Purism's Librem 5 To Rely On Secondary Processor For Binary Blobs

[edenist]

This isn't a binary that runs on the host processor. It's binary firmware that gets loaded into sram on the DDR PHY and runs on that. It's basically just changing who loads the firmware into the DDR PHY.

Get out of here with your logic based on an actual understanding of the article. We've got pitchforks which need things to wave at dangit!

[edwaleni]

I want to know how (if) they are going to set up the radio? Between Qcomm, Ericsson, Apple and Intel there is a lot of territory to avoid.

If anything, it may not support *any* advanced modulation schemes for data. So it will be great for GSM voice but any scheme beyond GPRS may require those blobs once again.

I am more fascinated to see how far they will actually get with this than actually owning one.

[starshipeleven]

Why do people get so worked over these things? Im willing to bet that most of the whinning here comes from people that couldn't write this low level code if they tried.

I'm sure they could not write even shell script if they tried.

[starshipeleven]

This means we have a Trojan Horse in the DDR4 standard:

No it means that the ram controller is too complex to have its logic baked into the silicon, just like every other RAM controller.

[starshipeleven]

I want to know how (if) they are going to set up the radio? Between Qcomm, Ericsson, Apple and Intel there is a lot of territory to avoid.

If you mean modem, they will use a commercial modem chip over USB connection (no DMA possible) with a hardware killswitch.

They have no other choice, really.

[Delgarde]

This sounds more like wanting to keep a cheap ARM chip than to actually investigate options. I know POWER9 won't work in a phone, but surely the appropriate technology could be licensed from e.g. IBM and a proper chip fabbed at the quantities expected for a phone?

The "quantities expected for a phone" are, in the case of this phone, near enough to zero as makes no difference. Purism isn't Apple or Samsung - they're not shipping millions of phones, and they certainly don't have the resources to be paying licensing and chip development costs if they can avoid it.

[coder111]

Just one nitpick:

That's the ONLY Arm chip on the market that's supported (somewhat) by open-source software. It's not like Purism has much choice...

I think this move is simply about time-to-market. Maybe with time they could implement open-source DDR PHY training for the phone, but if that delays the launch by 6 months, maybe this is not the right time to do this. They can do it for V2 if there is a V2.

This sounds more like wanting to keep a cheap ARM chip than to actually investigate options.

[starshipeleven]

Just one nitpick:

That's the ONLY Arm chip on the market that's supported (somewhat) by open-source software. It's not like Purism has much choice...

Nitpick of the nitpick:

Afaik only POWER processors have open firmware for the RAM controller at all, so it's not like they could have chosen a "more open" SoC anyway.

[euler271]

I dont know what's the big issue. 1) I think this is being designed taking into account privacy AND performance, and making the best available decision to have some of both. As some ppl said before,it's basically low-level firmware that goes and stays unmodified, so it behaves basically as hardware. but in this case, it's even BETTER than hardware, because 2) being all in a separate flash device, eventually anyone can load an opensource code to replace the blob and have the entire cake and eat it too. So it's just about having a good understanding of what's actually happening and the technical implications of the decisions, instead of just pulling a Stallman without the actual knowledge of how things work.

[starshipeleven]

I dont know what's the big issue.

People don't understand and rage when they see pre-programmed keywords. Full news at 20:00.