Tell HN: System76 has critical firmware issues unresolved for over 3 years
Posted by jaime-ez 16 hours ago
I am system76 lemur pro (lemp13) owner since dec24.
At the time when looking for a new linux laptop I came to hacker news looking for information. The framework laptop was the most recommended one, but they don't ship to my country, and second to it was system76, mainly because of their hacker-centric vision, which I wanted to support.
Bought the best my money +shipping+taxes could afford. 56 GB ram.
On August 2025 started having problems with the fan, looking for help found this issue: https://github.com/system76/ec/issues/369. Several overheating episodes occurred.
On January 26 started having crashes, whenever plugin-unpliggin thunderbolt, hdmi, even usb sticks. On May the computer became unusable so I contacted support.
Support was good, the problem was the ram died. Never happened to me before so I didn't recognize the symptoms. And now I understand the most likely cause was due to overheating.
Bought a new ram, updated firmware, operating system to their cosmic OS (which I like), but the fan issues persisted.
Went back to the same issue looking for an update, nothing. Decided to look for a fan manager in the cosmic store, found one, installed, and realized the setup instructions resolved the fan issue.
I'm no expert and I am sure it is not a definitive solution but it works! So I went back to the issue and posted this comment if it was any help to others in distress: https://github.com/system76/ec/issues/369#issuecomment-5146109013
Yes, I did a snarky remark, but it comes from the frustration of not seeing this solved in over 3 years! When they know the consequences of overheating on system components. And turning the computer off every time I move is no solution.
Issues are sometimes where users find solutions to their problems, that's why I posted that workaround. However, in less than 30 minutes (more or less), my comment was flagged as off-topic and thus hidden from the thread.
This is what led me to post this. They are actively hiding useful information (from actual experience) to new users that will have the problem. Instead of focusing on solving it. Unacceptable.
I leave this as a warning to anyone who is looking to make a purchase decision.
Comments
Comment by naishoya 13 hours ago
Have you identified the code changes which effect the solution?
Are those able to be integrated back with the firmware via a PR?
IFF those are possible, it would be most interesting to see what the maintainer(s) [Tim Ziegler (@tizcza), Jeremy Soller (@jackpot51)] responses are to a PR for fixing this.
Without trying to solve it via those channels, one probably cannot say whether this specific "snarky" post being hidden constitutes bad faith on the part of the company.
Comment by jaime-ez 3 hours ago
In the issue comment I posted I identified the command that I assume fixes the problem. I am not a firmware expert so I won't dig deeper. What bugs me is that my post provides the only proven workaround (for my model at least) to prevent your laptop from overheating. And they hid it. There are tens of other ways of reacting to that comment, but hiding the only sane workaround for a system76 customer is just bad.
Comment by naishoya 2 hours ago
It seems like implementing a fancontrol tool, and perhaps making a one line edit to /etc/udev/rules.d/60-fan-control.rules so that the tool presets the content as a rule for udevadm to make it controllable for users without sudo.
That seems like it might work for your system, maybe for many laptops with similar firmware.
This 'closure' instead of engaging with the potential solution might not feel fun when you're proposing a band-aid.
However, instead of directing your conversation to the firmware developer, the comment used system76's github repository to directly speak to system76 customers, and try to shame the one developer into 'being motivated'
So, I can see the reasoning behind the closure as it has gone off the topic of reporting a bug, does not add details useful for improving development toward a fix of the bug, and proposes an assumed solution which would potentially add more noise to later responses from other users who attempted that solution whom it didn't help which has the potential to add more noise to the diagnostics process for this particular bug.
As you say, not a firmaware expert, and i understand the frustration at the lack of explanation from system76 in 'Hidden as off-topic' for your comment, but it's not deleted, anyone who want to read it can click the 'expand' button and read it.
Sorry that fan at "0"rpm and the heat spike from that seems to have cooked off your RAM, that really hurts. I would expect that system76 should honor the full extent of the warranty that came with your purchase, if that includes the RAM it would be simply inconvenient, not inconvenient and costly.
That has happened to me, not from this kind of firmware driven failure, but lots of time pushing systems to the limit of design has given me maybe too much experience with 'little puffs of blue smoke.'
If it's of any relevance, my experience going back 20+ years at this point makes me almost exclusively purchase Kingston RAM. Simply because I have had them honor a no-questions lifetime warranty for sticks of RAM. Even recovered from eWaste, physically damaged, off the floor. If it has the KVM number legible, I've had replacement sticks shipped to me for the cost of postage. Once even had them offer an option to upgrade from DDR to DDR2 as they didn't actually have much DDR "1" in stock, and it was better for me also at the time.
It makes me purchase new systems with a minimum of RAM installed, and then swap whatever brand it comes with for matched sets of Kingston after a basic functionality test on arrival.
I keep the original as fail-back for warranty / service.
Hope your new RAM is performant, and hope the udev solves the issue until the firmware gets fixed.