Re: extend pgbench expressions with functions

>> I guess the question here is whether we want the part-c patch, which
>> removes the historical \setrandom syntax.  That's technically no
>> longer needed since we now can use random() as a function directly
>> inside \set commands, but we might want it anyway for backward
>> compatibility.

This patch is indeed a proposal.

>> Anybody have an opinion on that?
>
> +1 for nuking it. That's not worth the trouble maintaining it.

I share Michaël opinion.

Some arguments for removing it:
 - \setrandom is syntactically inhomogeneous in the overall syntax,   and is now redundant
 - switching to the \set syntax is pretty easy, see attached script
 - custom scripts are short, they are used by few but   advance users, for which updating would not be an issue
 - the parsing & execution codes are lengthy, repetitive...

-- 
Fabien.