(untitled)

>vhs degradation >disc rot >data rot >drive failure When are we going to be able to store data forever, without having to worry about it? It makes me anxious to think none of this stuff is permanent.

>>2173 Even carving your data in huge granite blocks won't store it for ever - picktish standing stones in Scotland are now faded beyond reliable data retention, and thats just the stones that still stand! - (thats Scottish granite not some soykafty sand stone!) I guess somthing like Storj coin might have potential to last "forever" as a distributed storage could be more resistant than granite - as long as Storj coin is still a thing in 10yrs let alone 10,000! Failing that and using current "normal" media its gold film CDs - 100yrs or so life expectancy - maybe more if vacuum sealed in a dark cool dry room? I think the real solution is to keep your data moving - as tech progresses transfer to keep data current. Of course 3 copies (not all at same physical location) is the only way to give your data some degree of longevity. - guarantee at some point in 500yrs 2 copies will fuarrrk up!

IPFS, freenet zeronet and similar stuff... It's not indefinitely, but it's good enough.

There was for a while hype about storing data on 5D glass discs. Sadly it seems like this is already rather old news with nothing new in the tech progress. But if it became true to what it claims and popular enough then worrying about data loss would disappear. https://www.allaboutcircuits.com/news/5d-data-storage-how-does-it-work-and-when-can-we-use-it/

>>2180 >5D sounds like vaporware

Do magnetic disk drives fuarrrk up if they aren't used? Would a hd be fine after 50 years in a draw?

You think you want data to last forever, but you really don't.

>>2203 IDK but i get the feeling that older 500MB drives might be ok as the magnetic density was much less and so probably alot more stable. If kept at a constant cool temp, sealed in dry air tight bag... maybe a drive could last for a very long time? - I dont think ive got much older than 15yrs - and some of those (Maxtor) 15yr old drives are dead - some are not (Seagate)...

>>2180 I heard about this but how long will it take to be commercially implemented? Are we sure it can last several millennia?

>>2173 Nothing is permanent, eventually even black holes will disappear and it'll be nothingness again. >>2203 Yes, magnetic drives are prone to magnetic breakdown, anything magnetic is far from ideal for long term storage. I think the best solution is something physical, like CD's. You can always read them, even if the technology is long obsolete a laser reader isn't something too complicated to devise. Verbatim claims to have a "1000 year old lifespan disk" called the M-Disk, it's targeted towards archivists and similar.

>>2223 CDs decay eventually too, even if they're not in use.

Have less stuff on ur pc and use Super Disk!

I wonder how well ROMs will hold up. Imagine space-grade ROMs of all your data

>>2290 "space-grade" memory just means a soykafton of error-correcting codes and redundancy across backups t. someone who interviewed for a ISS engineer job

>>2173 Film will outlast all of this.

>>2173 do what everyone did before recording was a thing tell your kids and your mates let them copy it spread your software seeds with your meatspace seeds

>>2204 no, we really do before data recording, all we had were sticks and stones, and they got boring very quick

DNA, biocomputing. https://en.wikipedia.org/wiki/DNA_digital_data_storage https://phys.org/news/2017-03-short-movie-dna.html https://www.technologyreview.com/s/607880/microsoft-has-a-plan-to-add-dna-data-storage-to-its-cloud/?set=607915

>>2224 It should be noted that the longevity of optical media is incredibly variable as the format itself was standardized, yet the manufacturing process was not. Some burned CDs from the 90's are already decaying, and others may last for hundreds of years.

>>2209 I've got a Maxtor 160gb disc running my gentoo backup server, it's probably 13yo. Should I be worried?

>>2177 >I think the real solution is to keep your data moving I call this the torch passing method. Think of your files as a fire that eventually burns up whatever physical substrate you store it on. The metaphor breaks down when it comes to integrity checking and such, but oh well.

>>2568 True enough. The oldest CDs I had stopped working a long time ago, and I always took proper care of them. They were first gen CDs I got from a (dead) friend who also always took care of his media. CDs will wear out anyway as people abuse them but that's another matter. Having CD players in car stereos definitely encouraged a lot of that.

>>2547 DNA is easily corrupted by radiation

>>2765 >even your data isn't safe from cancer

>>2765 Yeah, and even when you have all the error-correcting proteins working 24/7 the strand breaks down over time. Still, if you got the right system and layered a digital ECC on top of the DNA repair processes, you could preserve a lot of data in bacteria for *centuries* reliably.

>>2342 Their memory will fade and there's no sufficient error correcting algorithm for people.

>>2173 > It makes me anxious to think none of this stuff is permanent. Try to get over fear of change, brother.

Hitachi has experimented with quartz blocks that would last hundreds of thousands of years with zero data degradation, with the data lazered into them. Only downside, really the only one, are huge spaces taken up for relatively little data. Other than that there's only media transfer or carving things into sealed off rocks.

paper is still a good medium for long term storage

>>2795 xxd -p file.txt > paper

Microfilm. No, seriously. It can last up to 500 years. That's probably well past whenever anyone will need your data. Of course, I have to assume it's quite expensive. Regular camera film coated on PET will also last incredibly long. It would be a viable alternative for your photography needs.

>>2793 This anxiety is something I've forgotten with age. I used to worry that all my cassettes would wear out, there'd be nowhere to store old VHS and vinyl and other formats that take up space (the BBC purged 100+ Doctor Whos for a reason) and nothing to play them on anyway. Besides media wearing out, there's also fire, flood, eviction, etc (most of humanity's ancient text were lost to fire, insects, persecution or war anyway). Now that I've actually lost most of what I owned twice, I don't care that much anymore. The only thing I miss in a painful way are some music demos I played on and some others friends made, friends that are unreachable for different reasons. All in all, less than 3 hours of audio.

>>2794 If you were able to machine tiny holes into rock you could make punch slabs.

Tape and paper are the most time efficient for archiving ( 40-50 years )

>>2950 ...assuming you have space to store it safely.

>>2173 all you need is a proper backup system and youll heavily decrease your chances of data loss

Proper rotating backups. You can get cheap ass ESXs etc., and some cloud storage, keep it rotating, take snapshots of your VMs and migrate them off site, and keep them backed up onsite on separate media. If you've got the know how to do it, then you can make the money to do it.

>>2173 >It makes me anxious to think none of this stuff is permanent. Well nothing is permanent. Every sinner, every saint, every dictator or king or popstar, every war, every love, every truth that's ever been spoken by man is right here. On a dust speck suspended in a sunbeam. Nothing is permanent. Every ebb has it's flow. All rivers run to the sea. Make peace with it.

Nothing is. but Tapedrives arnt bad

BD-R HTL CD-ROM stored at 45 degrees temp, assuming minimal use and multiple copies are made your data is 99% safe guaranteed for 25 years You're welcome.

Keep your discs on a multi rotator Disc degradation is caused by molecules shifting.

Floppy Disks hold well

The limits of data storage depend on the technology to write and read such data. For example, an 8″ × 10″ (roughly A4 without margins) 300dpi 8-bit greyscale image map contains 7.2 megabytes of data—assuming a scanner can accurately reproduce the printed image to that resolution and color depth, and a program can accurately interpret such an image. A similarly sized image in 2400dpi 24-bit true color theoretically contains 1.38 gigabytes of information. en.wikipedia.org/wiki/Paper_data_storage

>>2224 Speaking of CDs. A couple months ago, I loaded a cd that was burned 20 years ago and it was fine. Bunch of VJ clips, 320x240 res.

What are the best methods to storing plain text?

>>3778 Inscribing it in clay tablets

>>3786 Bronze is better, clay is good but under certain climate conditions only.

It'd be an interesting idea if someone rich started a foundation with a very large storage facility where they spent the time transcribing important films, books, art, and music into punch cards

>>3795 hypothetically, wouldn't gold or tungsten plates be very good too, because of the lack of elemental degradation? >>3809 DNA encryption is capable of 700tb/gram of storage (in a cold area) longevity allows for possible storage of over 50,000 years of possible stability, current cost is around $7000 to encrypt and $2000 to decrypt a 2mb sample also possible destruction due to sensitivity is an issue

>>2173 http://gizmodo.com/5925719/a-sapphire-hard-disk-will-last-1-million-years-but-you-cant-afford-it >million year hard drive for 30k, used to preserve locations of nuclear waste, on that note, what are some informaions that should be preserved besides dank memes?

>>3825 >what are some informaions that should be preserved In complete seriousness with no hint of joking or irony, all the Grateful Dead recordings ever made. This is the greatest treasure humanity currently has and it's of the utmost importance it be preserved indefinitely for posterity.

I recently installed a Blu-ray drive on my PC so I could archive things on M-Disc, pretty neat. Other than that I guess the next cheapest option would be tape, maybe D-VHS?

>>3799 VHS can last way longer than 15 years with careful use

Might be bullsoykaf but worth mentioning. >This alien space cathedral contains a vault with the works of Scientology's founder, the science fiction writer L Ron Hubbard, who claimed that humans are extraterrestrial beings. >Hubbard's writings are preserved for eternity on stainless steel tablets, on gold discs and encased in titanium capsules. http://www.ibtimes.co.uk/scientology-l-ron-hubbard-new-mexico-alien-499692

>>3799 What does excessive care mean in the case of CD/DVDs? Just not using them and keeping them sealed, or what?

>>3823 Gold is a noble metal, but too soft. I wouldn't recommend wolfram due to chemical reactions. But so does bronze. Hmmm..

>>3825 Chemistry, physics, math, and a guide to decode the message.

>>3869 Keeping them on plastic cases, never leaving them on tables or anywhere the bottom can slide and scratch, not using DVD folder sleeves, etc. Basically keep them in their original case, take them out carefully, never touch the bottom or let anything make contact with the bottom, and make sure your DVD player is positioned carefully and won't make any sudden movements that could scratch the discs. I also suggest you don't leave them face-down as dust can gather under the disc and you'll have to blow them off with something most likely. Cloth or napkins can harm the disc.

My recommendation is using a bunch of BDs for routine backups and M-DISC for important archives. A quality Bluray disk uses some special coating on the recording layer. Some manufacturers claim it improves the lifetime from 20 years (DVD) to 100 years, insufficient evidence, but surely less prone to deterioration than DVDs. I have some cheap blank DVDs stayed in the case for two years and I discovered some of them have stopped working completely, while another box of blank Sony DVDs I bought 7 years ago, when I was still a skiddie, are still working properly... M-DISC, as mentioned, is readable for standard BD readers, and recorded by a compatible BD writer. As long as the discs are properly stored, your personal data is probably okay for more than a decade. Archiving data for the Internet or the whole civilization is out of reach of my post, obviously. A final thought, I believe optical drive is still an awesome and useful technology. I don't understand the normal person's "discs? $current_year, who use discs" narrative. Perhaps some of the reasons are * the widespread use of portable devices, particularly those iThings and drive-less laptops. * The unpopularity of Bluray, at least in my region. Most people on the Internet use BitTorrent and never buy a disc of movie, BD is still an edgy and expensive technology in their impression, so they never think about it. It is understandable, I also don't need a mechanical device for my laptop which is unused for 95% percent of time. But at least you need one for your main workstation for data storage. A BD recorder is one of the best gadget I ever purchased. The price of BDs and recorders has reached a reasonable price for an average hobbyist or gamer, and for cold storage there is really no cheaper alternatives. Or is there? THE CLOUD! THE CLOUD! NOW I KNOW WHO'S BEHIND THE EXTINCTION OF OPTICAL DISCS... but there is still a bandwidth problem... oh ALL YOU DATA IS GENERATED WHILE CONNECTED, no upload needed...

Archiving data is for CIA tyrants

>>4013 By the logic you're applying here right now, I could make a simple analogy that knives are for murderers. fuarrrk off

>>4013 Hmm.. not archiving data.. seems like this retarded theoretic backfired on us recently

Regarding data preservation, I own a pretty big collection of old optical discs, many of them contain original retail versions of games, historical computer software, hardware drivers, and other things not to be found on the internet anymore, often from magazines. I'd really want to make it possible to preserve this for next generations. The catch is, the most of that data is copyrighted, or the legal status of copyright is unclear at this point in time. Aside of making local backups of it, i'd really love to upload this to some public repository. Are there any options to do so without risking any legal action? Maybe some services that deny public access to such data until the copyright expires? Thought about Internet Archive, but i have to dig and read about their policy more, or maybe contact them directly to clarify the situation.

>>2173 put it in a safe and bury it underground there it won't be affected by the wind, water, and living organisms.

You should look into ZFS and FreeNAS.

>>4137 There's far more legally dubious content on the Internet Archive than games, abandonware and drivers. Just start uploading before you forget about it. Also ask on winworldpc about sharing your abandonware.

>>2173 If we want to prolong the life of our history as the human species and information, we need to take many steps: >Prevent external and internal degradation using all possible measures (Radiation, EMP, etc.). >Spread data across multiple mediums of storage in the event one medium becomes corrupt. >Ensure the mediums of data storage are at their top performance and upgrade when new methods arise. If you wanted to store a picture of a cat, you would store a copy of the picture in twenty various hard drives that are made by the best company (Confirmed with testing by large database companies and schematics), and these hard drives would be stored in a controlled environment that is heavily protected to decrease radiation exposure and external factors such as seismic waves. Thin solid plates with a massive area could have small holes drilled into them to represent data, and plates could be stacked on top of each other to create a cube with various layers of data. Books are almost always more reliable than computers though, due to the fact that they are easily interpreted by humans without external equipment, and do not require the amount of resources and care that computers require to run, they are extremely convenient and long-lasting. If you want to store books, you should have a climate-controlled room with the books individually stored within plastic cases, possibly airtight if it was for long periods of time.

Best way to prepare for EMP strike?

>>4596 shielded grounded boxes separating internal equipment from the bock with insulation there are some DIY methods with chicken wire, microwave ovens , and surplus ammunition cans that may work effectively if used and designed properly, but yo would probably need to test it. See if you can find any old NASA publications on EMP shielding, which is extensively used on spacecraft to protect systems and sensors from EMP from the sun and other space events

http://www.dpworkshop.org/dpm-eng/resources.html#obsolete http://www.dpworkshop.org/dpm-eng/oldmedia/obsolescence1.html http://www.dpworkshop.org/dpm-eng/oldmedia/index.html https://en.wikipedia.org/wiki/Digital_dark_age https://en.wikipedia.org/wiki/Orphaned_technology Some interesting reads on the topic

>>2173 >When are we going to be able to store >data forever, without having to worry about it? Never. See: 2nd Law of Thermodynamics.

>>2330 it really will

Why not share all your important non-personal (archival things) with anons and such on here and allow them to be spread to as many places as possible. Wouldn't that be the most effective for longevity?

>>5049 Datahoarders tend to be greedy recluses.

The oldest stored data on Earth that I know of are ~4 billion year old fossil records of the first microorganisms preserved in rocks. That's almost a third of the age of the universe. So if you want someone to read your message a billion years from now, I suggest storing it on/in buried rocks. It has a proven track record. Metals oxidize. Most other things disintegrate or decompose. Rocks just sit there (or float around in space) for billions of years unless they're actively changed by some external force.

>>2795 http://ronja.twibright.com/optar/

>>5049 just give me the bandwidth

encrypt the fuarrrk out of your data and upload to google drive or such

>>5125 the "cloud" is just somebody elses computer baka it still has to be stored somewhere on something physically, and thats just what the cloud is: a bunch of harddrives saving your data, nothing more

I'm reminded of old calculator ROMs small enough to see the bits visible in the medium. I'm also reminded of not-so-ancient paper mistreated and reused from only half a millennium or so ago in which the message was still readable. So, encoding binary onto metal, rocks, and paper all seem good ideas and, even if these are only reliable for a few centuries, having someone read and transfer the data before it's lost is a good solution to this issue.

>>5128 what about combining >>5091 (peper encoding tool) but engraving on platinum or tungsten plates/sheets, you would be able to seal it (in glass or acrylic or something to prevent scratching/damage like how tree sap formed amber and preserved small animals) with plaintext instructions on the other side on how to retrieve the message

Egyptians figured out papyrus and it's all still readable, so you could learn some of their ancient recipes

>>5127 The Cloud is not just "a bunch of harddrives"; it's an organization's dedication It's the guys who are going to routinely check those hard drives because they will get fired if the data dies because their company needs a reputation for reliability Like the other guy said, GDrive is great. Google will never ever ever go bankrupt, and you know for sure that they're going to warn you if they're ever shutting down Drive. Or that Amazon Iceberg thing, whatever they call it You're not renting out "the hard drive space"; you're renting out "storage, under these terms". They could be using magnetic tapes or M-Discs or solid-state drives or ANYTHING. It's immaterial. You're storing it in the company hosting said cloud. "Other person's computer" is accurate, part of the fact of its being is their maintaining it.

>>5181 This got me thinking Would it be wrong to say the best way to reliably preserve information across time is to manually move it from an old substrate to a new one? I'm not saying long-lived substrates are a bad idea, but keeping information actively circulating probably keeps more of it preserved.

>>5285 That's exactly what language is, the only thing that kills the data stream is destruction of the media that allows the chain to be broken, this would be the major flaw of constantly transferring data, but as can be demonstrated by cultures who still speak the same language (or a variation thereof ) that was spoken thousands of years ago, Hats off to people who devote their lives to committing certain texts and other information to memory instead of writing it down, by comparison, you can obviously see how flawed that system is too, suppose your data got corrupted by faulty hardware, cat jumped on the table and corrupted the data stream, you just forgot to backup certain files etc. you introduce more danger during the transfer phase, but if done securely, that's basically the alternative to the cloud (having someone keep updating and backing up/transferring for you)

What makes anyone think that anyone else will give two short and curlies about their data in 500 years? Are you JS Bach or something? >>5144 > it's all still readable Except for the overwhelming majority of it that didn't survive at all.

>>5287 Like >>3825 said, the purpose of information preservation can be to stray future generations from danger, or it could be you just really want to have someone a million years from now see that .jpeg of your tiny benis (if it can be read at all, like the south park episode where cartman goes to the future and cant play his gaming system because of the lack of reverse compatibility), even some ancient texts and wall writings actually have vandalism (egyptian glyphs included)

> But we can't know if the caves themselves were particularly sacred spaces. It's possible that Paleolithic rock art was concentrated entirely in caves, but it might also be true that caves, sheltered from the outside world, are simply where these images survived. It could be that the people of the Pleistocene made their entire world into a gallery, that animals charged across every rockface, that wherever the tremendous herds of Ice Age beasts roamed, they were surrounded on all sides by echoes and images of themeselves, in a world where image and object had not yet torn themselves apart.

>>4598 Good stuff

>>5286 I thought so, but it's nice to hear it from outside my own head. Though I didn't think of language first; in my daily life I think of it more as one torch lighting another because of the way I make my backups. I still use mechanical hdds since they're so cheap, but when one shows signs of wear, I copy everything on it to a brand new drive and use the aging one for data that's easily replaced until it finally fails.

>>2950 >>3255 what kind of tape? i have plenty of experience with audio reel-to-reel/cassettes and those need to be stored in climate control or they'll practically turn to dust. plus tape degrades with use. if i needed my data to survive the apocalypse, my MO would probably be "find some crazy polymer that takes millions of years to fall apart and punch holes in it". pragmatically speaking i think the best way to save data is just to make sure there are plenty of copies floating around.

>>6571 Google and facebook use high density magnetic tape for user profile backups https://newatlas.com/sony-ibm-magnetic-tape-density-record/50743/ https://www.itbusinessedge.com/cm/blogs/mah/google-gmail-outage-underscores-relevance-of-offline-backup/?cs=45779

you will die also eventually anon

If you want data to be preserved just make it really really good and it will be

>>6627 that's not how it works

>>2173 That's becouse you are a nooblet, i have been using the same floppy disc since 1972, in fact my OS is right now running on one 10/10 it has outlasted several HDD's and solid memories. Btw HDD's and Solid merories are a scam

>>6623 So... If you want to store some data for long amounts of times, you make sure it shows up in your G/F user profiles? Well, rev up those captcha breakers...

>>2173 Welcome to my world. https://en.wikipedia.org/wiki/Impermanence

>>6627 >>6639 fuarrrk this is really deep. I mean, look at the bible. I don't know if "good" is the way to go, but there definitely is a way of using civilization itself as a storage medium.

>>6672 >I mean, look at the bible Dozens of different translations and iterations on basis of same basis material? Hundreds of different denominations believing different things, cherrypicking from Bible? Misinterpretations, both willful and otherwise? >using civilization itself as a storage medium hah, good one Putting an idea in ever evolving culture and thinking you will get the same thing when you come back to it few years after is either wishful thinking or you never were around to play the game of telephone aka "chinese whispers" with your friends.

Really stupid and very impractical idea, but what if we stored data by having it chiseled into stone or some more sturdier than stone? and one day when we have the ability to shrink things or something we can infinitely store books, source code, recipes for a very long time with stone tablets. Really, really dumb, because imagine giving your kid all the Harry Potter series in stone tablet form.

>>6678 Congrats, you just invented the dvd. Just a "stone" tablet, with very small engraving writing a binary message.

>>6681 I get what you mean, but DVDs aren't nowhere near as impractical as a literal stone tablet. And nowhere as funny to imagine.

>>6682 stones are notorious for environmental damage, acifdification and other hings in water cause erosion. probably the best "stone" would be things like titanium or tungsten, or other metals that are extremely strong or corrosion resistant but not brittle. now carving things into stone and hiding them on the moon would prevent environmental damage, and maybe the ayys would find them when they walk their moon dogs

Threads like these always make me think of the inevitable. I've lost a lot of VHS tapes that I had as a kid. I need to seriously put my home movies from the 90s onto a DVD as well as digital back ups. Does anyone know an affordable way to do this? I know some of the VHS DVD recorders are pretty expensive or at least they were when I last looked them up. I also have some old video games that are already starting to develop disc rot. And What is the current lifespan of typical 4Tb hard drives?

Take a copy of your files and bury it underground https://youtu.be/Li4J1wFNzik

>>6712 how would that help?

Our rulers don't want us to have technology that lasts. That would cut the money coming to them. =D

>>7155 Sad but the honest truth and it'll always be this way. Unless one of us invents our own resolution to the problem. Even still I don't see that happening ever.

>>7156 Getting rid of rulers and hierarchies in general is the solution, but I don't think we can make it happen within our lifetimes

>>2223 >M-Disk Best we have right now, give it a shot, OP.

>>7166 "There are decades where nothing happens; and there are weeks where decades happen." t. bald russian statue man

>>4144 >underground >not ve affected by living organisms What kind of underground are you talking about?

>>2173 Best you can do is backup periodically hdd raid, cold storage, and cloud backup is best M-disc might work but it's dubious (no practical test) fuse based prom or mask prom within any container may last over a geological timescale(still wears away though) longest surviving storage medium of the human race I can think of would probably be the Voyager golden record It's out in space, so matter interaction would be sparse (assuming it doesn't crash into something), only source of wear would be from ionizing radiation from cosmic processes transmutating the atoms on the record (very long time) and onboard Uranium-238 decay doing the same (very long time and there isn't enough U238 to completely destroy the record to begin with; most it can do is corrupt a small part of the record) Nothing lasts forever, not even the universe

Never. Any physical representation of data degrades with the universe and requires genuine, skilled effort to counter reliably; if you don't want to worry about it, you will not fit the job. yet you cannot reliably outsource the job to anyone else as they have other reasons that they will always prioritize over you. If you only want a good enough solution, you should probably go with redundancy: use all the various methods at once, they have negligible chances of failing all at once.

datamites are corrupting your porn as we speak Present Day, Present Time! AHAHAHAHAHA!

Nothing lasts forever, immortality lies within our offspring. Use that floppy, make lots of copies!

>>7551 Sterilized and hermetically sealed safe. The underground part is to keep humans from finding it

>>9045 No thanks, I don't feel like spending a chunk of my life being a slave to my genes.

>>6702 >What is the current lifespan of typical 4Tb hard drives? Don't worry about hard drives and just keep backing your important stuff up. And should obviously be caring more about SSD than just HDD.

>>2994 why rotating?

>When are we going to be able to store data forever, without having to worry about it? When we prove the law of the conservation of energy invalid

>>9166 You can't prove a law invalid though. It is a law, so it must be true. It might be better to say that the current models which have this law may not be accurate to the physical world.

>>3799 >Grandmaster Flash Had me a smirk at that.

>>9045 >use that floppy clever

>>2290 Rom as in cartridges? those already are degraded. Its rare to find a working sega game these days.

>>4137 IA or host an onion service. Simple HTTP repo

>When are we going to be able to store data forever, without having to worry about it? It makes me anxious to think none of this stuff is permanent The same way every big company does it: 1) multiple copies 2) verify integrity through hashes 3) write to new media regularly (especially tape libraries age when you access the catridges) 4) repeat forever until you die or they archive you on tape so you can continue working

>>11272 260 full-file passes, 5000 load/unload, and 30 years shelf life to be precise

>>2223 > M-Disk Am i reading this right? $22 for a 100GB disk? That's almost as cheap as a flash drive for storage that can last multiple human lifespans

>>3255 I have old reel to reel tapes (1/4 inch 4 track and 1inch 16 track) - they are not quite that fussy, I have had them just kicking around in box storage for 20+yrs now - they still sound and work ok for the most part, but 1 small issue is because they are rolled up magnetic tape (same with my normal tape cassettes) I can hear sound from 1 part of the tape has over time, transferred to the layer of tape above, resulting in a very very week recording pre sound. i.e. you hear the music very softly 2 seconds before its recorded place and full volume hit. obviously for most digital storage this will be filtered out, but if data density is very close to tape max it might possibly start to cause data loss.

>>2604 >>2177 >I think the real solution is to keep your data moving In theory, if we set two drives to constantly be transferring contents between eachother, would this be somewhat effective at giving longevity to data?

>>11497 Are you pretending to be retarded?

>>11499 If you continue to swap out the storage media this could be viable

>>11503 Well, of course it would be. But if I'm reading his post right he assumed you would just be moving it between the same drives, and called you an idiot for the suggestion.

>>11497 Holy soykaf lainon just invented RAID over here.

>>11499 >tfw not pretending

>>2173 Accepting the impermanence of things is an important step towards enlightenment

Distributed File Systems with redundancy exist. A drive fails, you add another. Life itself is primarily an information propagation medium with the ability to determine better methods of said data survival over time. Embrace Life: Every cell of your body contains your genetic information. You are a decentralized cluster of data, incredibly redundant and able to heal all but the most grave wounds. Continued replication of the data is paramount to the survival of such data. The battle against entropy was already won. The result is sentience. Now the task of the awakened ones is to look back at the path they have walked and consciously apply the ancient principals to their daily lives.

>>11632 the battle was won but the war is unwinnable. Eventually all information will be lost.

>>11632 nigga dont know bout my regenerating jellyfish new proposal >write data to DNA, increasingly cheaper in modern times >buy a bunch of special immortal jellyfish >embed jellies with "junk" DNA, their bodies are simple enough to potentially handle it >potential to encode DATA into jelly genes so offspring will also be data carriers (i.e. self replicating storage) induce trauma to keep jellies from dying (since they just turn back into polyps and then restart life cycle) >getting stung every time you want to recover your long lost dank memes/10

Sell me on checksum filesystems. What are they? Will they save my data?

>>11701 IIRC checksum just ensures things are still running smoothly, it does not prevent damage or degradation to memory

>>11700 t. soon-to-be molecular biologist One problem, eucariots "leak" DNA. After about three or four reproductive cycles, jellyfish won't have that data. Invitro storage is far better

>>11702 Error correcting codes like CRC32 can fix one-bit flip errors. There are other error correcting codes that can fix larger errors

>>11721 CRC32 is a compression algorithm. The best error correcting code is simply a comparison of many uncompressed [attempted] identical copies. See: >>11632

>>11733 >A cyclic redundancy check (CRC) is an error-detecting code commonly used in digital networks and storage devices to detect accidental changes to raw data >CRCs can be used for error correction From https://en.wikipedia.org/wiki/CRC32

delete your data, you don't need it

Data permanence scares me more.

>>2330 Until they catch fire of course. >>5181 >it's too big to fail! A bunch of hard drives owned and maintained by a big company are still a bunch of hard drives. soykaf happens.

>>11744 > CRC is an error-detecting code Learn some information theory. CRC is able to detect and correct some errors because it stores a highly compressed and incredibly lossy representation of the data it is REDUNDANTLY checking. The post about multiple redundant copies being better that a compression algorithm like CRC or SHA fingerprint is correct.

>>11733 >CRC32 is a compression algorithm. Compression implies that you can get the original data back in some form which is absolutely not the case for CRC. >>12414 >Learn some information theory. The description >>11744 gave is quoted from Wikipedia. It doesn't say compression anywhere on the page because CRC is not a compression algorithm.

>>12414 Well aren't you a pedantic little nothing. >>12449 I think all error correcting codes can be used for compression. Compression using LDPC in particular has been the subject of a lot of R&D for a while now.

>>12381 I'm sure it does, CIA-man

So Lainons, you all have inspired me! I want to see how long data lasts on flash and optical media when in cold conditions. I want to see if it's possible to create a time capsule and have a CD, a thumbdrive, a DVD, and maybe some different types of memory cards. I'll buy some small ones and write all kinds of test data to them. I want to see if the capsule can be made out of things poor people have access to. I'm taking inspiration from geocashes. Like peanut butter jars or what not. I'll bury it in a public park. Let it set through the harsh winter. Then I'll read back the data!

>>13714 >the year is 2143 >humanity is recovering from WW3 after AIs were given autonomous control of military hardware and politicians made fun of them >Little is left of society from before the war, just a few computers that were not connected to the new when humanity destroyed the wired >unfortunately, the modern data standard was wireless holo reading, which none of the recovered computers can use >but today, a reconstruction and salvaging corporation discovered a number of vessels buried in what appears to have been a suburban yard from before the war >ARE THESE HIDDEN DATA CACHES WITH INSIGHT INTO THE PEOPLE OF THE PAST AND LONG LOST TECHNOLOGY PRESERVED FOR ALL THE AGES!?!? >a dozen or so flash drives in a variety of food and beverage containers are presented to the neocommunist genderneutral leader >though they have been sitting underground for over a century, they may still be functional >the mayor plugs each flash drive into the computer following detailed instructions someone had scribbled on paper and taped to the side >autorun.exe >ITSWORKING.mp3 >the mayor opens the folder labeled important data >2 terabytes of cat videos, pepe memes, smug anime girls, and furry porn >the mayor announces all the data was corrupted and all the flashdrives are subsequently burned, the memory of that fluffy animal thing's massive cartoon benis eternally burned into xer memory

>>13720 Lay off the ether, it's hard on brain cells in the long term.

shatag (https://bitbucket.org/maugier/shatag) is a program that adds checksums to any files' extended attributes. It can be used to detect silent data corruption in a filesystem-independent way. I've started using it about 15 months ago and it already saved my ass. However I don't see it in many distros' repos and there aren't many people who know about it, so I'm posting here because I think it's great and more people should know about it. Of course one could always use a proper checksumming filesystem, but they're all too immature or a pain in one way or another, so meanwhile I think shatag is a great way to get some checksumming on good old, stable and proven filesystems.

>>13720 >2 terabytes of cat videos, pepe memes, smug anime girls, and furry porn Specifically because you posted this, this will be what I put on the drives.

>>2173 Let it be a relaxing thought that your consciousness is finite and all will end some day for you, without risking eternal life

>>13720 >>13714 >>13811 I hope it's not too late to add this

https://en.wikipedia.org/wiki/Linear_Tape-Open ;)

>>15004 That soykaf demagnatizes my boi, Yall niggers need some M-Discs

>>15004 I actually used to use those until I got a SSD

>>2173 Just think - thousands of years from now, technomonks on some island blasted by nuclear weapons will painstakingly trawl through vast data archives, using rickety generators powered from home grown rapeseed oil, and diligently copy this thread into a book of vellum for future generations of technomonks to copy. Kings of petty villages across the barren wastelands will spend enormous sums of turnips and bottelcaps to acquire copies of our ancient conversations, as both a means to declare status and to have it read to them while dining on what was once, deep in its genetic heritage which is now long since mutated, a squirrel.

Sites that I use to download good old things: https://winworldpc.com/library/operating-systems - Many operating systems before Windows XP and several old tools. https://www.myabandonware.com/browse/platform/ - Old computer games: Apple II, DOS, etc. http://ftp.romhustler.net/roms/psx - Consoles games, good for PS1. https://the-eye.eu/public/ - Archive site with lots of content: roms, OS, documentations, softwares, pepes, etc. http://web.archive.org/ - whenever a site returns 404 use the time machine. Anything else?

>>14992 maybe data impermanence is a good thing

hey lains, whats the best long-term digital data storage method, in terms of time until the data will be unreadable(in ideal atmospheric condiions)?

>>16150 tape, assuming there's not to many magnets around you

The oldest data we have is engraved in stone, so Id say that's your best bet.

>>15287 if the entire universe is just data, is it though?

>>2223 >Nothing is permanent, eventually even black holes will disappear and it'll be nothingness again. that has to be the most poetic thing I've ever read on the permenance of data storage...

>>3823 platinum plated with rhodium would probably be the most corrosion resistant. very few things will corrode rhodium the longest lasting would probably be pure diamond (synthetic should be fine) though, probably also one of the more difficult things to work with. you might be able to meet a middle ground by useing tungsten carbide and having it plated with Rhodium after inscribing on it.

>>12400 Store them in multiple big company's hard drives then

>>13714 any word on how well the data held up over the winter, or are you just going to leave it and use it like a time capsule?

I doubt that there's any way to keep data permanent. Whatever you do, one day it will be gone.

>>2200 but it exists. two of them belong to musk, one of them is on board of his space tesla.

I've always wondered if it would be possible to devise some kind of hash algorithm that would take an input file, generate a hash of it, but unlike a normal hash algorithms which are deliberately hard to revers, would act more like a seed, which the original data could be "re-grown" from. sort of like a seed world for a game that has procedural generated maps. so re generating the file would take significant processor time but the hash itself you could write down on a piece of paper, or cut into a stone or whatever. no idea if that would actually work

The use of etched glass works like a multi-dimensional QR code IIRC, and it has been around for some time. The problem is that read/re-write becomes difficult (unlike paper or stone).

>>16438 I think the system would have to be increasingly difficult to produce. If you used >Paragraphs based on sentences >sentences generated based on words >words based on letters you could potentially get a code word, or a sentence, but you are limited based on the structural composition of language itself. Think of it as a highly complex digital shorthand that translates back into text. I only put about 2 minutes of thought into this, so I am sure that there is some critical flaw to such a system (Like having to remember long phrases with absolute accuracy) IIRC this is how mnemonic devices work >ROY-G-BIV (phrase to remember visible light spectrum distribution in order, i.e. rainbows) >RED-ORANGE-YELLOW-GREEN-BLUE-INDIGO-VIOLET >*EACH WORD REPRESENTS A SENTENCE*

>>16440 I was thinking more along the lines of "Tupper's Self-Referential Formula"; in that its a plot of all possible arrangements of pixels and to find your picture or whatever, instead of saving it you just remember what number along the pattern it is. self referential formulas only deal with pixels and graphs so I was wondering if it could be extended to files and general storage. if a file you'd want to save is just a specific arrangement of 0's and 1's, and if you produced a complete list of all possible arrangements of 0's and 1's (for a given file size) then your file would be in there somewhere. then all you'd need to do is remember the number in the sequence and you'd have your file. no idea if this is practical though.

forgot to post this

>>16438 This sounds like a compression algorithm to me.

>>16439 >5d dem wot

>>15031 >I actually used to use those until I got a SSD I'd like to know how that worked for you.

>>16442 >if you produced a [sequence of all possible messages] then your [message] would be in there somewhere[;] all you'd need to do is remember the [index] this is the old-as-dirt argument, the "perpetual motion" machine of information theory tl;dr: TANSTAAFL/entropy makes this "right out". ("the classical" argument against this basically says "read the halting problem and lurk moar research, you can't compress information for free", but imma try to "actually" explain it here Like the difference between saying >this perpetual motion machine can't work because 1st law of thermodynamics: conservation of energy AKA perpetual motion machines aren't real vs >this perpetual motion machine can't work, because if you look <here> you'll see that it's exerting <x> force on the apparatus, which if we write the formulas for the force generated by the "work-doing" part you'll se physically cannot be more than this I intend to do the latter so, that said, I'll proceed with the post) say you've got a 40096-bit message, you want to uniquely identify it from a pool of messages big enough to contain it well, from a pool of >1044388881413152506691752710716624382579964249047383780384233483283953907971557456848826811934997558340890106714439262837987573438185793607263236087851365277945956976543709998340361590134383718314428070011855946226376318839397712745672334684344586617496807908705803704071284048740118609114467977783598029006686938976881787785946905630190260940599579453432823469303026696443059025015972399867714215541693835559885291486318237914434496734087811872639496475100189041349008417061675093668333850551032972088269550769983616369411933015213796825837188091833656751221318492846368125550225998300412344784862595674492194617023806505913245610825731835380087608622102834270197698202313169017678006675195485079921636419370285375124784014907159135459982790513399611551794271106831134090584272884279791554849782954323534517065223269061394905987693002122963395687782878948440616007412945674919823050571642377154816321380631045902916136926708342856440730447899971901781465763473223850267253059899795996090799469201774624817718449867455659250178329070473119433165550807568221846571746373296884912819520317457002440926616910874148385078411929804522981857338977648103126085903001302413467189726673216491511131602920781738033436090243804708340403154190336 messages, how many bits will it take to pick out your message uniquely? If the message-IDs are mapped uniquely to messages, 4096. Now, if your message occurs before the halfway point in the Giant Message Sequence then you'll be able to just drop the high bit off your 4Kib message! (25% chance of saving 0.0244%, if you're using 512B chunks!) If it's even closer to the start, before the quarter-way-point, you'll be able to drop the highest TWO bits (12.5% chance of saving 0.0488% on your 512B chunks of data!) etc etc to infinity, assuming your indexer-program-thing maps messages to keys in an essentially uniform[ly random] way. So there's a 93.75% chance that your message-index-in-the-list will be less than 0.0732% shorter than your original message, ie you've got a ~94% chance of having an compression ratio worse than 99.9%, and a ~99% chance of having a compression ratio worse than 99.8%. (And that's in the extremely generous case of you're using 4096b==512B blocks; these numbers actually get proportionally _worse_ the larger the messages are. And if you try to just chain a ton of 512B messages together, you'll have to spend AT LEAST AS MUCH, BY [entropic] LAW as you saved "compressing" them into mere sequence indices, because you can't just "assume they're all $originalMsgSize"; the space you saved was due to the fact that the message length was changing!) I hope this at least communicates the basic concept, as to "why this perpetual motion machine doesn't work", on a technical level then you go further into actually "messages are low-entropy if there exist measures/orderings/indexes [whose definitions/generator-algorithms are themselves low-entropy*] which the messages are very very very early in the sequence, so much so that" for instance, the source code for GZip algorithm, DEFLATE, is kinda big, but most commonly-transmitted or likely-to-want-to-transmit messages are so very far [literally]exponentially earlier in the "index" of the enumeration of all GZip-compressible strings (i.e., their compressed counterparts are earlier in the enumeration of all—technically, valid GZip-decompressible messages—but what matters in the real world is, their [the indices'] position in the sequence of all messages) than they [the messages] themselves are in the sequence or enumeration of all possible messages *yes, entropy/simplicity is context-dependent and abstract, hard to quantify, and not as _obviously_ physically "real" as, say, potential energy, but nevertheless, its conservation seeming to be just as ironclad as a law

>>16442 >if a file you'd want to save is just a specific arrangement of 0's and 1's, and if you produced a complete list of all possible arrangements of 0's and 1's (for a given file size) then your file would be in there somewhere. then all you'd need to do is remember the number in the sequence and you'd have your file. Think about it this way: if you made a sequence that is just like 0,10,11,100, ... then your file which is the bit set F would be indexed in that sequence by the number F. Your file is *already* an index in a sequence. But, there is a catch: if you want to *uniquely* assign numbers to bit strings, you need at least as many numbers as bit strings, otherwise you would have one number which refers to two bit strings by the pigeonhole principle. So you need 2^N numbers to refer to bit strings of length N. There's not a way around this. The whole theory of lossless compression algorithms is trying to find such sequences(aka compression algorithms -- the output represents the index in an abstract way) that put files we are interested in into small indices, so they are more convenient to store as their index. Most strings humans are interested in are very patterned. If u kn engl u cn rd ths. I just reduced the size of that sentence by almost half and it's still readable by a fluent English reader. We exploit patterns in the things we want to compress to compress them. Very random(unpatterned) data is very hard to compress and also usually very uninteresting to humans, so we don't worry about it too much. >>16489 I can count 4 that are obvious from that picture (3 spacial dimensions and color) maybe they're connected by some graph scheme that gives them an extra dimension?

>>16150 Tape drives are usually the best option for the long term storage, the average lifespan is about 40 years (but if you keep them in good condition they can probably last longer, if you don't then they will not last anywhere close to that) good luck getting a tape drive though, they usually cost around $5000 and the tape cost about $20, but the prices vary depending on the amount of storage you want, the cloud is probably the best method for long term digital storage, but remember that the services can (and probably will) shut down after some time, so keep a back up in mind if you want to use cloud storage.

>>16565 the bit set F would be indexed in that sequence by the number F tiny nitpick: it's a bit more complex than that, since these files are all unique (one line per file) > 00 > 00 00 > > 00 01 > 01 > 00 00 00 00 00 00 01 but I want to emphasize that that's a tiny point and what you said in that post is correct, especially this, which is the key: > lossless compression algorithms …[are sequences] that put files we are interested in into small indices though I, that the compressedmsg having "small[er] indices" is understatement of the year, to compress the data by a factor of X, you've got to have the compressdata's messageindex be the following amount > 2^( X × len(data) [× 8 if using bytes] ) times smaller than the message's messageindex in the [same/]used medium (and an interesting note: this is "detrimental" from a pure-index-conservation standpoint, but realistically not a problem and actually kinda beneficial given other scarcity consideration, that there are (ironically) actually many possible compressdata strings fed to a given compliant decompresser, hence the different levels of compression available for a given decompresser. Compression [ie, given M, finding a message C, such that Dr(C)=P and index(C)<<<<<<index(M)], is an art, but Decompression [ie, executing D(C) where D(C)=Dr(C) for all C, Dr being the reference decompresser] is in most cases a robotically straightforward process )

>>11272 >write to new media regularly >(especially tape libraries age when you access the catridges) u wot if accessing it ages it, shouldn't you do the opposite and not make a point of accessing it regularly?

>>16583 New media means fresh new cartridges that haven't been used(heavily anyways). Even new old stock tape media suffers from degradation with sticky shed, stretching, etc. Only real way to guarantee the safety of your media is multiple formats, multiple backups. Even if the "cloud" is like kicking dead whales down the beach from a privacy point of view, it's still a pretty solid idea if your data's encrypted.

>>4596 Faraday cage

>>2803 `cat file | col -b | fmt | lp -`

>>2173 Entropy and the degradation of all physical systems will always be a fact of life anon.

you can store data in the blockchain

>>2223 This, it's good that nothing is permament, we need to move forward. So as I mention, it's good but it makes me anxious that we don't have any good commercial storage mechanisms to store data for at least >50 years.

Aborigines managed to retain data over 6000 years. Human Raid-System with people and frequent checkups. They got geographical data encoded in stories that would be retold frequently with everyone sitting together and correcting each other whenever they remembered it different by consensus.

It's funny to worry about losing data. I mean think about all the Greek and Roman fan fictions we have no idea about because they just weren't that important. Do you wake up sweating because you don't get to read some dudes romance novel of Zeus storming his anus? No. Because you don't even know if it exists. Everything you have now will probably live as long as you do, so who really cares if the last Harry Potter bluray set is lost in a plane crash 70 years from now. I'm pretty sure movies aren't even going to be as important as they are now. Hell, most kids might even see it as primitive garbage. Living in the now is way less stressful, since you have actual control over it.

>>18188 >Do you wake up sweating because you don't get to read some dudes romance novel of Zeus storming his anus? Kek I kinda want to read that now. No homo

>>18188 May be true, but consider how happy historians are when some graffiti is discovered in Pompei or some piece of racy message between lovers pops up from 2000 years ago. Hell, the first written example of my own native language is a piece of parchment from the 8th-9th century with a stupid riddle on it. Don't be so quick to discount the value of mundane pieces of information to historians. History is not just made of wars and kings.

>>18188 > Do you wake up sweating because you don't get to read some dudes romance novel of Zeus storming his anus? No. Because you don't even know if it exists It does now

>>18209 This. I honestly think people think people are getting dumber because we don't have enough documentation of commoners doing dumb soykaf back in the olden days

>It makes me anxious to think none of this stuff is permanent Hoo boy have I got some terrible news for you about literally everything that exists.

>>16566 Yeahhh we don't normally term things like color or anything of that sort a dimension though. Would you call a painting three-dimensional if it was in color but two-dimensional if it was black and white? I've also read that the term 5d has something to do with the angle of the reading head, which also isn't a dimension. Any angle in 3d space can be described by a 3d vector; no 5d needed. Just kind of reeks of marketing hype, which makes me doubt its claims of longevity as well. Pretty sure any media would last a bajillion years if you stuck it in a heavily shielded box and cooled it to near absolute zero. When CD-ROMs were brand-new tech, similar exaggerated longevity claims were made about their durability. Turns out they were based on ideal engineering models and shaky extrapolations and ignored countless real-world factors, which is usually the case. I would be skeptical of the claims for these; probably off by orders of magnitude or simply fabricated.

>>16159 Yes. I welcome the heat death of the universe, and my only regret is that it will be 100 trillion years too slow. Something has to erase my knowledge of this: >>14992 Oblivion, take me

>>2204 Can't tell if trolling or not, why the hell would information want to be lost ever

>>18225 Information wants nothing. *You* may want to preserve it. And >>2204 is saying that you don't want to preserve all the information. That's the predicament with facebook and co. Dumb soykaf you did when you were younger will haunt you forever because data doesn't go away in any meaningful sense and on a meaningful scale (say 10-20 years, enough to make your life misarable).

>>4137 torrents or directconnect (think dc++)

>>18143 yeah but universal heat death is just the big bang so dont worry about it

>>11497 Well then it depends on the technology you're using. If you were using solidstate memory constantly being in use would degrade the nand gates. If you were using a mechanical hard drive the drive will eventually experience mechanical failure. the only way to preserve things long term is carving it on rock and then protecting the rock in a temple or some kind of bunker.

>>18385 lol no it ain't. Maybe you're thinking about the big crunch?

>>3869 keep it well out of sunlight. If I've heard correctly.

>>5091 Very interesting, 200KB on a single A4 page is impressive. These kind of encoding schemes are always so interesting to me. I wonder though, all these encoding schemes use only black and white units, wouldn't it be more efficient if they used different colors?

>>18599 >I wonder though, all these encoding schemes use only black and white units, wouldn't it be more efficient if they used different colors? Efficient, sure, but prone to bit-rot. B&W schemes are very effective against ink fading, yellowing paper, etc. Think of the increased number of printed pages due to B&W encoding as a dead-tree RAID array.

>>3825 lmao preserving the locations of nuclear waste on a sapphire hard drive? Hopefully they have IDE cables in the year 32,019 AD I guess. That's dumb. Just carve it into a rock. Why do people think warning future humans about nuclear waste is a big problem? Just bury it deep as fuarrrk in the center of a tectonic plate and it's fine. Even if somebody miraculously stumbles upon it, it's not going to kill a statistically significant number of people over such a gigantic time span. Know what's going to kill literally all of us way before the thousands of years it takes for nuclear waste becomes inert? All our non-nuclear waste. And that stuff is literally everywhere. That's going to make all these long-term data schemes completely pointless if it kills us all in the next century.

>>18537 https://www.youtube.com/watch?v=FVDJJVoTx7s no im thinking about conformal cyclic cosmology

>>2173 The only issue is VHS. Digital media will be preserved as long as someone cares about it. VHS can even be preserved but I don't know what the limits are to remain authentic.

https://gizmodo.com/the-case-for-storing-humanitys-most-important-data-on-g-1839615098

redundancy /thread

>>18616 Fun fact: there's a lot more radiation per energy produced by coal-fired plants by concentrating natural uranium and thorium.

>>2173 I relate lainon. I listen to a bunch of obscure music (usually 500> views) and knowing that if the artist took down the video or the channel got a copyright strike it'd be all gone scares me. $waggot recently deleted their soykaf and I couldn't find it anywere on YouTube. I had to get some compressed russian pirate site verisan. Whats even sadder is that I get sentimentally attached to 'no one would care if it's gone' type files. I don't know were else to share this, but I'd like to post some 'almost wiped form the internet' soykaf on YouTube https://www.youtube.com/watch?v=4gNsveCCRLw&list=UUncNhxg7pqUoEy4cg8drykQ https://www.youtube.com/watch?v=NuWs_Nvjp4U https://www.youtube.com/watch?v=J5i75Qro9zQ

>>21893 I highly recommend using youtube-dl to download and archive these. I agree fully, it's a real shame to go through an old YouTube playlist (for example) only to find half of the videos missing :( Youtube-dl also handles SoundCloud and other sites, it's a great tool.

>>21894 Does youtube-dl actually still work? I've been using a version from a PPA that's been busted by youtube for ages.

>>21896 It does! :) --version shows 2019.09.28 for me but it gets regular commits on GitHub, sounds like the PPA might be a bit behind

>>21893 Can relate, more than with youtube obscure videos, even with fuarrrking yahoo groups content.

>>21896 youtube-dl and youtube-viewer tend to work when you use the github versions; ppa and other downstream versions tend to be out of sync. Youtube continously tries to stop people from downloading.

>>2173 there's no way to preserve data forever, but some mediums like m-disc are probably good enough for personal use.

>>21907 0.00000000003 laincoins have been deposited in your account

I've been working on something to OCR classic greentexts for archival from pictures to text files, so far this is what i've come up with #!/bin/bash echo "Resizing" convert $1 -resize 200% ocr.png echo "Converting to grayscale" convert ocr.png -grayscale Rec709Luminance ocr.png echo "Sharpening" #convert ocr.png -sharpen 10 ocr.png echo "Converting" echo "OCR conversion of file $1" > $1.txt tesseract ocr.png stdout >> $1.txt rm ocr.png echo "Done, output to $1.txt" Now this is like kicking dead whales down the beach and my bash programming is rusty, but it works altho it has trouble with multi-row caps and sometimes the ocr output is just blank, and i don't know why. I figured that greyscale and sharpening could help make the letters more distinguishable. Anybody got some advice or experience?

>>22228 >bash when not needed >unquoted variables Would you useless niggers at least use shellcheck on your abominations? >resize using the default filter instead of -distort Resize with some like LanczosSharp >two convert operations >sharpening using -sharpen instead of -unsharp Mate. >rm ocr.png Use a trap to delete it even if your script soykafs itself. Like so: trap 'exit 1' INT TERM HUP QUIT trap 'rm ocr.png' EXIT You may be rusty, but at least do the following: use /bin/sh unless you really need bash, put `set -eu` at the top or `set -eu -o pipefail` if using bash and run shellcheck on your script.

>>22231 sorry anon what i meant by rusty is that i totally suck

>>22243 >sorry anon dont empower replies like this. the way to get better with any shell programming is to run the script until you notice it does something you don't like, then read about how to do it differently. just do it the way you currently know how to do it, unquoted variables and all, until you run into a case that makes doing so impossible. then you will know why you shouldnt do that (the reason is not because some dickhead reply said so, it's to avoid unintended expansion issues). using traps without knowing why won't help you use them in other scripts. just hack and let the learning happen organically.

>>22254 That's how you end up with a whole infrastructure not handling spaces in filenames (I have to fix this soykaf at work) or very bad performances because "lol, it's bash, it'll be slow anyway". This idea that sh is a special language that you can properly use while only knowing 10% is retarded and lazy. A hacker meant someone who strived for some kind of excellency, at some point; don't confuse it with lazy bums happily and knowingly rolling in their mediocrity and ignorance. By the way, I may have roasted him a bit, but that was also to recommend shellcheck.

>>22256 we're all looking forward to the story of how you wrote your first shell script with 100% knowledge of all shell programming languages before starting to create it. you can post it on /g/ and link to it in this thread

>>22256 i did literally just google the most basic "how do i X in bash" because i don't bash script often, i did write better than this but i couldn't remember or find the previous ones

>>2547 I look foward to the day where I can breed my computer with other strong computers to form a super computer baby.