On 2 Oct 2012, at 22:20, Ruth Ivimey-Cook wrote: The ease of writing this stuff is why (correctly) Peter is suggesting that Occam-next has explicit support for protocols, rather than just individual communication, which IMO should have been added much earlier.Which is why I advocated rooting the only widely applicable safe protocol – that of Brinch-Hanson's client/server (or master and servant, as he described it) – in the core of Honeysuckle : not just the connections, but the way you use them is specified (at least in part) ; a guarantee against the most common cause of deadlock can then be offered. I'm delighted Peter has come to agree. Ian Ian East Open Channel Publishing Ltd. (Reg. in England, Company Number 6818450) |