Purism Librem 5 Linux Smartphone Campaign Set To End At Around $2 Million

[phoronix]

Phoronix: Purism Librem 5 Linux Smartphone Campaign Set To End At Around $2 Million

Tomorrow marks the end of the crowdfunding campaign for Purism's Librem 5 smartphone campaign...

http://www.phoronix.com/scan.php?page=news_item&px=Librem-5-Ends-Tomorrow

[Guest]

If purism won't go with at least quad-core A53, then they write off themselves. It's great goal to bring "pure" Linux to smartphones, but investment into old ARM cores is waste. They would better be going with at least 4xA53 + 2xA72. It will be more future-proof for them.

Personally, I wouldn't buy any smartphone not having at least 4xA53 and 3GB of RAM. I want my smartphone to last few years, and being capable to run latest resource-hungry apps.

[Dawn]

If purism won't go with at least quad-core A53, then they write off themselves. It's great goal to bring "pure" Linux to smartphones, but investment into old ARM cores is waste. They would better be going with at least 4xA53 + 2xA72. It will be more future-proof for them.

Personally, I wouldn't buy any smartphone not having at least 4xA53 and 3GB of RAM. I want my smartphone to last few years, and being capable to run latest resource-hungry apps.

The plan is to use an i.MX8M (which is 4xA53) and 3GB RAM. The limiting factor in going beyond that is finding a chip that's sufficiently open, which is harder than it once was.

[Guest]

The plan is to use an i.MX8M (which is 4xA53) and 3GB RAM. The limiting factor in going beyond that is finding a chip that's sufficiently open, which is harder than it once was.

Why is it hard to find ARM chip with open specification? All the value lies in the HW logic of the chip, not in the "driving" the chip. Even, if chip is completely open, there's no guarantee, that there isn't any micro software hidden within it, which could be used as a vendor backdoor.

And, what do you mean by "sufficiently open"? it's pretty much ambiguous term.

[Guest]

Also, aren't chips used in ODROID boards made by hardkernel sufficiently open? They already work with Ubuntu 16.04.

[Dawn]

Also, aren't chips used in ODROID boards made by hardkernel sufficiently open? They already work with Ubuntu 16.04.

I can't comment in detail on other ones, but at least the XU4 uses a GPU with no open driver support.

[starshipeleven]

Personally, I wouldn't buy any smartphone not having at least 4xA53 and 3GB of RAM. I want my smartphone to last few years, and being capable to run latest resource-hungry apps.

This smartphone won't run Android anyway, so you don't need so much resources.

[starshipeleven]

Why is it hard to find ARM chip with open specification?

Because the hardware manufacturers like to keep their stuff secret. Publishing decent hardware documentation is completely horrific for most.

Even, if chip is completely open, there's no guarantee, that there isn't any micro software hidden within it, which could be used as a vendor backdoor.

Embedding such things has a significant cost and pretty low return of investment, vendors aren't evil villains, they only do stuff for profit.

And, what do you mean by "sufficiently open"? it's pretty much ambiguous term.

Has open drivers for the most important SoC features, like GPU and media acceleration subsystem.

[starshipeleven]

Also, aren't chips used in ODROID boards made by hardkernel sufficiently open? They already work with Ubuntu 16.04.

"already"?
They work by running a distro with an older and blobbed kernel to support the GPU (Mali) and the media acceleration subsystem (if they have one), or at least that's the most you can do with an Amlogic SoC. You can't update the kernel or the SoC-specific binary drivers stop working.

Same as with last supported Android version, the 5.1.

This is the same situation for most SBCs not based on FXP SoCs, or raspbery Pis.

[starshipeleven]

AAnd.... as predicted in the older thread, they are getting 2 millions for this project. Not bad at all, I really hope they can pull this off.