Talk:Internetworking

I'm sorry it's moving in this direction

"Internetworking" is more of a lower layers term than "Internet", and I'm afraid that moving the article here does not feel like a willingness to collaborate.

There are a number of redlinks in this article that could be going to existing articles on quite the same subject, but they are not using the correct article names. Preferably change the link, or at least pipe.

The bulk of the architecture article is about layering, which is not a strong Internet context. Don't take my word alone for this: An updated IETF architectural document, RFC3439, "Some Internet Architectural Guidelines and Philosophy" by Randy Bush and Dave Meyer (2002), has a section entitled: "Layering Considered Harmful": Emphasizing layering as the key driver of architecture is not a feature of the TCP/IP model, but rather of OSI. Much confusion comes from attempts to force OSI-like layering onto an architecture that minimizes their use. For higher-layer terminology, see RFC3466, "A Model for Content Internetworking (CDI)".

Internet architecture is much more onion-like than layer-cake, with principles such as end-to-end, local vs. remote, robustness, etc. Locality of networks issues also enter, be they link-local or scoped. I'm afraid I cannot, as an Editor, support something called a description of Internet architecture that is mostly about layering. Howard C. Berkowitz 05:47, 20 September 2009 (UTC)

You perhaps have misinterpreted my motivation. I moved the subgroup to the new name because two issues were being conflated: 1) what material belongs in the article called "Internet", and 2) the formation of the "Internet" (now "Internetworking") subgroup. Keeping these two issues separate clarifies the discussion. I will study your comments and respond to them in the next few days. Dan Nessett 06:02, 20 September 2009 (UTC)

Explanation of history

This page was first edited as a new version of Internet and then copied to "Internetworking".

I have replaced the copy by the original page (including its history). The corresponding talk page is at Talk:Internet/Archive 1.

Peter Schmitt 22:22, 23 September 2009 (UTC)

---

Stepping on toes

While not wanting to step on toes here it can't sit as an article without having some metadata. I just added metadata so it can exist in article space. Hopefully we can develop in in a useful direction or merge it in to other articles Chris Day 04:02, 9 December 2009 (UTC)

"Internetwotrking" is certainly not a good title. The problem is to find a good one. If no other suggestion is made, I would call it "The architecture of the i(I?)nternet". Peter Schmitt 19:56, 9 December 2009 (UTC)

Actually, internetworking has a meaning of which the Internet is a subset: the interoperable interconnection of networks of different technology or under different administration. Dave and I are starting to brainstorm about it, and all the Computers Workgroup Editors should get involved. There is an article, Internet Protocol Suite, which is somewhat architectural. For the Internet case, Routing policy, Routing Protocol Specification Language and Border Gateway Protocol deal with some of the issues of separately administered aspects.

There are other aspects dealing both with the applications, and also social and economic features of the Internet (e.g., value of networks, telecommunications provider economics Howard C. Berkowitz 20:07, 9 December 2009 (UTC)

This does not answer the question: Where to put this article? "The architecture of the internet" would not conflict with a technical article on "Internet architecture", and could be the place for a rather popular overview over its structure. Peter Schmitt 20:40, 9 December 2009 (UTC)

There is very great overlap over this article and Internet. I need to print them both and get a cleaned-up article, then delete one. Could you give me an idea what level would go into "architecture of the Internet" vs. "internet architecture?" I'm guessing that autonomous system and the end-to-end principle need to be at the introductory level, but, for example, per-hop behavior does not. Howard C. Berkowitz 21:08, 9 December 2009 (UTC)