Other Arch Flavors I've tried (some are no longer with us) include:
- ArchBang
- EndeavourOS
- Manjaro
- Chakra
So with that out of the way, I've found my Garuda experience incredibly painful. From messy repositories (Chaotic-AUR plus their own stuff), to an overly involved upgrade process (when using the helper) - the distro screams of a team that has no freakin' clue how to maintain an actual distribution.
It's basically Arch on hard mode with so many settings rolled into their own packages which need to be removed before customization.
Then we get to the purported performance enhancements and, honestly, this is the worst performing distro I've ever used, by multiple miles. I'm not sure if its the scheduler settings, or something with the zram settings - but this distro hitches and hangs constantly. (5950x, 64GB of Ram, Samsung 980 Pro drives, NVIDIA RTX 3080Ti - NOT a weak machine by any standards)
I'd normally chalk it up to compositor issues on Wayland (yes, I prefer Wayland and it works fine for most Arch derivitaves even with Nvidia). However the performance issues even crop up on basic terminal commands on a TTY with lots of weird hangs and lags.
The ONLY thing that was easier on this distro was installing the various Proton GE builds and other specialty stuff found in the Chaotic-AUR. But given the above, it's definitely not worth it when one can configure an Arch box to do the same things without all of the problems.
Perhaps I'm not doing something right? Given all the praise for this distro, perhaps it shouldn't perform like this?
To be completely and utterly clear - I'm an advanced user trying out these distros for fun and discovery. I can indeed "just use a different distro" but wanted to give this one a fair shake before moving on.
I understand the sentiment... But... This is a terribly reasoned and researched article. We only need to look at the NASA to see how this is flawed.
Blown Capacitors/Resistors, Solder failing over time and through various conditions, failing RAM/ROM/NAND chips. Just because the technology has less "moving parts" doesn't mean its any less susceptible to environmental and age based degradation. And we only get around those challenges by necessity and really smart engineers.
The article uses an example of a 2014 Model S - but I don't think it's fair to conflate 2 Million Kilometers in the span of 10 years, vs the same distance in the span of the quoted 74 years. It's just not the same. Time brings seasonal changes which happen regardless if you drive the vehicle or not. Further, in many cases, the car computers never completely turn off, meaning that these computers are running 24/7/365. Not to mention how Tesla's in general have poor reliability as tracked by multiple third parties.
Perhaps if there was an easy-access panel that allowed replacement of 90% of the car's electronics through standardized cards, that would go a long way to realizing a "Buy it for Life" vehicle. Assuming that we can just build 80 year, "all-condition" capacitors, resistors, and other components isn't realistic or scalable.
Whats weird is that they seem to concede the repairability aspect at the end, without any thought whatsoever as to how that impacts reliability.
In Conclusion: A poor article, with a surface level view of reliability, using bad examples (One person's Tesla) to prop up a narrative that EVs - as they exist - could last forever if companies wanted.