@@boomerz2478 FileCoin is literally designed to be the opposite way around lmao. It is to give people an incentive to use IPFS and serve those files. Additionally, having a browser-mined coin for those _viewing_ but not _serving_ would accomplish nothing. The entire purpose of IPFS is to have people download and serve to increase availability (and as a result, speed). Additionally, it's pretty peer-to-peer. If I visit an IPFS-enabled website, I download the stuff it needs, then serve them myself (until I delete it ofc). Also, they can't really close down or block IPFS, which is also what it was designed for. I mean, how can they block it? Just look at BitTorrent, they aren't able to block that either and well, torrents are far, far more notorious for illegal stuff. They can't close it down either, because it's open-source, so people would just fork and continue on as usual. And it probably already hosts a bunch of illegal stuff, you can never prevent that. Then again, every "censorship resistant protocol" is prone to this. Just look at ActivityPub (which is a W3C-recommended protocol), BitTorrent, TOR (more or less), Most cryptocurrencies etc. etc. And even *if* you manage to block it, it's a matter of time before a workaround will be found, which is exactly why those protocols are so powerful.
@UCv3D-0cNQeL7BRPE5RpO2LA Correlation does not equal causation. IPFS can be ran standalone just fine like I do. It wasn't designed to give more money to the whales, it's just a side-effect of FileCoin, which is a seperate project. Also, I never said IPFS was *fast*, I only said that by having more people serve a file, the *faster* it becomes. Faster, being a relative term. Also, they haven't blocked torrents lmao. In Germany, you can torrent just fine (I use torrents to send stuff like documents and some tools I made between a bunch of friends). Yes, they have blocked sites like TPB but they have not blocked the BitTorrent protocol. They can fine you based on things like finding you in the DHT for a specific (known illegal) file and/or scraping you from trackers. They can't fine you for using the protocol itself (else using the Blizzard Launcher would get you fined since that uses the BitTorrent protocol under the hood as well). Also, they can't really done it that easily. How would they? Blocking IPFS specifically would mean blocking the underlying connection protocol (be it TCP/IP or UDP). They can block commonly used ports, but at that point, just swap ports. They can block every port but the ones they allow, but at that point, a lot of different stuff will also start breaking. The Great Firewall of China does exactly this, block a lot of ports and only allowing a select few, that they can monitor heavily (for obvious reasons). Do that in, let's say Germany or The Netherlands and well... you're bound to cause massive uproar. And ofc, as you know, in China, they'd probably just gun you down if you start rioting buuut that's a different topic. Yea ehm... suddenly your "most of your points are just incomplete and largely wrong" starts backfiring to yourself...
we dont have to necessarily just have one or the other. if private capital wasnt anti decentralization to the point of trying to sabotage it in order to maintain the status qou, anyway
Wait. This still has some big flaws. Since hdd’s are the cheapest storage medium most of the data will be stored on it. Wich would make the internet slower again. And if you alter a pixel in an image this will just still be a new image wich would bloat the network. And how would compression work on this network? We just all sending raw video files? Also wouldn’t it be more sufficient to have a big cashing database on mars instead of distributing it. So you can have tiered storage wich would make the internet feel faster because the most visited sites are the most requested. And then to communicate with earth you could work with quantum entanglement since distance doesn’t seem to matter with entanglement.
A data storage method which by its very nature requires infinite storage. I shudder to think what the backend looks like 20 years from now if it takes off.
I like where web 3.0 is going but I am very much worried about its immutable nature. And IPFS being core part of it, I wont be able to delete any uploaded content even if I want to. Not sure how everything will turn out
I think the current answer is blacklisting hashes. Nodes would have to voluntarily accept to ignore/block replication and sharing of hashes for blacklisted files.
@@projectpegasus1297 of course one expect those blacklists to be created and maintained by some legal entity, but we can imagine different entities providing different blacklists.
IPFS is cool but Filecoin is a scam. If they were not scammers they would've used Bitcoin to pay people. Bitcoin is the only money that no one can get for free. Every other form of money that someone somewhere is getting for free but you are presented at face value is a scam. If every project starts creating it's own coin we end up with the same situation that Bitcoin was created to fix - every country with it's own money and banks getting rich for moving money around, where exchanges will be the new banks and they will get robbed and lose people's money. It is actually happening already, search "exchange hack" and you will find articles about dozens of hacks with billions of dollars stolen from the people and all because of greedy people wanting to print their own money by creating their own coin for their project. Completely killing the point of the blockchain innovation. >.
@@dkf2711 are you saying only Bitcoin is not a scam... Why Bitcoin? why not Ethrium or XRP or Tether or litecoin or EOS or TRON or Stellar or Monero, what about Dash or Tezos... etc etc etc.... Is Bitcoin special in some weird way, please share.
IIIRotor Bitcoin was created to make the world better. Everything else was made to enrich the creators. Bitcoin is like Linux, it’s open and it’s community driven. The other projects are like Microsoft, they are businesses that have marketing budget and use it to create the perception of community that OneCoin and BitConnect did and also out of the crypto world Theranos which was the Ethereum equivalent in biotech. There is a reason why everyone that builds open source software builds new layers on top of Linux and not incompatible completely new kernel operating systems that you can show off idling on the desktop and say they are faster and more secure. That stuff works for startup scams which in crypto are ICOs but in general in the computer world if you want to build new software you do that on top of the most secure and stable kernel like Linux so that if your software fails the whole system doesn’t collapse. All useful new functionality is developed on bitcoin, it has smart contracts, anonymity features and scaling optimisations that actually work and are designed to be long term solutions. Everything else is recycling bitcoin improvement proposals from 7+ years ago that got rejected and is monetising on people not doing their research and throwing money at the next pump.
I liked this concept until it introduced yet another cryptocurrency. Why does it need crypto? And how are the nodes different from just having a community sourced CDN?
A long time ago, we used to download content (Movies and Music mostly) with IRC, Kazaa, and other apps/platforms. The files would be downloaded from beginning to end, serially. I worked with a team to create torrent downloads, where you can have multiple connections for a single download file and that the download would come in as "chunks" in no specific order, just fastest, best connection. The only credit we get as a team, is the message boards we collaborated on, so Long ago. This torrent method of data exchange now rules the internet world. Blockchain would never have existed without this tech. I don't care for credit, I am just SO HAPPY to see how far we have come since the 1990s.
I don’t get how versioning works… if you request Earth’s Wikipedia page using its hash, you will receive the page from the IPSF node from Mars. But what if Wikipedia updated it? The commit won’t be seen, unless you know your version is outdated
okay but, what if i publish a unethical and very bad video on dtube? is there a way to remove it? there probably is, but in that case, can't turkey or any other country, just block the IP address of dtube or the site version of wikipedia hosted on ipfs?
Reminds me of the early days of the Gnutella networks, Kaza, Bearshare and other P2P clients. They had hashing of files too for verification but I don't recall a file size limitation. There was no incentive involved for persistent nodes however.
You’d like use IPNS which can assign names to hashes to remember them more easily. Just like the current DNS system that translates a domain name into an IP address
Great video! I keep wondering though if there’s also a solution which makes it possible to store an entire interactive website and all its data in a decentralised way instead of hosting it on centralised servers. Anyone got a clue? Cheers
Can you make a video on the Oyster protocol? It takes a slightly different approach to the same goal of ifps, also adding volatile memory functionality. It is built on top of the IOTA tangle so could potentially be distributed across all smart devices in the near future
it wasn't. it was made to provide unavailability-resistance, temper-resistance, and censorship-resistance. all of which are desperately needed in the current economical and political landscape.
Ok but about the versioning- how will you know about a file v2 if youre content-addressing file v1? Assuming the old links to the new, but how there will be other copies v1 which dont link to v2. Is there active discovery for versioning (i.e. if i commit v2 i update anyone who queried v1 from me) or is it just passive? (ie anyone who will query v1 from me in the future will be made aware of v2)? Also how would merge conflicts work if someone else created their own v2?
IPFS has a great epic privacy fail! IPFS is a solution to get content from an computer near you to protect the long range internet data link and to reduce the latency of respondig. Rearly a great Idea! But if yo can find content on computer near you, you can also determninate on wat computers a specific content was seen. This is - depending on the content - enourmes privacy influencing. The classic example: You are using products of the Adults Entertainment Industrie. Or you have see stuff, your goverment does not like you see ist. For example some Hints about Korruption of Erdogan in the Tukey. Or some History reports About the Amenias Holocaust in Turkey. Within the USA i dont know such exact, but looking Videos from Parties the USA is in War will it also create some Questions from Official sides? To Bring IPFS to a success, it must include the privacy Ideas of IAN Clark's Freenet. For Example: the Chunks are Hashed, Then Encrypted with the Hash as Key and then a second time Hashed for Address creation. The Link is the concat of the both Hashes. This makes the Content Denieble. Wy there is a "Privacy Mode" in the Browser when you can look in the ipfs folder to check what was used? But for Every Content you Know you can Calculate the Link. So the Data must be never stored at the related Computer to make IPFS rearly a private transport layer. And also it ist Important that the Content never comes direct from the source Computer. How to do this is all finaly evaluated in Ian Clark's Freenet. In Order to Avoid "Cambridge Analytica" by everyone, it is Madatory to include Ian Clarks privacy Ideas to IPFS In its current form it is only siutabel for Operatind system Distribution.
I disagree, the data is so broken up and spread out across the network that you wouldn't be able to get a clear picture of what files were seen where unless you too had seen the files, but your very access of that file will effect the network if you look at it. It's almost like a box that destroys itself if you open it up to see what's inside.
@@georgekaplan6204 sorry you are wrong. The IPFS uses the methodes of Bittorrent to distribute Content ... and we all know the lawsuite against people who are downloading and redistribute stuff stored in a bittorent form without owners premission. This will work also in the same kind against IPFS like it works against Bittorrent. Using "IPFS" as standart layer is more destructive as the use of Bittorent. While using Bittorent you aknolege every single Download an you can thing about is it legal or not. Within a browser using ipfs as background system a malique copyrigt owner my download this unwantet work in the Bachground by using javascript an then … after removing the hidden downloader from the webpages … sue every of his victims on court.
IPFS is like the way-back machine on Mars, but cannot solve the issue of wanted freshly upended information in a network. This is the same issue of double-spending that plagues blockchains, and to say they "solved" it by only establishing the baseline scenario of the problem is scarily deceptive.
Thank you for your explanation as always! One question tho, what if I put a file online and I want to delete it? Let's say an embarrassing picture of me that finally don't assume anymore. It's going to be on the network forever, yeah?
Btw. the bigest problem isn't "to keep the files available" as you said, but in fact the bigest problem is "how to find your file". Just imagine there are milions of IPFS nodes and like 4 of them have the content you need. How do you find them ? .... and that's why IPFS is sooooooooo incredibly slooooooooow. As well that's the reason why IPFS will never become "interplanetary" because the metadata about "who has what right now" will just oversaturate the inerplanetary link. Just an example, my IPFS node I'm running to host a small website (approx. 18MB) is permanently talking with like 5000 peers and "wasting" on average 1Mb/s of my bandwidth. So it's not all that peachy as you all are explaining IPS. Maybe worth being more objective next time.
@@Moodboard39 In fact I was running my IPFS node on my Synology NAS in a docker container, but I had to switch if off, because as my IPFS node become more "famous in the network" it started to generate so many packets per second that it in fact overloaded the wifi access points of my ISP and I got threatened to be disconnected if I don't turn the service off. The funny part was that the actual network load in Mbps was negligible, but the IPFS node basically sendins soo many packets to soo many hosts that it overloaded the CPU and the routing tables of these mid-range AP's and managed to disrupt the internet connection for every user connected through that ISP provider in that area. So for me IPFS is dead - long live IPFS ;-)
I'm extremely interested in the IPFS. Can't wait to see it explode with growth. It would be good to see a way to host files with a currency other than filecoin.
@@basspoett torrent download speeds are actually quite good. If you are complaining that it takes an hour to download a gigabyte size then just lol....
I don't really get it about the "Versioning" capability of IPFS. I tried to find about committing file or uploading new version of old file to IPFS, but I can't find how it is really done (I still can't imagine how it is done practically, like is it automatic or we need a specific command to do that?). Is this versioning done by IPFS automatically, or is it done with other software/add-on/something that is separated from IPFS? I've also read about Commit File Object of IPFS, but I don't really understand about it and how it is implemented. Anyways, your video is so good. I am actually still learning english, but your video is clear enough for me to understand. Salam dari Indonesia 😇
I like this explanation, but it seems like this NFT "ownership" is completely meaningless, saying you "own" artwork, but not controlling publishing rights is like "owning" air, or Moon, as in there is no meaningful use you could derive from it. If you disagree I have a Moon NFT I'd to sell.
Who here would be browsing the web and go "ohh cat picture, let me see this". Pay wall appears saying send 1 filecoin to see this photo. Now I have to go purchase some token to view a file. I have to also learn what the hell filecoin and a token is. NO ONE IS GOING TO DO THIS! IPFS on it's on is just fine.
Hi, don't you think the core objective of this is same as the 'Substratum' Project. Their project is already live (early beta). What do you think about it? Thanks for making these videos.
It is basically torrent thing. This makes your Hardwear to be used extensively and so it will loose it life resource much faster. So to me it looks impossible. And not only that internet will be slow - your Hardwear will be used for not your tasks so that you will gave to wait more for everything. You wish to solve it buy dedicating special Hardwear? Ok, now you are at the same point you were before. There is no problems to make a baking server, you don't need a blockchain. You just need Hardwear, so you need more money, so you need to be rich so you need to be living in another dimension.
You can rent out space on your hard drive and get paid? Who's going to do the paying? Other users? The people who created IPFS? Honestly, while a decentralized system of storing websites and files sounds nice, I can't see it working in actual practice at scale. I just saw a torrent for a movie get uploaded ten days ago and it's already dead. Even with incentives, average users just can't be replied on to reliably store files.
What a disappointment. I was all for it until you mentioned the god damn coin. Not that I don't realize the point but involving "market forces" into this makes it totally not worth supporting.
is there a way to configure it so that if I pin a video, then link the video to a personal site, that anyone who visits the site to view said video gets it directly from me only and not other? (so long as my ipfs desktop app is running, they'll get it from me every time). my problem is that the video takes forever to load, but eventually does.. (once after an hour of loading), other times maybe like 5 mins. i want a "direct connection" between the file link and my storage. how can i make this happen?
on the issue of content versus location address, how does one get the right content, if you don't know the hash? do you have to trust someone that knows the correct hash for that content that you are seeking? thanks
This is something that IPNS can solve. This system allows you to link a hash to a name. That way it becomes easier to remember and to use. You can compare it to DNS. This links a domain name (google.com for instance) to an IP address of a server somewhere.
I should've subscribed a long time ago. Just did. Your videos are always brilliantly done. I'm going to go do even more research because this video was that interesting. Well done, Sir!
3:45 You're taking about kilobytes and yet what you're showing on the screen are kilobits which are a different unit. kb = kilobit. kB = kilobyte. 1 kB = 8 kb. If you're talking about an IT subject, you CAN NOT mistake these. It makes you look untrustworthy. I hope you know that already, since the video was uploaded 4 years ago. If not, well then, now you know.
Hello could you help me I'm trying to add remote pinning on IPFS desktop but when I go to pin it gives me an error telling me that the site is not authorized
So are you saying that all content that hasn't been cached can tak 45 minutes to download? That's one hell of a performance issue if you happen to be the first one to download the content.
What about privacy? How could you ensure that the files are private and that people that might store them can’t access them? Also, what happens when you want to delete a file? Will it remain available forever as long as people have the address to the file? Nowadays, once something is taken down, although someone might still keep a copy, it is kinda hard to efficiently distribute it. Would it be impossible to stop this using this system?
ShadPayback not sure exactly, but in theory you could store your private files encrypted, if the encryption is strong enough. You could only access them if you had the key.
I am sure the whole concept of distributed internet is so that you cannot delete/censor any file once it is out there. If you don't want people to have some file, store them on your own drive would be the better option. Once something is in the cloud, you should always assume it is shared with everyone.
Like others mention: it's indeed the goal og IPFS to work like this. Completely distributed and immutable. Not possible to delete a file (but it will probably disappear on its own because nodes will stop caching it when its not requested)
Yeah.. I believe content ownership is difficult target for all decentralized p2p file storage applications. The concept of ownership demands centralization.
@@simplyexplained ((but it will probably disappear on its own because nodes will stop caching it when its not requested))----- This is a option? or a time line obsolescence programmed is implanted by defect? I LOVE FOUND OLD CONTENT. Like my lovely old EMULE
