Just like clients can use COMMAND INFO to find where the keys are for cluster support, I think there should be an info flag if the command can block, this can be useful in a context to know if it can be pipelined (multiplexed with other commands) or not.

Comment From: tzickel

I guess Redis 6 ACL provides this information and this can be closed (guess it won't be back ported to a normal command flags for some reason).

Comment From: itamarhaber

Actually, afaik, ACL doesn't let you know whether a command can block.

Comment From: tzickel

Seems like it does ?

127.0.0.1:6379> command info blpop
1) 1) "blpop"
   2) (integer) -3
   3) 1) write
      2) noscript
   4) (integer) 1
   5) (integer) -2
   6) (integer) 1
   7) 1) @write
      2) @list
      3) @slow
      **4) @blocking**

although I never understood why all this nice stuff couldn't exist in part 3 of the response (not just blocking)...

Comment From: itamarhaber

Lol - I must be getting old :)