I wonder why everyone is making super unrealistic predictions. Maybe because there is no project plan? Just think about how many tasks are open. Assembly, Certification, Manual, etc. So end of 2018 is more realistic in my opinion. (but before R.I.P)
Certification takes one day and labs currently have a waiting time of about 4 weeks. A quick test (to check if your system complies) doesn't take as long, so you'll get a lab for that in less than one week usually.
The final prototype is what's needed for the certification and for the pre-check, all we need is the new motherboard (which Nikolaus is currenty soldering).
So if the quick check is successful, the appointment for the certification can be made and in parallel, the new CPU boards be produced.
Once that has happened, the boards are ready for mass production.
For the case production, the tweaks are the main things (though they should mostly be done now, according to my contact).
There's no real setup time for the new company. You attach the molds to the machine and optimize parameters like heat, pressure and cooling time with samples until there are no visible marks left.
Assembly takes 10 minutes per person per Pyra, which is similar to the Pandora, and GC could easily deliver 100 Pandoras per week (even more, if need be). That's simply a question of man power.
The manual will be similar to the one of the Pandora and that had been written in two days and printed in one week. The packaging has a production time of 2 weeks.
I'm doing similar things regularly, I'm handling the production of the Retrode and all plugins, am producing shells, manuals, stickers and packaging for Everdrives and am now also taking care of the production of Cinemaware games in Big Boxes with goodies. Oh and I published a few Dreamcast games, where I also handled the complete production run.
Nope, the issue is neither a missing organization nor am I concerned about the production itself too much.
It's perfectly planned already when certification, manuals and packaging will happen. But that's not based on a date but on the status of the project.
Yes, the hardware (PCBs) turnd out A LOT more complex and beastly like me or Nikolaus ever imagined. Just take a look how long it took to analyze the CPU and memory instability issues. Earlier hardware (OMAP3) wasn't that fiddly but a lot more tolerant to design - but I guess the reason for that is that everything gets smaller and faster...
Also, the best organization doesn't work well if a main player (like the case company) doesn't do their job as planned AND you don't have the money to simple kick them and move to a new one tod finish the design.
That will ALWAYS throw you back - and you can't do much about it.
There might've been a few organizational things I could've done a lot better, if you loom back, but that's life. No one always makes perfect decisions.
I am not aware of big organizational mistakes I made, but feel free to mention what you think I wrongly and let me know how you would've handled it to prevent any unexpected delays.
Sent from my XT1650-03 using Tapatalk