That's the weight of 2000 cases that are packaged up and leaving for UK tomorrow via air.
What - why tomorrow? They were packaged up and finished last week, weren't they?
Yeah... but guess what: There was the Moon Festival - and naturally, all exporters wanted to get THEIR stuff on the plane, and that tripled the shipping costs.
We won't get any more boards before the new nubs have been mass-produced anyways, so we decided to NOT pay a fortune to ship them.
Wait - no more boards until the new nubs are there?
Right. The tested nubs worked better than the ones before, but they still break more easily than the final ones. Using those slows everything down, as we need to desolder broken ones from otherwise working boards, and quite a few might fail after a few weeks, too, which would've lead to a lot of RMA costs and bad reputation.
The last testing prototype of the nubs lasted 7 times longer already in the stress test, the new one is already being tested right now. We aim for at least 500K cycles (last one had 300K, the first ones under 100K). Hopefully we get more good results of the nubs until the end of the week (well, testing takes some time and can't be sped up
)
Hey - does that mean: If I have I Pandora, I need to ship it back to get the nub replaced?
No. While the amount of breaking nubs is higher than it should be, it doesn't mean your nubs will break for sure. As long as they work fine, there's no need to replace them.
Okay... so what's currently happening and what's the plan?
Once the cases arrive in UK, Craig and team are preparing the cases so that we only need to pop-in the boards when they arrive while I keep soldering the LCD-Cables.
I'm pretty fast with that already and can solder 100 in two hours. If I solder 10 hours a day, it's 2 1/2 days for the rest of the cables (got about 1500 finished so far).
The company in Texas keeps producing and testing boards without nubs. Hopefully they stick to their promise that they finish them all until October 18th - but until now, they still claim it.
When the nubs arrive, they will be soldered onto the boards and the boards will go to UK ASAP to be put together into final units.
Some other news, regarding software
I started to work on Hotfix 5. I already fixed a few bugs and added some features and are now extensively working on the Tweak settings where you can set the OPP level, default clock, maximum overclock, etc.
I also plan on including a theme selector for XFCE4 (or better: GTK+) which allows people to install themes with bigger scrollbars, etc. more easily.
I also started to rebase our current OS on the latest OE Testing tree... which means: The whole system will be updated to a newer version.
I have no idea yet how this will workout, but it's something I want to try. DJWillis and notaz are helping me there, so some bugs like Midori rendering problems or the non-working detailed view in Thunar will most likely be fixed then.
TI has also updated the 3D drivers, though I don't know how fast we can include them, but we'll work on that, too.
But much more interesting: If you check this thread on gp32x.de, you can see that TI actually released a demo version of the Adobe FlashPlayer for the Cortex A8. And yes, it runs on the Pandora!
It's still slow, but it's just a first demo release, TI will most likely improve it. As Flash is running way faster on the N900 (which uses the same CPU), we know there's still a lot of potential.
Well, that's it so far.
Any question / feedback / etc. welcome as usual
What - why tomorrow? They were packaged up and finished last week, weren't they?
Yeah... but guess what: There was the Moon Festival - and naturally, all exporters wanted to get THEIR stuff on the plane, and that tripled the shipping costs.
We won't get any more boards before the new nubs have been mass-produced anyways, so we decided to NOT pay a fortune to ship them.
Wait - no more boards until the new nubs are there?
Right. The tested nubs worked better than the ones before, but they still break more easily than the final ones. Using those slows everything down, as we need to desolder broken ones from otherwise working boards, and quite a few might fail after a few weeks, too, which would've lead to a lot of RMA costs and bad reputation.
The last testing prototype of the nubs lasted 7 times longer already in the stress test, the new one is already being tested right now. We aim for at least 500K cycles (last one had 300K, the first ones under 100K). Hopefully we get more good results of the nubs until the end of the week (well, testing takes some time and can't be sped up
)Hey - does that mean: If I have I Pandora, I need to ship it back to get the nub replaced?
No. While the amount of breaking nubs is higher than it should be, it doesn't mean your nubs will break for sure. As long as they work fine, there's no need to replace them.
Okay... so what's currently happening and what's the plan?
Once the cases arrive in UK, Craig and team are preparing the cases so that we only need to pop-in the boards when they arrive while I keep soldering the LCD-Cables.
I'm pretty fast with that already and can solder 100 in two hours. If I solder 10 hours a day, it's 2 1/2 days for the rest of the cables (got about 1500 finished so far).
The company in Texas keeps producing and testing boards without nubs. Hopefully they stick to their promise that they finish them all until October 18th - but until now, they still claim it.
When the nubs arrive, they will be soldered onto the boards and the boards will go to UK ASAP to be put together into final units.
Some other news, regarding software
I started to work on Hotfix 5. I already fixed a few bugs and added some features and are now extensively working on the Tweak settings where you can set the OPP level, default clock, maximum overclock, etc.
I also plan on including a theme selector for XFCE4 (or better: GTK+) which allows people to install themes with bigger scrollbars, etc. more easily.
I also started to rebase our current OS on the latest OE Testing tree... which means: The whole system will be updated to a newer version.
I have no idea yet how this will workout, but it's something I want to try. DJWillis and notaz are helping me there, so some bugs like Midori rendering problems or the non-working detailed view in Thunar will most likely be fixed then.
TI has also updated the 3D drivers, though I don't know how fast we can include them, but we'll work on that, too.
But much more interesting: If you check this thread on gp32x.de, you can see that TI actually released a demo version of the Adobe FlashPlayer for the Cortex A8. And yes, it runs on the Pandora!
It's still slow, but it's just a first demo release, TI will most likely improve it. As Flash is running way faster on the N900 (which uses the same CPU), we know there's still a lot of potential.
Well, that's it so far.
Any question / feedback / etc. welcome as usual
Last edited:
