1. 《From ARPANET to Internet some interesting things》

1.1. 前言

  1. history repeats itself
  2. the future is a door,the past is the key. the farther backward you can look the farther forward you can see.
  3. 工程师和科学家伟大之辩:互联网上工程师更伟大
  4. 互联网是没有蓝图的(不变量 Technical Building Blocks,即便是design principles,如:无连接、尽力而为,也是后来总结出来的。

1.2. 网络发展脉络

1.2.1. cases 1:OGAS vs ARPAnet:(苏联 vs 美国)

image.png

  1. Paul Baran's Paper
  2. 去中心化和分布式
    1. 这是两个事情,去中心可能是多中心,而不是无中心
    2. 中国不是centralized也是decentralized,而美国是distributed
  3. RFC-1 Network Working Group Request for Comment:1

1.2.2. cases 2:NCP

  • Paul Baran
  • Rand Corp.:message block switching
  • Donald W. Davies:packet switching
  • Louis Pouzin,Vint Cerf, Alex McKenzie,Hunert Zimmemannm Donald Davies:International Networking Work Group,INWG

image.png

  • CCITT现在就是国际电联ITU
  • The Cyclades Architecture(Louis Pouzin)
  • 法国人的思路

1.2.3. cases 3:Split IP from TCP

Vint Cerf,Robert Kahn

  1. RFC 675:Specification of Internet Transmission Control Program December 1974 Version

  2. TCP早期是IP和TCP的耦合,从版本TCP 3分离出IP 3.

  3. Power of the Internet Architecture(Vint Cerf,principles)

    1. Not designed for any specific applications:just move packets
    2. Designed to run over any communication technology
    3. Permission innovation at the edges
    4. Design to scale
    5. Open to new protocols,new technologies,new applications
  4. INWG transport protocols:如果有自治域头部就没有跨域问题了

  5. 互联网和X.25及ATM比较

Internet X.25,ATM
网络类型 无连接分组交换 虚电路分组交换
复杂度 端系统 网络
网络服务 尽力而为 保证服务质量
互联互通 容易 困难

1.2.4. cases 4:TCP/IP 和 OSI

  1. 为什么TCP/IP战胜了OSI

  2. OSI is a beautiful dream,and TCP/IP is living it

  3. 网络体系结构

1.3. why?

  1. Why IPng?
    1. Problems with IPv4

      1. address is running out!地址耗尽
      2. routing table explosion!路由表爆炸
      3. David Clark:A Cloudy Crystal Ball
    2. ATM -- A really big elephant

      1. The last force on us -- us
    3. Why design IPv6?

    4. IP versions

      1. 愚人节 RFC 1606

image.png

  1. Evolution

image.png

  1. Open Internet Keys

    1. Voluntary adoption of technology
    2. bottom-up innovation
    3. Functional Interoperability
    4. Collaboration where needed
    5. Different Players at different layers
    6. Global generic and universal
    7. competition where possible
  2. why ARPAnet?

image.png

  1. 网络的十二条军规(April 1,1996,RFC1925)【但是这个日期发布的RFC,emmmm,恶搞RFC呀!】
  2. 亲自写code,运行network