This post could have been a prompt. It is 1000% generated with an LLM.
Tell-tale signs: fluffs up the "key features for x". Each feature follows bold title, highly generic explanation of the title. Then it lists some examples, all following the same pattern "why x".
Then the conclusion is full of gpt-isms. Doesn't say anything of value, just re-iterates fluff words.
> NetBSD's presence in space demonstrates how a well-designed, community-supported operating system can play a pivotal role in advancing space exploration and technology. As satellite missions grow more ambitious, NetBSD’s adaptability ensures it will remain a key player in the evolving field of aerospace engineering.
I know there's a policy encouraging users not to cry "llm generated" every time, but this is quite egregious.
I really want an excuse to run NetBSD on something. I wonder if ESP32's are capable of running NetBSD? Even just running on a RPi would be fun.
Also, for embedded devices the coherent whole system design of *BSDs with the package builder seems like it'd be much nicer than BuildRoot or Yocto.
I've come to very much dislike trying to keep Linux images for embedded stuff running over years. Perhaps it's because Linux claims not pride itself on "not breaking userspace" but seems to add a new GPIO sub-system every 5 years. Ain't nobody got time for that.
This Substack article should not be trusted - there are no sources given. I google'd each of the 4 example satellites and "netbsd" and the only results were the Substack article and people referencing the article. NetBSD may have been used in ground systems, but (i) that would have been a non-story and (ii) I couldn't find any evidence of that, so I wonder where the author picked up this information.
I am most familiar with SAMPEX, which was launched in 1992. The initial release of NetBSD, version 0.8, was in 1993 according to Wikipedia. Okay, the article says the project "transitioned" to NetBSD in the "extended mission". Okay, maybe in the 2000s, let's say they decided to replace the original OS/real-time-executive on a working spacecraft with a new OS. So you abruptly replace the old OS/RTE-based flight software with software based on a new OS/RTE. (You don't gradually transition from one OS/RTE to another.) I don't buy it. On a working spacecraft? No.
(I realize the article was probably AI-generated.)
5 comments
[ 3.4 ms ] story [ 32.7 ms ] threadTell-tale signs: fluffs up the "key features for x". Each feature follows bold title, highly generic explanation of the title. Then it lists some examples, all following the same pattern "why x".
Then the conclusion is full of gpt-isms. Doesn't say anything of value, just re-iterates fluff words.
> NetBSD's presence in space demonstrates how a well-designed, community-supported operating system can play a pivotal role in advancing space exploration and technology. As satellite missions grow more ambitious, NetBSD’s adaptability ensures it will remain a key player in the evolving field of aerospace engineering.
I know there's a policy encouraging users not to cry "llm generated" every time, but this is quite egregious.
Also, for embedded devices the coherent whole system design of *BSDs with the package builder seems like it'd be much nicer than BuildRoot or Yocto.
I've come to very much dislike trying to keep Linux images for embedded stuff running over years. Perhaps it's because Linux claims not pride itself on "not breaking userspace" but seems to add a new GPIO sub-system every 5 years. Ain't nobody got time for that.
One case the mentioned date launched. I would like to know:
* dates launched
* Satellites still in active use
* What NetBSD release was the software built from
* Are the Satellites still being produced in 2025 and if so are they still based on NetBSD
So in a way, the article had nothing "new".
I am most familiar with SAMPEX, which was launched in 1992. The initial release of NetBSD, version 0.8, was in 1993 according to Wikipedia. Okay, the article says the project "transitioned" to NetBSD in the "extended mission". Okay, maybe in the 2000s, let's say they decided to replace the original OS/real-time-executive on a working spacecraft with a new OS. So you abruptly replace the old OS/RTE-based flight software with software based on a new OS/RTE. (You don't gradually transition from one OS/RTE to another.) I don't buy it. On a working spacecraft? No.
(I realize the article was probably AI-generated.)