Re: [gambit-list] Module boundary performance