Totally reworked ping timeout detection

[user/henk/code/ruby/rbot.git] / ChangeLog

diff --git a/ChangeLog b/ChangeLog
index 9c9e71201f8baa11696b6c157084c6f08d0a3c95..19fa2d14a5fb2625687065e82b18560f61572a8d 100644 (file)
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,36 @@
+2007-01-12  Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
+
+       * Server timeout: rework the server timeout code. Instead of PINGing
+       the server unconditionally every server.ping_timeout seconds, we only
+       PING it if we don't receive anything in the user-chosen timeout (lazy
+       PING). The code rewrite also seems to have fixed the "bot stalling
+       doing nothing" problem, which seemed to have been a consequence of
+       @socket.select not having a timeout.
+
+2006-11-01  Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
+
+       * SSL support: patch from Robin H. Johnson <robbat2@gentoo.org>
+
+2006-10-28  Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
+
+       * A-Z game: try to guess the word the bot is thinking of: every miss
+       helps by reducing the range of allowed words, until it's so small
+       around the correct one that you can't miss it.
+
+2006-10-27  Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
+
+       * Flood protection: first attempt at penalty-based flood protection.
+       This should make rbot much less prone to Excess Floods *and* still
+       serve normally without excessive delays.
+
+2006-10-25  Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
+
+       * HttpUtil: Strings returned by get_cached now have a cached? method
+       which returns true or false depending on whether the result was taken
+       from the cache or not; this can be exploited by plugins that parse the
+       result in some particular way, since they can now skip the parsing
+       stage if they cache the parse result.
+
 2006-10-24  Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
 
        * HttpUtil: initial implementation of proper caching based on