The amount of people that design and repair electronics is significantly smaller than the people that use electronics, that doesn't mean RasPi should ignore them. They boast opensource development, but seem to be working counter to that.
It makes is substantially easier to work with known design docs, and RasPi want people to use their hardware in embedded applications.
It's not like you can replicate their hardware with just the schematic.
Its awesome to see other filling the gap, but its a gap that has no need to exist.
The original mission was to get kids into computing with a low cost board, something akin to the C64 back in the 80s. But apparently that is only as user and not as a creator. I'd argue that getting kids into computing at the hardware level is even more important now.
But I think in some small part the RPi Foundation has captured itself and turned into a for profit company where the original mission takes a back seat.
Out of lack of understanding: what does this new knowledge allow? Will it allow open-source developers to write better drivers? Or does it allow the making clones? Or nothing in particular?
This is a PCB layout and board schematics. So this helps component level repair (uneconomical in first world) and diagnosing some of the weirder failures or hardware limitations. It also helps make a stronger case that RPI Foundation should have released this in the first place.
You could use this to make clones - but only if you could source at least the BCM2712 SoC, preferably RP1 too. I can't imagine that happening in practice.
Yeah they've become a real commercial outfit now. You could see this already during the shortage period, they always prioritised system integrators. I stay away from them now. It was nice while it lasted.
ten (10) layer board! :o Following twitch stream where TubeTime was doing it live gave some interesting insights, afair it did look like layout was done by more than one person and some things seemed to suggest time was more valuable than optimizing for less layers.
The elephant in the room: Why is it not Open Source Hardware to begin with?
NDA's with Broadcom.
Non-answer. BCM is far from the only vendor.
Broadcom has its hand so far up RPI Foundation's ass they might as well be conjoined.
True, this is the real answer.
The number of people who care is small enough not to affect any outcome.
The amount of people that design and repair electronics is significantly smaller than the people that use electronics, that doesn't mean RasPi should ignore them. They boast opensource development, but seem to be working counter to that.
It makes is substantially easier to work with known design docs, and RasPi want people to use their hardware in embedded applications.
It's not like you can replicate their hardware with just the schematic. Its awesome to see other filling the gap, but its a gap that has no need to exist.
I agree that they shouldn't be ignored, but... how to compel companies to share schematics, in the absence of enlightened legislation?
The original mission was to get kids into computing with a low cost board, something akin to the C64 back in the 80s. But apparently that is only as user and not as a creator. I'd argue that getting kids into computing at the hardware level is even more important now.
But I think in some small part the RPi Foundation has captured itself and turned into a for profit company where the original mission takes a back seat.
Yeah. The IPO really baked in some negative things.
Discourage cloning.
Impressive!
Out of lack of understanding: what does this new knowledge allow? Will it allow open-source developers to write better drivers? Or does it allow the making clones? Or nothing in particular?
This is a PCB layout and board schematics. So this helps component level repair (uneconomical in first world) and diagnosing some of the weirder failures or hardware limitations. It also helps make a stronger case that RPI Foundation should have released this in the first place.
You could use this to make clones - but only if you could source at least the BCM2712 SoC, preferably RP1 too. I can't imagine that happening in practice.
It has. At one point HardKernel "found" some Broadcom chips and made a one-off unauthorized batch of the Pi-compatible Odroid-W. Rpi weren't amused :)
Yeah they've become a real commercial outfit now. You could see this already during the shortage period, they always prioritised system integrators. I stay away from them now. It was nice while it lasted.
Thanks!
This helps placing an RP1 on a PCI-E card.
Looks like a complex board. Who was the board design outsourced to and who did the high speed trace consulting?
Isn't this the same as previous? https://news.ycombinator.com/item?id=45018509
No, the previous post focused more on the Lumafield scans of the Raspberry Pi boards. This one is about reverse engineering the schematics of the CM5.
Different project, one which was referenced in the post for that thread
Those are lumafield scans.
This is the old analog CT method: sanding
ten (10) layer board! :o Following twitch stream where TubeTime was doing it live gave some interesting insights, afair it did look like layout was done by more than one person and some things seemed to suggest time was more valuable than optimizing for less layers.
[dead]