WebCit on a dedicated server, then you might choose to
- use port 80 after all.
-
+ -> http_port: the TCP port on which you wish your WebCit server to run. If
+ you are installing WebCit on a dedicated server, you can use the
+ standard port 80. Otherwise, if port 80 is already occupied by some
+ other web service (probably Apache), then you'll need to select a
+ different port. If you do not specify a port number, WebCit will attempt
+ to use port 2000.
+
-> tracefile: where you want WebCit to log to. This can be a file, a
virtual console, or /dev/null to suppress logging altogether.
-> tracefile: where you want WebCit to log to. This can be a file, a
virtual console, or /dev/null to suppress logging altogether.
- WebCit contains a small amount of graphics (icons, etc.) which are kept
+ WebCit contains graphics, templates, JavaScript code, etc. which are kept
in its "static" subdirectory. All site-specific graphics, however, are
fetched from the Citadel server.
in its "static" subdirectory. All site-specific graphics, however, are
fetched from the Citadel server.
"Server: %s\r\n"
"Connection: close\r\n"
"Pragma: no-cache\r\n"
"Server: %s\r\n"
"Connection: close\r\n"
"Pragma: no-cache\r\n"
- "Cache-Control: no-store, no-cache, must-revalidate\r\n",
+ "Cache-Control: no-cache\r\n",
*
* * setInterval() makes it auto-run this code every 30,000 milliseconds (30 seconds)
*
*
* * setInterval() makes it auto-run this code every 30,000 milliseconds (30 seconds)
*
- * FIXME b0rken IE is still caching it
+ * The random number parameter forces b0rken MSIE to fetch a new page instead of going to
+ * its cache, even though it's been specifically told not to cache.
+ *
*/
wprintf(
"<script type=\"text/javascript\"> \n"
" setInterval(\" new Ajax.Updater('fix_scrollbar_bug', '/who_inner_html', "
*/
wprintf(
"<script type=\"text/javascript\"> \n"
" setInterval(\" new Ajax.Updater('fix_scrollbar_bug', '/who_inner_html', "
- " {method: 'get'}); \", 30000); \n"
+ " {method: 'get', parameters: Math.random() }); \", 3000); \n"
"</script> \n"
);
wDumpContent(1);
"</script> \n"
);
wDumpContent(1);