Project update 75 of 75
Long story short, Mike’s successfully done a production run of 100 EOMA68 A20 Cards, and sent two of them to me in the UK and five to Chris of ThinkPenguin, for testing.
Here’s the longer version. After Mike’s manager quit with zero notice taking all of the manufacturing knowledge in his head with him; after the replacement manager was unable to create working cards using the PCBs that turned out to be sub-standard; after Mike decided to pay out of his own pocket for newer, more expensive, higher quality PCBs; after moving the entire PCB assembly factory equipment to a larger warehouse a couple of months ago… after all that, we decided to do a short run of 100 Cards.
The reason for doing a short run is very simple: if they turn out to be faulty it is not the entire remaining community-funding cash budget gone. Given that the last run (using the PCBs that we had no idea would turn out to be sub-standard) did in fact fail, requiring 50 units to be discarded, this is just practical risk mitigation.
Both Chris and I will test them, and, if they boot up from microSD,
we’ll go into full production. I will then give Mike some preliminary
testing procedures (e.g.,
lsusb), then they’ll be sent to Chris,
along with version 1.7 of the Micro Desktops, for full testing and
Remember, please don’t ask yet for a schedule — you know the answer is, "when all unknown and unknowable issues have been identified and resolved." It’s only when we are on a regular shipping schedule and have stock that such questions can be answered immediately and with confidence.
We’re nearly there: I am so grateful to everyone for being supportive, patient, and understanding. In particular to the anonymous person who donated 0.8 BTC some time a year ago. This basically pays outright and in full for the testing of all of the cards, plus the boxing and assembly which will be done by ThinkPenguin. Whoever you are, you made that possible, not just for me but for everyone.
Remember that this is not about the product itself. Yes you receive a manufactured good (a Computer Card), however what you’re really doing is helping shift the balance of power in computing, from a position of pseudo-ownership and corporate control to a position of upholding your rights to repair, repurpose, redesign, refurbish, reuse, replace, and much more.
Dave Hakken’s Phonebloks (900,000 supporters) and the Intel Compute Card (a knock-off of EOMA68) showed that both end users and OEMs are hugely excited by the potential, flexibility, cost-savings, and reduced carbon footprint of modular computing.
However, Dave’s initial enthusiastic support of Google Project Ara turned to serious annoyance when Google and select partners in the "Open" initiative created proprietary and closed chipsets (based around the patented MIPI UniPro), and Nexdock‘s $200,000 USD high-profile and widely reported failure of the Chinese ODM to be able to cope with the DRM built into Intel Compute Cards led Intel to shut that down, too.
These are expensive lessons that large corporations seem pathologically incapable of learning, meaning that it is down to you and I to create something different, and show the world how it can be done.
We may be close to delivering the EOMA68 A20 Cards, the first in the series, however the journey to bring mass-volume, inexpensive, carbon-conscious eco-computing to the world is just beginning, and your continued support makes that possible.
Update: an article in the Guardian newspaper shows the clear extent of the problem. Apple (like FairPhone) firmly and devoutly believes that their products are helping, because they’re made from recyclable materials. Whereas, Slashdot readers of the Guardian article show that, actually, people don’t want their product recycled per se: they want to keep their old one running. They want to easily repair it, replace the battery, and keep on using it for another 5-8 years. The Fairphone II product is indeed easily repairable, but unlike Desktop PCs you can’t upgrade the processor or the RAM. Right to repair legislation in the EU has just passed, and will in no way be enough, causing more problems than it tries to solve, because, fundamentally, products are "designed for manufacture" (assembly, not disassembly), and to be short-term. After all, why make a durable product at all? A durable product adds your own old products to your list of competitors, making it a hazard to any corporate bottom line! Modular upgradeable Computer Cards at least allow companies to sell upgradeable Computer Cards.