X-Git-Url: http://www.privoxy.org/gitweb/?p=privoxy.git;a=blobdiff_plain;f=TODO;h=bdd88964b9a0f5c32d04587e59eb4ce0c7b1e341;hp=8eabebe12bd79483a48003eab0c15929c61f08ad;hb=b74fedb146c49ba5374965b1a0d38d74a920b520;hpb=0ae492a4a295b1f638ee251c6d47d4f5c648abb2

diff --git a/TODO b/TODO
index 8eabebe1..bdd88964 100644
--- a/TODO
+++ b/TODO
@@ -1,4 +1,4 @@
-$Id: TODO,v 1.161 2017/04/20 11:10:54 fabiankeil Exp $
+$Id: TODO,v 1.163 2017/05/20 09:25:14 fabiankeil Exp $
 
 Some Privoxy-related tasks, sorted by the time they
 have been added, not by priority.
@@ -414,7 +414,7 @@ http://www.privoxy.org/faq/general.html#DONATE
      forward-override). Investigate and fix or document.
 
 141) Port Privoxy to CloudABI, which, despite the name, is actually
-     rather neet. https://github.com/NuxiNL/cloudlibc
+     rather neat. https://github.com/NuxiNL/cloudlibc
 
 142) Remove or update the "internal" pcre version.
 
@@ -471,6 +471,14 @@ http://www.privoxy.org/faq/general.html#DONATE
 
 157) Add a directive to control the currently hardcoded receive-buffer size.
 
+158) Use a single thread to wait for new requests on reused client connections.
+     Currently the thread that handles the first request on a connection
+     stays responsible for the client connect until it gets closed.
+     In case of lots of idle connections lots of waiting threads are used.
+     While it's conceivable that this ineffiency is irrelevant from a
+     performance point of view, using a single thread should reduce Privoxy's
+     memory footprint a bit which may be noticeable in case of multi-user setups
+     with hundreds of idle connections.
 ##########################################################################
 
 Hosting wish list (relevant for #53)