Re: Lock compatibility matrix

Oleg Bartunov <oleg@sai.msu.su> writes:
> Besides formatting improvements, it has addtional lock with
> temporary name UPDATE EXCLUSIVE (UE), which is the same as 
> EXCLUSIVE, but doesn't conflicts with SHARE UPDATE EXCLUSIVE (SUE),
> which aquired by VACUUM and autovacuum. The reason for this is that
> at present we have no lock mode, which doesn't conflicts with *vacuum.
> The problem was described in thread 
> http://archives.postgresql.org/pgsql-general/2006-12/msg01476.php
> What is the reason why we don't have such lock ?

I don't think the case was made that we need one.  There was certainly
nothing in that thread that I found convincing.  My opinion is we have
too many lock modes already ...
        regards, tom lane