MWeston said:
Hey there,
This first 100 debate might not be as big of a deal as you might think. It's likely that the time difference between the first 100 and the next 1000 will only be one week if any delay at all. The first 100 will be going nowhere until the case is done and shipped since we are no longer shipping out bare boards to anyone. They will go through very rigorous testing, over and over again. They will be used to establish the test procedure for the next batches of thousands. It's even possible that all 4000 will be done testing at the same time the cases show up and they will all start going out immediately after being put together.
Thanks MWeston. I always appreciate these longer more informative posts over shorter, more emotive posts
It sounds like there may be precious little difference between the 100th "finished" Pandora and the 101st one - and so on. Like possibly a matter of seconds. I'll come back to that point, if I may.
MWeston said:
I would say 95% of the parts are in Texas this week and the remaining parts (some connectors) should finally ship this week with the taxation stuff better understood. The case molds should be going to production this week as well. Dave was hoping to have it done this past week but it didn't happen so it seems reasonable to assume an extra week is all he will need.
Cool - so hopefully only one more week for the D-Pad modification
(I'll come back to this if I may).
MWeston said:
The mold creation is the long 45 day process that will involve several reviews and tweaks but we are getting the mold company involved with this so that the time is used as efficiently as possible. We're not making anymore CNC/SLA prototypes because it seemed like every time we got one, something new was not quite right on it.
I THINK I understand, but even so I must admit that I am slightly worried by/about this. The most... what word to use? The most contrary way to read this is that you keep getting it wrong so you're going to stop doing prototypes and just produce the "real thing". So... what's to stop "the real thing" from being "wrong" if you can't get a working prototype right? Apologies of that sounds overly pessimistic, but I believe it's an honest (if as I say deliberately as contrary as credible) reading of what you seem to be saying.
MWeston said:
It got to the point where we had to ask ourselves if we made the error, or if they did!
I'll come back to this later, if I may (this seems to be a recurring theme!).
MWeston said:
This could go on for years if we are attempting to trust prototypes to get it exactly right. In fact, I don't think any company does this. I think we are to get it very close and fitting together and then allow the plastics company to make it perfect. I'm sending them everything they need to build a Pandora unit so that they can tweak the molds to have the tightest tolerances and guaranteed fit for the tens of thousands of units that will get squeezed out after this. By taking us out of that equation, there is no need to ship us anything or get approval on little things that could add weeks to the process. We stay involved in the major decisions and wait for real test units to ship to us for approval before pumping out thousands more.
Hmmm... It seems odd that you're not sure who is responsible for CNC model problems - them or you - so you're going to turn everything over to them. I hope you have some contractually binding bits of paper, signed in blood, that they have to actually meet the standards required. With no wriggle room. I've no idea how many mass production molds you guys could afford to buy, but I'm assuming it's more expensive than CNC cases. Because it seems that this puts a great deal of faith/reliance in their laps.
MWeston said:
Based on these calculations, it looks like the end of August to start seeing fully assembled units and then shipping to customers in early September. Having to ship things back and forth around the world really sucks, but it's unavoidable on this first batch due to taxation laws. Only the developer units provided free of charge to those involved with Pandora's development can ship direct so, in a way, shipping to developers first is the only way this can be done without needlessly wasting time.
Ummm. Only devs who are getting free Pandoras make sense? If there's little to no time difference between the first 100 and second 100, I can't see why at all. Especailly not from the point of view of someone who actually paid for the device. Of course, if you want to refund me and give me a freebie, I will happily see this from another perspective and agree that shipping out freebies is a priority
Seriously though - ignoring the appearance that Craig's query was of less value than we've given it (no comment) I often see "it doesn't matter/it's not important/there's little to no difference" used as an argument to favour one route over another, and I've never understood it. It's less clearly being used here, but still being used, and I've never really understood the logic. If there's so little difference between the first 100 units and the next 3900 that it shouldn't matter to the 3900 people, then why can't that be viewed in reverse to say that there's so little difference that the 100 devs - 100 freebie devs - can't wait while all paying and unpaying customers get their units delivered in "purchase order" order?
Maybe in the seconds, or hours, or days - even a week or two! - between the first 100 and the next 100, the freebie devs will do something useful? Looking at the progress to date - like the weeks to get the WiFi driver working (I hope that will happen) or the weeks to get the D-Pad design perfected (I can think of few things as crucial in a handheld gaming device, so this is not meant as criticism) what exactly will these freebie devs DO in the seconds, minute, hours, days or perhaps even weeks between getting their Pandoras and the paying public getting theirs? What is it they can do that they cannot do wtih their beagleboards or devboards or whatever in that tiny gap in time? Can they REALLY compile, test, patch/fix and optimise hundreds of games specificallt for the Pandora and Pandora-specific controls that they cannot do with any other equipment? Is there so little software "ready to go" now, or does it take weeks to update software to use Pandora's controls?
Of course, the longer the gap between the first 100 and the rest, the more devs could get done - but also the larger and more important the gap would be, and the more feedback good beta testers could give. So I'm torn between "the gap is so insigificant it may as well not exist" in which case I don't see a good argument for not supplying the units to people who have paid for them, and "the gap will be large enough that feedback could be provided that could affect the rest of the units before they are shipped" - in which case it seems vital to me to supply some of said units to some end users willing to knock it about a bit. If TAXes worry you too much, stick to customers in the UK perhaps?
{scratching head}
So - what's your plan if you get 4000 cases that NEARLY fit? What's the backup plan, in other words? (I am a great believer in having a "Plan B" for every situation - it means you rarely if ever get stumped by things not working as you switch immediately to the next step).
Apologies if I seem needlessly combatative or contrary, but as I say "It's too insignificant to make a difference" always seems to me to swing both ways, so I've tried to be as clear as possble about my awesomely cyncial reading of the situation, which may make it seem more rant-like than it should, and less puzzled than would be appropriate
Ta,
its still up to them to decide 
