might not even browse these community pages, nor be on discord and be unaware of any bios updates. the average computer user is annoyed when a system reboots multiple times and all scary white characters appear and multiple progressbars.
EDIT !!! SHA1 I took was for 3.06, now its for 3.08.
Framework_Laptop_12th_Gen_Intel_Core_3.08_EFI.zip SHA1:317771e616be990ff3c54b24f9864d41dbea8d23
is still on the FW servers.
Just wanted to report a success with the 3.08 Beta EFI upgrade on my 13 inch Intel.
I’m well aware that it has been removed for the moment, however, I have had a 61kWh battery waiting to be installed for months now so I really did want to do the update to enable full support for that, and took the risk.
Nothing unusual - disabled secure UEFI boot in existing BIOS; booted from USB-A stick formatted FAT with the update unzipped to it; watched and waited while it went through all the steps; booted back into Fedora 39; restarted to re-enable secure UEFI boot.
I’m likely going to be doing the same thing. I have a 61Wh battery coming and only have Fedora 39 installed. Don’t want to reformat to install windows.
Worst comes to worst, I remove the SSD as others have said they have been doing
You might want to check your download here. I just downloaded the file (for clarity, it’s still on the servers but Framework say you shouldn’t use it) again which shows up with a SHA1 of 317771e616be990ff3c54b24f9864d41dbea8d23, this also matches the file I downloaded Jan 11 2024 (my post in this thread) which I still have.
The link to 3.08b above resolved “The Installer can only be un on Framework Laptop 12 Gen Intel Mainboards”. I had a Batch 2 1280p board on 3.06 Beta. Now 3.08 Beta.
However, now with Linux, entering “deep” suspend (in the /sys/power/mem_sleep sense) consistently soft bricks the device. Once in this state, a mainboard reset is required.
If you use Linux, I would steer clear of 3.0.8 until this is resolved.
Strange. I updated a few weeks ago, but I don’t see that issue. Using deep sleep (not s2idle), Debian testing, kernel 6.6.13. No problems with sleep here.
Interesting! It repros consistently for me with the Fedora-Workstation-Live-x86_64-39-1.5.iso (without installing; running off a USB stick). I was triggering deep suspend with something like:
# Do not run the following commands.
echo deep > /sys/power/mem_sleep
echo mem /sys/power/state
I’ve got a case open and have been discussing with the support team.
I have the opposite problem. On Debian with BIOS 3.05 and 3.08, the laptop would sometimes never come out of s2idle (I’d have to power the machine off with a 10-second power button press). The “deep” suspend has consistently worked.
After around one and a half months with BIOS 3.08, I haven’t encountered the following problems (seen on BIOS 3.06) anymore:
“disappearing” of the USB-A expansion card after plugging a device in more than two or three times
non-working function keys (brightness control, media keys, etc) in Linux (seen occasionally after uptimes > 3 weeks, used to require a complete shutdown to get the function keys working properly again)
no tap-to-click in Haiku with PS/2 emulation enabled after rebooting from another operating system
No change:
in Haiku, touchpad still works better after a cold boot (and only with PS/2 emulation)
New (but only once so far):
sudden and complete power-off (no shutdown) during full-screen remote-ing into another computer on Linux while connected to power and battery at 80 % (limit set in BIOS) - black screen, LED of power button off, no power indicator light between expansion cards - power indicator light came back after plugging the USB-C power cable out and in again, pressing the power-button resulted in some colorful blinking of the power indicator light (not the full failure sequence) before the laptop just started again
External Devices/Other: Charger (45W) plugged in (left rear)
EXPANSION CARD TYPES: 2x USB-C (left rear, right front), 2x USB-A (left front, right rear)
BIOS VERSION: Coming from 3.04
DRIVER PACKAGE VERSION: Initiallly released version (are there any newer??) with all updates coming in via Windows-Update
OS VERSION: Update performed on Win11 23H2 (3155) using 3.08b-MSI.
System configured as dual boot with Debian 12, Win is encrypted using Bitlocker (with PIN only, no TPM), Debian with LUKS.
Within hours of me reporting this, FW actually built a 3.08b revision of the installer that fixed this broken check. It is in this thread a few posts down from my initial report.
FW simply failed to mention this fixed installer anywhere in their own opener or update the link to the fixed installer…