We’re working on improvements to the box and the bends of the cables are something we take into account. Also the power cable sticking out of the back.
As written elsewhere, it’s a reference device, which was then also used by other open source projects like OSMC: https://osmc.tv/2016/08/announcing-the-osmc-pidrive/
So being a fully open source project we are happy about any contributions, especially if you are knowledgeable in that field. In any case I will forward your recommendations to the team, so thanks a lot already! (I’m sure they also thought about these things so I’ll let you know.) :)
All 100% open source at https://github.com/nextcloud – and any contributions are very welcome! :)
Currently testing out an OrangePi Plus 2E strapped to a touchscreen to replace the all in ones we use for another product, with good results thus far.
"Have never done it before" would not be a description I would apply to Western Digital regarding hardware products.
The website admits however, that this is more of a "reference device, meant to inspire and invite anyone to build their own", so I would not expect the most rugged quality. In any case, as you need to bring your own Raspberry Pi, it seems to be more of a product for tinkerers than for enterprise-strength reliability.
1. Each cable has a bend radius recommended by the supplier. If this bend radius is exceeded (too small) then it is very unlikely that the cabling will function properly electrically, especially for high speed communications. Cross-talk, dropped packets, etc. If anything the supplier is off the hook for meeting their spec since it wasn't installed correctly.
2. The plastic insulation will take this for a while but two things happen to polymers with a constant stress on them.
2.a They suffer stress relaxation via creep. The long linked molecules will reach a new alignment over time. They will no longer be in their plastic range and settle in a new configuration. This could result in more crosstalk or even a complete insulation failure as a function of time.
2.b Flexible plastics typically slowly lose plasticity over time due to the breakdown of the plasticizer that made them flexible. (That nice new cable smell) Most cables have PVC made flexible due to these additives. If this box is warm due to the heat from the drive and pi, then it's entirely likely that the cable could further degrade under the constant strain and creep and fail. Causing not only thinning but brittle breaking of the jacket. It doesn't have to be visible breaking for it to stop working electrically.
Basically you can't trust your wire anymore. It could work forever. It could fail randomly. It could just cause communication errors that are hard to trace.
As for the box, they just used too much plastic and pushed the thing through the machine too fast without cooling. But then again, I am not an ultra expert on injection molding. It's an art. My recommendation would be "Have a sit down with your molder, these faults do not need to be."
I'm sure you might be right but it unless I'm missing something I'm not sure any of your points can be entirely proven until you know who sources the parts and what extreme(s) they can handle, no?
Considering the cables are made by the same company as the box itself (WDLabs) and that we've sold almost 1000 boxes and haven't had a single person tell us the cables were problematic and considering I've personally put together a dozen boxes and taken them apart and together again many times and still encountered no issues I think I can safely say: it is just fine.
Nobody could source cables with a right angle bend in them?