From 171ef6813d54882e4eebd018183f382f74db9102 Mon Sep 17 00:00:00 2001 From: antirez Date: Mon, 2 May 2011 00:12:37 +0200 Subject: [PATCH] TODO updated --- src/SCRIPTING.txt | 25 +++++++++++++++++++++++-- 1 file changed, 23 insertions(+), 2 deletions(-) diff --git a/src/SCRIPTING.txt b/src/SCRIPTING.txt index ff91a59e..2d16ac9f 100644 --- a/src/SCRIPTING.txt +++ b/src/SCRIPTING.txt @@ -1,6 +1,27 @@ TODO: - redis('get',1) => crash - Check what happens abotu MULTI/EXEC + MULTI/EXEC/...: should we do more than simply ignoring it? + Prevent Lua from calling itself with redis("eval",...) + Implement log() function, and define REDIS_WARNING, REDIS_NOTICE, ... + + SCRIPT FLUSH or alike to start a fresh interpreter. + + Max exection time for scripts, even pretty high, but in order to + prevent that an instance gets completely freezed by error. + Hint: Use a debug hook for this (lua_sethook() with LUA_MASKCOUNT). + +Blog post TODO: + + Explain that for now this is just an experiemnt and there is no way to + tell if this will ever enter Redis stable, but that since it is pretty + self contained it is trivial to rebase with the unstable branch. + + Explain how replication / AOF mix with scripting and why it is this way, + and that scripts should be written to don't depend from timing or other + external events if you want to use AOF / replication. + + Show how fast it is! + + Mention that dynamically generated scripts are not a good idea.