[spicybright]

While I don't think he should be flipping the bird, his technical arguments are absolutely spot on.

An aside question for readers: How often does RAM error, and is it a significant problem in practice?

I don't have any direct experience with this (that I could tell) using consumer computers for the past 15-odd years of programming.

[ahelwer]

That photo is from him saying "fuck you" to NVIDIA back in 2012: https://arstechnica.com/information-technology/2012/06/linus...

FTA: "Bit flips can happen for many reasons, beginning with cosmic-ray impact or simple hardware failure. A large-scale study[0] of Google servers found that roughly 32 percent of all servers (and 8 percent of all DIMMs) in Google's fleet experience at least one memory error per year. But the vast majority of these are single-bit errors—and since Google is using server CPUs and ECC RAM, this means the machines in question keep right on trucking."

[0] http://www.cs.toronto.edu/~bianca/papers/sigmetrics09.pdf

[monocasa]

Something that doesn't get said enough is that Google had a habit at the time of buying RAM chips that had failed manufacturer QA, stuck them on DIMMs themselves, and revalidated those DIMMs. They were really leaning into the whole "embrace failures if they're going to happen anyway and you can get cheaper servers out of it". So those numbers need to be taken with a grain of salt.

[mywittyname]

Even still, a one-third of a memory error per year is a rate that I'm totally comfortable with.

[thinkmassive]

Also FTA: "ECC RAM ... can generally stop Rowhammer attacks—in which rapidly flipping bits in one area of RAM cause bits in an adjacent area to change."

[kodah]

> While I don't think he should be flipping the bird, his technical arguments are absolutely spot on.

Linus is certainly known for some inappropriate rants against people which he has since toned down, apologized for, and worked to correct.

That said, this image is taken from an event where Linus was prompted about Nvidia in 2012 during some filming. For those who don't use Linux, these times were extremely chaotic in graphics, especially for Linux with respect to Nvidia, who were largely catering to Windows users. It's what kept Windows gaming in such a strategic position. This affects people who have lower end machines, people who can't afford Windows, and people who seek freedom from predatory data collection. Throwing up his middle finger at a company with those kind of practices is hardly revolutionary or inappropriate.

Source: https://www.wired.com/2012/06/torvalds-nvidia-linux/amp

[spicybright]

Ah, thank for you for the explanation on the pic. I definitely remember that being a huge issue in 2012.

[vbrandl]

More often than you'd think. There's a paper from black hat 2011 [0] where bit flips and domains with a small hamming distance were used. Take google.com and woogle.com which in their ASCII representation only differ by one bit. The paper showed that these domains will receive traffic for their intended counterparts because of bit flips in the RAM in DNS servers.

[0]: https://media.blackhat.com/bh-us-11/Dinaburg/BH_US_11_Dinabu...

[taneq]

I think a measured dose of profanity has its place as a way to emphasise a point beyond what stern wording can convey.

That said, I also would like to know more about how often RAM errors actually happen. My gut feeling is that on mass produced consumer hardware, software errors FAR outweigh hardware errors in terms of how much they inconvenience me personally. Maybe the very occasional "that's weird" moments where a restart fixed it were hardware errors but they're few and far between. That's just a feeling, though, and I like numbers.