Is par2 enough for bitrot protection?

Is par2 enough for bitrot protection?

Attached: B9w0-9OCIAELslB.png (450x300, 398K)

>bitrot paranoia
It's 2018, user.

>inb4 autists justifying their waste-of-money raid setups

There's no need for RAID.
Just a modern file system like btrfs or ZFS,
plus backups.

NTFS and other deprecated file systems should not be used for anything important. It's OK to store you child's games on NTFS, but nothing else.

EXPLAIN PIC????????
BIT ROT ???????????????

Why are you shouting?

magnetic hard drive data is prone to losing their magnetic orientation which can in turn cause bit rot
bit rot is just the degradation of data over time
pic looks like a dozen flipped bits although i smell bs
t. databending

>have a fucking old toshiba internal 120gb stored in closet for fucking 9 years
>hdd have some family stuff
>turn it on
>spinning
>check files
>no files damages or even lost
>?????????
Who the fuck fell for this MEME?

My screen appears like pic related for just a split second any time I close a container in i3.

maybe for detection

>Is par2 enough for bitrot protection?

I'm not that familiar with par2, but I know there is some problems with unicode support.

Why not use WinRar with a recovery record instead? It's like using par files, but they are magically embedded inside the rar-file it self. You can set the size of the recovery record. The larger the size, the more damage you can recover from. 10% let's you recover 10% damage, etc.

Attached: winrar.png (377x131, 6K)

If the files are not compressed, i.e. .txt, .wav, .bmp etc, then single bit errors will not be noticeable unless they happen in the header of the file (unlikely). They will manifest as a number or letter changing to something else in a text-file, a pixel changing color in an image, etc. If the files are compressed, you'll see more damage as even a single bit flip will mess up all subsequent decoding of the data. The chance of it happening is small, most consumer drives are rated at 10^14 URE (Unrecoverable Read Error Rate). That means that there is a chance that you get a read error every 12.5 TB of data. But this doesn't tell you the whole picture. Non-ECC RAM are affected by cosmic rays creating energetic neutrons and protons that pass through the DRAM chips. One estimation by IBM was that such events cause one error per month per 256 MiB of non-ECC RAM. Thermal neutrons and inductive or capacitive crosstalk also increase the chance of soft errors like these. All these factors add up, for instance every time you copy a file it pass through RAM and the northbridge or equivalent on your motherboard, where there is a non-zero risk of error happening. Same with reading from and storing the data on disk/other mediums. The risk of errors increase with time in terms of warm or cold storage (demagnetization, cosmic rays), or with the number of times the data is read and written (pass through ram, cables and circuitry and becoming exposed to cosmic rays and crosstalk).

Attached: Mars1_small.jpg (256x256, 14K)

Moreover, what's described here is the best case scenario, i.e. what everyone will experience as a bare minimum. On top of this you have the risk of faulty hardware or non-optimal storage conditions or operating environments which may further increase the risk of data errors. Examples include bringing your laptop or storage medium on a plane, cold storage in hot and/or humid environments, hot storage in close proximity to hardware producing excess heat, excess thermal fluctuations, close sources of Ionizing radiation, flash storage such as ssds, sd-cards and USB-sticks kept in cold storage for >6 months, optical media and/or tape stored for >3 years, especially if kept in hot and humid environments, vibrations, physical damage, dust, hard resets/powerouts during operation or any combination of the above.

u smart u kewl

Also: Solar flares, software corruption (fucked up partition tables, issues with sectors and all kinds of esoteric bugs caused by not dismounting the storage medium before ejecting it etc), bugs in file system handling, bugs in drivers, etc.

u dumb, u go back to plebbit pls brainlet

Corruption.

Attached: corruptcirno-3.png (682x1014, 1.01M)

thats called nouveau

>ZFS without ECC
lmao

Le stateur of Jow Forums.

>ZFS without ECC
>lmao
My desktop has ECC. Why wouldn't it?

rotational velocidensity

actually, the man knows what he's talking about, albeit, petty or nonsense to most people. when i've gone back to some really old mp3's from way back 'in-the-day' (90's, Napster, 56k modem) most of my mp3 library sounds like crap, mostly due to the technologies available at the time. i notice an unusually large amount of "pops" in a lot of songs and a 128k rip sounds more flat than a new 128k rip (both sound terrible regardless). anybody who knows how data is written to a disc would know that bits do get lost over time.