This article will talk about the HTTP protocol, developments with the new version 2.0 and the differences between the new version 1.X and 2.0.
Also, we will echo that last December 3 the famous service CloudFlare activated the option of using HTTP2 in their communications.
But first, to better understand the reason for existence of this protocol, we will give a brief explanation of HTTP.
Where HTTP is?
After the birth of the Internet, was the need to establish a simple communication protocol. Scientists at CERN (thanks to the father and creator “Tim”) then created a protocol combining several technologies:
- Hypertext documents
- Information retrieval based on the net
- The language Standard Generalized Markup Language (SGML)
These technologies, well established, are encompassed in the HTTP protocol, giving a common way to communicate Internet distributed systems.
Now we will explain briefly the HTTP protocol.
Read also: Essential tools when creating a successful Blog
HTTP
This protocol is stateless, which means that each communication between client and server maintains no information. As usual maintain certain status information; they have developed strategies to maintain a state outside the HTTP protocol (Cookies and Sessions).
It also defines a nomenclature for “call” resources located on the machine that makes server, as follows:
http: // [domain or IP]: [port] / [resource] [attribute]?
HTTP 2.0
Last February, the version 2.0 was published after more than 15 years without updating.
The Version 2 is compatible with running versions 1.1 and 1.0. Maintaining the same status codes (1XX, 2XX, 3XX, 4XX, 5XX), methods (GET, POST, PUT, HEAD, etc.) and the structure of the call chain. The differences are listed below:
- Headers in binary format instead of text.
- Compressed headers, so beside the binary format compress the amount of information transmitted per request (through is reduced HPACK).
- Server push technology: Allows you to upload files (CSS, JS, images, etc) from the server to the client without the latter’s request. HTTP 1.X when asked for a page, the HTML first, then all the CSS files, JS and images linked by downloading it. These files could turn reference other and as a result were discharged last. With HTTP server push technology 2 can send all the files needed for the page is loaded from the start, without having to wait for the browser interprets HTML.
- Multiplexing: HTTP 1.x has a problem called “Head-of-line blocking” where only one application is used by a connection, leading to congestion. Multiplexing addresses these problems by allowing multiple requests can be resolved at once. This allows the clients only need to use a link to load a page.
In short, we can have faster loading for all page resources.
Read also: The success of a website starts with good analysis
In the end
To operate the HTTP 2 protocol, it is necessary that both the client and the server has support. In the event that some of them do not have support in the negotiation protocol would be used previous versions.
On the client environment, updated modern browsers are supported (Chrome, Firefox, etc). Some browsers that do not support it will have soon.
In the server environment, there are many implementations official environments (Apache, Dart, Jetti, etc …) that you have available. Moreover, the servers that make intermediary as CloudFire increasingly offered support for HTTP 2.
Leave a Reply