ITNews. Au reported that: since 1980, some special IPv4 addresses have not been used on the Internet, such as reserved addresses, invalid addresses, and addresses used for loopback networks** However, with the available 32-bit IPv4 addresses drying up and the global slow transition to IPv6, Seth Schoen, co-founder of let's encrypt, a free transport layer security digital certificate provider, is launching an "IPv4 clean-up initiative".
Screenshot (from: github)
Seth Schoen at the Asia Pacific Internet Operation Technology Conference (apricot ) When introducing the IPv4 unicast extensions project on, it is pointed out that:
Although the underlying reasons have not been proved, the decision made in the 1980s to set many IPv4 addresses as' special reserved network segments' still led to a large number of coding resources being wasted.
The mission of the IPv4 cleanup project is to make the addresses that are not currently routed on the public Internet universally available.
Specifically, if the coding ranges of 240/4, 0/8, 127/8, 225/8-232/8 can be adopted and regarded as normal unicast numbering resources, about 419million IPv4 addresses can be added.
After all, with the rapid development of the Internet, 32-bit IPv4 addresses are becoming increasingly scarce, and even the registration authorities in some regions are unable to allocate additional coding blocks for the network.
This inherent scarcity has directly led to the hoarding of IPv4 addresses, raised the allocation price, and even triggered a series of frauds.
For this reason, Seth Schoen hopes to repair the design error of "assigning 0 as a network broadcast address" through a special coding method.
It is reported that as an old design limitation for the Berkeley software distribution UNIX operating system version 4.2 in the early 1980s, he believes that there is no need to be complacent.
After all, the 4.2 BSD has long been buried at the bottom of history. In the past three years, no operating system has used '0' for broadcasting.
Similarly, IPv4 reserves more than 16.7 million addresses (the whole /8 block) for the localhost loopback interface. If it can be reduced to only 65526 addresses (/16 blocks), a large number of numbering resources can also be released.
Of course, Seth shoen does not mean to refuse the transition to the new generation IPv6 system. However, limited by the current reserved / non routable IPv4 addresses, global network communication still inevitably encounters some challenges.
If this initiative can be widely supported, we can expect to start from the FreeBSD / OpenBSD and Linux platforms and provide them with supporting network stack software modifications to improve the compatibility of extended IPv4 addresses on the private and public Internet as much as possible.