We spoke to the factory over night, everything is cool. Those errors were because the machine was not warmed up/tweaked/cooled. They would never mass produce something with such errors.

They just assumed we would know that.

So, green lighted.

(craigix)

Source

Update: EvilDragon elaborates on the green light and reflects on one of the most tumultuous weekends the project has seen, in a new post on the official blog. Congratulations, team Pandora.

Introducing the Pandora Angst Coding Competition, an adventure in creativity/distraction devised and funded by the community. Entries are invited for GP32, GP2X, Wiz, and Dingoo (native or Dingux) games & applications, with a current prize pool of around $400AUD on offer. If you’re not a coder but you’d like to get involved, you can donate to the prize pool, offer your services as a judge, contribute splash screen artwork, or hook up with a coder and collaborate!

For full details & entry form, head to the 2010 PACC website!

[EDITORIAL]
It’s been a fun weekend, hasn’t it? Ok, perhaps showing us case samples on a Friday, giving us a whole weekend to wait for answers from the factory, didn’t work out all that well. But information has surfaced in the last two days that gives us cause not to freak out. To fully understand what happened here, you need to fully understand injection moulding – which most of us don’t. That means it’s time for a ludicrous cake analogy. The chef has been asked to do some quick tasters. He knows the recipe, and he’s got a totally sweet oven. But in his rush to please us, the cupcakes came out a little sloppy. And although we asked for icing, he didn’t want to mix up a metric ton of the stuff for a handful of cakes.

Injection moulding facilities are designed to pump out units in the thousands. It’s not cost effective, in fact it’s borderline impossible, to provide a showroom quality batch of seven units. The production line needs time to warm up. Settings need to be bumped on the fly. In a run of 10,000, the first handful are likely to be sub-par. Why then, were we shown panic-inducing photos of half-baked samples? Because we demand to see them, for better or worse. It’s the nature of this project, and it’s not always a picnic. What we need to understand (and we’re all learning this together) is that sample cases are not necessarily representative of what a large scale run will produce. This is why most companies don’t share such information.

Onward to some forum quotes, starting with some expert testimony from Madcad77.

From what I have seen of these new mouldings, your moulding company needs to run a small batch to validate thier parameters, things like mould temperature, injection pressures, cycle times etc.

These coupled with venting added in the correct positions (tiny modifications 1- 2 days max) will sort out your sinkages, short fills etc.

The greasy finish you mention is most probably due to excessive release agent used when running the mould for the first time (reduces the risk of the moulding sticking in the tool).

I have been designing plastic components for over 10 years now, and alot of your issues with Chinese moulding companies are well known, regardless which company you chose over there, they just work diferently to us.

I am very optimistic though, when these parts are run to the correct production processes, perhaps with extra venting, your cosmetic issues will be resolved.

(Madcad77)

I tend to agree with you here since there has to be a reason for the way things are. I’m still up assembling units as I’ve made some promises for the week and I am noticing that several of the parts feel like they are almost dripping with grease (release agent). I even started wiping a few off. These came right out of the mold and into the shipping box.

(MWeston)

Source

Further reading: This recent post from EvilDragon, and a rather good post on Screeny’s German blog which contains some direct communication from MWeston (note: you’ll need to use the Translate button).

Let’s just be patient, and let the team get more answers before we make any assumptions about where we are at.

The case photos we’ve been waiting for have appeared in MWeston’s album. As you will have gleaned from the various tweets and forum posts in the last 24 hours, there are once again some imperfections that have left Michael and the rest of the team vexed to say the least. Before we get ourselves in a twist about further mould “tweaks” though (hands up if you now hate that word), it has been pointed out that the cases are structurally, and functionally, perfect. The source of vexation here is that the finishing on the cases appears to be no improvement over the revision 2 sample.

It would only be natural at this point to ask what in tarnation the case factory has actually been doing since rev2, if not improving the finish (and ask it MWeston did). But the answer to that is not necessarily nothing, and a couple of possible explanations are being explored. Production cases go through a finishing process after they’ve been moulded, to give them that nice, even matte sheen. It’s likely that the case factory doesn’t deem a small sample batch worthy of this extra effort (even if it’s been asked for). Another likelihood, and not a mutually exclusive one, is that recycled plastic pellets have been used to make the samples; a cost saving measure which can result in uneven colour. If both of these scenarios turn out to be the case, we can pretty well say that they won’t be issues when the factory makes a “real” batch. But they are questions that still need to be answered. Fatih and his trusty translator will no doubt be doing some overtime on this one. (Update: They will be seeking answers on Monday.)

Read on for photos and all the words from MWeston. And just a little bit of advice when you’re looking at the photos – be careful how you judge them in super high resolution. Not many moulded plastic cases will stand up to scrutiny at this level, not even from the big guys. Check out the details by all means, but zoom them back out to “normal” size for a more realistic indication of where the issues are.

The major part of making cases isn’t necessarily molding the parts, but carefully packaging up the parts for shipping. Final parts don’t get shipped like samples do which is crammed into plastic bags inside one big box.

(MWeston)

Source

I’m [trying] to find out if these were made from recycled plastic or new pellets. Maybe that is why it still has rough edges and a greasy finish and a production run will be done differently.

(MWeston)

Source

Well I made it through 17 pages of this topic now so I should post a few comments.

First, I am not looking for perfection. That is an infinite process but there is such thing as a standard quality that any commercial device should meet. I spent so much time perfecting the guts of this thing and it is a true insult to my efforts to see this mold company send me such incomplete samples and call them ‘final’. We will find out what level of purity the plastic was that made these cases. Maybe they know production plastic won’t be swirled and greasy and so when they said we would be happy with these, they expected that we understood that. Communication is a tough one when they refuse to hold conversations for more than three minutes at a time. Don’t ask me why they didn’t put the silk screening on these parts because I was explicitly told they would have them. More lies.

There were a few more internal modifications for rev3 but most of the important ones came from the rev2 samples. A couple of changes for rev3 that I am happy about are a tighter fit on the screws which makes them feel more secure and closing up the nub surround so that the PCB isn’t visible around the nub under certain lighting conditions. It also looks like they closed up a small gap in the lid that let a bit of backlight light shine through the front. The LCD cable slot took a step backwards with a bit of unevenness but since it was right last time, we can assume that more carefully made production parts will be back to normal.

The cases are still very solid. You can try to twist and bend them but they don’t flex much. It will feel like a very solid device in your hands and functionally, everything works exactly as we want. The gaming controls are perfect. The keyboard is perfect and there has never been an issue with buttons getting stuck in the case or anything like that. The problems now are cosmetic. There are rough edges around the connectors, around the edges of that recessed channel in the front the SD cards, volume wheel, etc. and some on the edge of the base parts where the two halves go together. There are some sink marks too which might hide a bit better with a full, proper matte finish but they’re there.

As some people have pointed out, a picture can say more than any amazing piece of software ever will and a reviewer crapping all over the case quality could spoil it for anyone considering the Pandora. I’m one of those people really. These cases aren’t a disaster but I would want something to show off to my friends and having to explain the colored patches and rough edges is a little defeating.

Some pictures are posted on my photobucket and I’m not too sure making a video will show better quality. I will probably still make one anyway but unless I can really master the lighting, my 640×480 videos won’t help you know if I’m holding a rev1, 2 or 3 case.

EDIT: Two more pages were written since typing this. Ah yes, the hinge area. There are clips in that area that hold the two halves tightly together and they screwed up somehow. On this revision, the recessed spots for the clips to sit in got completely evacuated and there are holes right through the case. They didn’t do this on rev2. Rev2 just had roughness on the bottom edges of these parts and not the destruction of the clip channels like this time.

(MWeston)

Source

From the bucket of Weston.

Update: EvilDragon elaborates on the official blog. A little.

Wow. Actually, there’s something that happened earlier than expected: The cases did arrive, including the final keymats.

Here’s a picture of the final keymat. Now please give Michael some time to thoroughly test everything out.

Update 2: Mr Weston will be putting together some photos (and hopefully a video) later today, detailing what went right and what went not-so-right with this batch of cases. At the time of writing this, it’s some time before dawn in his neck of the woods, so the next few hours will probably a waste of your F5 miles. Sit tight.

Fresh from the twitter of Craigix:

The cases have arrived early, wow, some good luck for once!

When I say early I actually mean weeks late I suppose.

Well, not perfect, but I guess we will have to go with them.

Sit tight, grab some popcorn, we’ll bring you updates as quickly as we can.

“Ah, breaker one-nine, this here’s the Rubber Duck. Looks like we got us a convoy.”

Missed a few status updates of UPS it seems. Anyways, the most important (and current one).

Loc.: WINNIPEG, MB, CA — Date: 05/02/2010, 5:46 — Description: REGISTERED WITH CLEARING AGENCY / RELEASED BY CLR AGENCY. NOW IN-TRANSIT

Source

From ze Twitter:

Status: In Transit – Rescheduled Rescheduled Delivery Date: 08/02/2010 Would’ve been nice before the weekend, but won’t happen sadly.

But it’s not all bad news, as MWeston has kindly informed us:

It might bounce back to Friday once it hits the US border (Alaska). These China shipments change the delivery date once or twice along the route every time. Sometimes they change back to the original delivery date, sometimes they don’t. We’ll know by tonight.

Tense?

Update: There’s a new post from ED on the official blog. It covers stuff you already know, but here’s the link anyway.

Update: Image courtesy of Screeny

Just a quick update on the cases.

Your package is on time with a scheduled delivery date of 05/02/2010.

If the cases reach Micheal by the 5th, he will have ample time to test them before giving China the green light. If China receive the green light by the 8th, they believe that they can finish all 4000 cases before Chinese new year.

Source

A micro update from EvilDragon’s Twitter account.

We got a tracking number! It’s not yet in the UPS system (always takes some time, even longer in China). Good news, I’ll keep you updated!

End micro update.

If you needed a reminder that China isn’t the only player in this saga, this is a nice way to get one. MWeston has informed us that 1,000 boards are being populated with components this week in Texas. Also interesting is the impact certain software developers are having on the hardware production process. What, you thought they just sat around porting Quake? (No offence to those who sat around porting Quake. You’re all doing a bang up job.)

It’s worth saying again that Notaz’s work on the kernel is clearly the reason the device is as advanced as it is today. Sure the hardware may have come along fine with minimal testing as the assembler learned how to solder the boards better but Notaz has helped me develop a comprehensive suite of test software that has helped the assembler perfect their process. So good in fact that they have started building a batch of 1000 this week after tweaking their pilot runs.

(MWeston)

Source

Fatih and his trusty translator are making their presence felt in China, and they’re all out of bubblegum. EvilDragon reports via Twitter:

Fatih and his friend are constantly phoning up the chinese company to make them work faster. Not easy dealing with chinese companies…

Fresh news from Fatihs chinese speaking friend. He phoned with the company.

he wanted me to explain you that you will thank him when you see the case he had to make a lot of working on current mould. and tested a lot

“now they are 100% sure, it’s perfect. he will send the samples tomorrow morning.” It’s night in China already. Let’s hope…

In between tweets, ED also made a long post on the German boards which was hurriedly transported to GP32X and dissected by a team of crack multilinguals. The salient facts from this post:- Production at the case factory runs up until Feb 13, at which point they break for Chinese New Year. They need the green light by the 8th in order to produce 4,000 cases before they hit the red cordial for two weeks. Confidence in MWeston’s green light is high, so if they get samples into the mail tomorrow we could slip in before the dreaded CNY. There is also some kind of plan B that involves Fatih flying to China wielding a large stick.

Read the first translation here, and scroll down for more.

Thanks to tipsters & translators (naw)mcx, mali, Vlynndar, Bosbeetle, and mcobit. Go team! Twitter Dragon image borrowed from this guy.

Click here to view the embedded video.

Our favourite preserved vegetable has resurfaced, bringing with him nine minutes and fifty one seconds of glorious handheld gameplay. The titles in order of appearance are Ur-Quan Masters (aka Star Control II), Hammer of Thyrion (aka Hexen II), and Pandora Panic (aka Pandora Panic).

Click here to view the embedded video.

Cpasjuste has uploaded a video of Android running on the Pandora. Android is an operating system that was originally developed by an independent company and later brought by Google. Android is a popular operating system for smart phones.

This is my first attempt to port android to the pandora, touch screen support in the work.

-cpasjuste

Source