He makes stuff that goes into space for a living and understands the design trade off on hobby vs not hobby.
He may well do, but I sincerely hope he has a technician to do his soldering for him.
But more seriously, if I was up for
"bodge it with turned pins with ribbon cable soldered to it" I wouldn't have asked for folks experience's with a particular latched, crimped connector I'd have just gone ahead and done it with ribbon cable and turned pin and socket strips - I even have all three in stock. My own experience of make-dos like that is that they are
anything but reliable. In fact back when I was originally playing with the GPS receiver I'm using I had flying wires soldered directly to the hard metric board to board socket that it uses - I've lost count of the number of times that "
wire broken off socket stopped play" happened. It was at least three, and this was just "try and talk to the thing", not a production version.
I don't wish to sound churlish, I hope that it's fairly obvious that I'm looking for something a bit neater and more reliable. Actually I'd rather solder individual wires direct into the board than rely on ribbon cable flapping about on connectors that aren't designed for wire-to-board use. I do use turned pin socket strips for board-to-board connections, but that's what they are designed for. On something that's probably going to run to a "rev B" I want the main board to be swappable in a fairly modular fashion. The PSU, hardware, LCD and indicators, and coax connectors ought to work flawlessly first time. If I do need a "rev B' I don't want to have to desolder all of them from the main board just to swap it. Especially if I want to do some closed case runs with "rev B" and "rev A" in relatively rapid succession.
I could have gone with classic ribbon cable on 2 row 0.1" IDC plugs, shrouded sockets and transition connectors but they're a bit bulky and use up a lot of board space outside of the contact area, I'd have struggled to get them to all to fit on the board
and keep the routing anything like the way I wanted it - it's already too messy for my taste. I'm heavily constrained by the keep out area necessitated by the GPS module, the bulk of the OCXO and the need to get 4 coaxial signals off the board as well. Hopefully by the time I'm finished I won't have contaminated every signal on the board with the 10MHz OCXO clock, and the 12.7 MHz odd clock the GPS module uses.
Despite Cymaphore's reservations about the Nano-Fits I think I'll give them a try. They're reasonably cheap (the headers are in the 80p - £1.50 range for the sizes I need), reasonably skookum and a better fit for what I'm doing and how I'm trying to do it than anyone else has suggested.