In case you want to know why this filesystem requires a such amount of RAM:
Because it's JAVA.
A fully working read-only FUSE, backed by my own MerkleDAG structure, namely the https://github.com/hurui200320/Ariteg
By caching everything in RAM and using 2.8GB RAM at the start-up, then growing to 5.5GB when playing 12 videos at the same time (bottlenecked by the GPU's video decoder), the filesystem itself is working pretty well. The random reading is fairly good, too: sliding through the progress bar in MPV results in an almost instant response (data stored on an NVME SSD).
I would call it a day.
Also, 295GB of video files only occupied 290GB after block-level deduplication. I would argue that compressing the same data would require an infinite amount of time (my program can handle more than 50MB/s when reading and writing happen on the same disk, while 7z can only get 20MB/s) and wouldn't get a better compression ratio than this.
hmmmmm, Chinese piracy vscode.
https://github.com/microsoft/vscode/issues/163798
Video: A Toyota pickup truck slowly pulled the shuttle across the bridge.
Me: That must be a modified pickup truck.
Wikipedia: The journey was famous for an *unmodified* Toyota Tundra pickup truck pulling the Space Shuttle across the Manchester Boulevard Bridge.
Me: WTF? 78 tons of spaceship, unmodified truck?
----
Endeavour's Wild Journey Through the Streets of Los Angeles (by Practical Engineering)
https://www.youtube.com/watch?v=tVzgHvTuwdU
The raspberry pi rebuild version is amazing.
The original version would be fantastic in terms of engineering.
----
This 1970s tank simulator drives through a tiny world (by Tom Scott)
https://www.youtube.com/watch?v=AcQifPHcMLE
I'm not sure if it's Mayday or not (maybe Mythbusters?), but from some TV program, I learned that modern airplanes are designed to be able to land safely by a passenger with autopilot and instructions from the tower.
The engineering is amazing.
----
I'm not a pilot. Can I land a 737? (by Tom Scott)
https://www.youtube.com/watch?v=AbTDzPUDxqY
Just watched the first video, amazed by the technique used in the vid.
Also reminded me of the second video, which is also pretty amazing.
----
I Collected Every Illegal Item In Minecraft Hardcore
https://www.youtube.com/watch?v=rFO1iqDpMZU
----
How far can you get from one block of bedrock in vanilla Minecraft?
https://www.youtube.com/watch?v=-OG4iOMGBMo
Copyright is not always a good thing. Of course, I don't want people to abuse my content, and I want to control them. But when your contents have some value to preserve, copyright is just a bunch of shit.
Especially when the owner doesn't care. I mean, Nintendo, closing the 3DS eshop and don't care about those old games.
----
Time to UNSUBSCRIBE! Own your content, cancel Netflix!
https://www.youtube.com/watch?v=RZ8ijmy3qPo
WeChat started banning people just because they sent this picture in group chat.
More info:
Rare protest against China’s Xi Jinping days before Communist Party congress
https://edition.cnn.com/2022/10/13/china/china-party-congress-protest-banners-xi-intl-hnk/index.html
So... Can we hack some blockchain using this feature?
For crypto operations, like the simplest one, comparing an input string against the known one. If we use xor to do the comparison, how on earth the time consumption will be different? (Only using xor is possibly not considered as crypto operation, but the idea is how can Intel use different amount of time for calculating the same amount of data?)
Or it's just Intel's excuse for don't blame me if your code don't work securely on newer CPUs?
Possibly a post contains political things and math from several youtube videos.
Those videos explained how altruism affects itself during evolution.
And I'm wondering: if the dissidents in China have similar movements, how would the number change over time?
(Here, I will just assume dissidents are good for the country. Also, assuming only dissidents will sacrifice themselves and alert others, possibly make other normal people into dissidents. ***And this idea is against the mathematical model discussed in the video, just a random thought. No serious political opinion was disclosed, aka I didn't say anything about if we need or we don't need dissidents in China. This is a proper disclaimer, and I do not accept any accusation for this random thought.***)
From the first video, if we don't know who is the dissidents, the result is not good. But considering those dissidents generally won't sacrifice themselves regardless, the reality won't be so bad. If dissidents can be identified by and only by other dissidents (unlike the green beard), things looked promising. However, when introducing the imposter, it goes terribly bad again.
From the second video, if those dissidents grouped together, it would be much better against different situations. But during that simulation, it's far from real-world situations.
Beyond the videos, I'm thinking about myself. I really appreciate those HKers, who can overcome their fear and fight for their own future. But in 2022, in China, I don't think there will be similar events happening in the future. And under this situation, when will be a good time to just move to another country, instead of trying to save this place?
----
Simulating Green Beard Altruism
https://www.youtube.com/watch?v=goePYJ74Ydg
----
Simulating the Evolution of Sacrificing for Family
https://www.youtube.com/watch?v=iLX_r_WPrIw
My profile image is actually generated by stable diffusion.
----
How AI Image Generators Work (Stable Diffusion / Dall-E) - Computerphile
https://www.youtube.com/watch?v=1CIpzeNxIhU
Abandoned since this instance is broken. Please follow my new account: @skyblond