What is the difference between HTTP and IPFS?

IPFS is a revolutionary model that can change the way we use the blockchain Internet. IPFS is not like the typical server-client model that we are familiar with, but more like bitstream (this is a content distribution protocol, it uses an efficient software distribution system and peer-to-peer technology to share a large number of files, and makes every user like a network The redistribution node provides upload services as well)
Now the network has a problem.
HTTP is the foundation of the Internet (World Wide Web), and any website we visit must be transmitted via HTTP. In fact, this is a relationship between the server and the client. Our computer sends a request to the server of a website, and the server returns a response. The host website is a centralized management model, and this centralized management environment is dangerous.
If a major host company or service provider (such as Baidu, Google, Amazon, etc.) suddenly collapses, the Internet (World Wide Web) will suffer a catastrophic blow in the short term.
In addition to the centralized model, the reliability of today's network is also worrying. Most sites and applications are hosted by a server or a redundant array of load balancing servers. If the owners of these servers, the managers of the data centers, or even natural disasters make these computers quit, you can imagine the results that websites and applications will not function properly.
How does IPFS deal with these problems?
IPFS is distributed. When a user visits a website, there is no need for a central server to provide web pages. Each running IPFS machine will become a node in the cluster.
The current working principle of bitstream is that if we are downloading a certain file and we use the BT application to download, we actually make a request on all computers connected to the same bitstream network, if any of them contains Once we have the files we need, it will be sent to our computer.
How does the IPFS node work? Every computer running IPFS can choose its own files to provide services.
Every file that exists on IPFS has a unique hash to indicate that any small change will generate a new hash. The hash value is how users view content. The client queries the system for the hash, and any node with available content can provide it to the peer node. "This cluster provides a torrent-like experience where partners can provide content to each other.
This system can provide content to customers quickly and accurately. Also, because hashing is used, it is possible to check whether the content on both ends of the exchange is accurate. Because any difference will lead to different hashes.
What are the actual advantages of IPFS distribution?
Reliable and durable.
Distributed peer-to-peer content is not affected by "direct denial of service" style attacks. This type of attack is mainly aimed at closing websites Or the host server of the service. However, what you need is information from different nodes. If you want to attack, you must clarify each part and attack one by one to achieve the purpose.
You can use the previously viewed content offline.
In IPFS In the cache system, many regularly viewed content can be used offline. Of course, no dynamic content is up-to-date, but regardless of whether you are within Wi-Fi range, the static content resources you have viewed before are all within reach.
IPFS It's definitely not a polished or oiled machine. This is more like a fascinating network prototype that will change the way content is provided and stored on the entire Internet.