GNU Libreboot Release Adds New Chromebook & ASUS/Gigabyte/Intel Board Support

[phoronix]

Phoronix: GNU Libreboot Release Adds New Chromebook & ASUS/Gigabyte/Intel Board Support

The Libreboot project has done their first official release of this Coreboot binary-free downstream now being under the GNU project label...

http://www.phoronix.com/scan.php?page=news_item&px=GNU-Libreboot-2016-August

[starshipeleven]

Hm, at least that's some server boards. They aren't total crap, but they are not that easy to find at decent prices.
But still, I sometimes wonder why bother. Support for anything new is impossible.
It would be more useful to actually make something new and free.

Or to at least make a coreboot firmware for AM1 motherboards that enables ECC ram, among other things. Non-free but at least useful and something I'd actually use.

[tigerroast]

About goddamn time, Libreboot! Now I don't have to build the binary myself for my Chromebook C201!

There's about 4 people that have documented how to get Debian working on the C201. After an update last month, it's possible to write a script that automates the installation process. After prep work, of course.

[torsionbar28]

Or to at least make a coreboot firmware for AM1 motherboards that enables ECC ram, among other things. Non-free but at least useful and something I'd actually use.

x2, at just 25w, those AM1 processors have a compelling TDP for a small server.... if only they had ECC support.

[Termy]

Maybe with AMDs ongoing OS-love we could be hoping for faster Libreboot-Options for coming ZEN-Boards/Chipsets?
I take it that also the Board-manufacturer would need to open up, but a big part of the work of libreboot-people should be related to the chipset?
i don't really expect this to happen, but one can still hope

[starshipeleven]

x2, at just 25w, those AM1 processors have a compelling TDP for a small server.... if only they had ECC support.

Also AES acceleration. (btrfs/ZFS, hint hint)

[starshipeleven]

ghetto edit: I mean they have already AES acceleration, so they only lack ECC memory to be great miniserver chips.

[Michael_S]

How often has the lack of ECC caused a problem for you? To my knowledge, I've never been bitten by that and I have four PCs with a combined 72GB of non-ECC RAM in my house right now.

[starshipeleven]

How often has the lack of ECC caused a problem for you? To my knowledge, I've never been bitten by that and I have four PCs with a combined 72GB of non-ECC RAM in my house right now.

Wrong question.
Right question is "do you really want to risk having to waste time to fix shit that should have NOT failed/reload data that had been corrupted/and so on or you prefer to pay some premium to be sure it's as low-mainteneance as possible"?

Because that's probability, and it's of course not common, but can happen and does happen. And if it happens you have to waste time and resources to get everything back in good shape. You can't know when it happens, it may even happen in a moment when you cannot devote thime and resources to fix it, so you get slammed in the face by it for a while.

For example, I know platoons of people that use the crappiest shit PSUs on earth in their PC just to save a few dozen bucks, but they never got hit by a full-scale foul-smelling PSU meltdown that kills hard drives and GPU.

But I do get people that got hit by that too.

Adding ECC on a home server isn't a major cost addition (as long as you aren't making it with BRAND NEW and CUTTING EDGE parts), on company stuff hardware cost is usually irrelevant.

[aht0]

ghetto edit: I mean they have already AES acceleration, so they only lack ECC memory to be great miniserver chips.

Some have reported success combining ECC-unbuffered RAM and Asus AM1M-A. Personally had no luck getting ECC work on that board. Yeah, I bought it. It's still alive and kicking too..