Howard Chu writes:

Since 2.4 slapd dies on unknown config directives, they won't get very far before it becomes obvious what needs to be fixed.

Ah, OK. Sounds worthy of an ITS though. And it makes the rest of my last message a different issue:

...

If we are going to redesign the interface, it might be better do do it more thorougly. E.g. might to let backend perl perl-load foo load foo.pm which defines backend foo, usable with 'database foo'.

Huh?

Something like this - except I've likely forgotten enough Perl to get it quite right:

back-perl creates a Perl module back_perl. Possibly foo.pm inherits from it. foo.pm calls back_perl::new() which creates an ordinary new slapd backend. It looks up foo::<search, add...> and creates bi_op_<search, add, ...> for the Perl functions that foo defines. It does not set e.g. bi_op_add if perl method foo::add does not exist.

Somehow foo.pm must be loaded, and we'll never say 'database perl' so we can't do it there - but we can do it under 'backend perl'.