Talk:Network topology

Readability of the Article

I recently referred my students to Wikipedia for research purposes, however they found the article difficult to read and understand. The text of the article was subjected to a readability text which indicated that the article was of an undergraduate level and therefore not suitable for my students who are aged 16-19. The main problem was that the sentence structure was too long. Would anyone be willing to rewrite the article in more simple English to make it suitable for a wider audience? User:Sarahhcfe 19:58, 23 October 2006

There's a special version of wikipedia written for those whose grasp of english is shaky. The articles are much easier to read, relying on simple vocabulary and grammatical constructs. Go to "simple.wiki.x.io". yandman 17:41, 12 November 2006 (UTC)

I looked on simple.wiki.x.io, they have two lines on networks and nothing about the topology. It would have been useless to anyone except someone who has just started using computers and has no idea about it, it does have a link to electron microscopes though

I thought of improving the language used in this article, and very soon realised that a complete re-write is required. I will do as much as possible, but can devote only a little time here. Need someone else to help out.

On another note, I am making a few changes to the first paragraph, if anyone is offended by it, please make the changes as you seem best fit, and leave a note about the same on my talk page.

User_talk:Darnir_redhat 14:02, 14 March 2010 (UTC)

We've apparently made progress since 2006. Current readability score is 7th grade. ~Kvng (talk) 18:36, 13 August 2021 (UTC)

Confusion concerning terminology...

I have rewritten this article to the best of my ability given my reference sources but am still confused concerning some aspects of the terminology used in this subject, especially concerning the correct usage of the term 'physical'. In some instances the term seems to be used to describe the actual arrangement of the layout of the cabling and the nodes in a network and in other instances it seems to be referring to the physical layer of the OSI model.

For example, the star network, which could exhibit, from a signal standpoint, any number of topologies while physically it is a star with reference to the central node being linked to the peripheral nodes in a hub and spoke fashion with cables. This topology could be used to implement the 802.4 Token Bus network if I am not mistaken but in this implementation (at the PHY layer) it is a physical bus and at the logical layer it is a logical ring which results in three topologies (i.e., physical star [cabling level], physical bus [PHY layer], logical ring) instead of the two referred to in most references (i.e., physical bus / logical ring). I am confused as to how to describe this network in network topology terms. My best attempt would be to describe it as a physical star, signal bus, and logical ring but can find no reference to this type of usage of the terminology.

For another example, the same star network could be implemented using a switch as the hub or central node and then it would exhibit a mesh (hybrid ?) topology as far as the signal is concerned and could be used to implement some type of a logical ring topology (given the correct software in the nodes), in which case I would describe this network as a physical star, signal mesh, and logical ring.

In my examples, the term 'signal' seems to be replaced by 'physical' in many texts and the fact that the physical and signal topologies (my terminology) are different seems to be entirely overlooked or ignored.

If someone could clarify this for me (with some references that are easily accessible over the internet if possible) I would be happy to incorporate this new (for me) understanding into this article.

Thanks! --mlewis000 19:35, 12 December 2006 (UTC)

We could clarify this by using wiring topology or physical topology to the physical layout and pyhsical layer topology for behavior at the physical layer. ~Kvng (talk) 18:36, 13 August 2021 (UTC)

Basic Topology Types

I feel that the section Basic Topology types should be removed or clubbed with Classification of Network Topologies. It serves no purpose on its own, alone out there. Any suggestions? Darshit 14:36, 14 March 2010 (UTC) —Preceding unsigned comment added by Darnir redhat (talk • contribs)

section is now called Classification. ~Kvng (talk) 18:36, 13 August 2021 (UTC)

Topology content in Computer network

If this article is to serve as the {{Main article}} on topolgy for Computer network, we need to merge most of the material from Computer_network#Network_topology into this article. ~Kvng (talk) 15:52, 1 September 2016 (UTC)

I have moved a bunch of content from Computer network to here. This seems to make Computer network more manageable. Further improvements or discussion is encouraged. ~Kvng (talk) 17:01, 24 October 2017 (UTC)

GliderMaven reverted these changes at Computer network with comment, "article no longer covers the topic." Do we just need a more comprehensive WP:SUMMARY of topologies at Computer network or is there some deeper objection to this reorganization? ~Kvng (talk) 22:07, 2 November 2017 (UTC)

Basically, it already was a summary of Network Topologies at computer network, and by removing almost everything about network topologies it meant that the article was no longer properly covering the topic of computer networks. For example you completely and utterly removed the definition of 'network link' and then continued to use it in the article. Like what is a network link? You don't say. What is even an example of a network link? You also don't say. But I'm not simply complaining about that, that's just one symptom of the problem. Another is that you also removed all the types of network nodes that people would be familiar with. Without your edit, the article is nearly B-class. With it, it was squarely C-class, and there is clearly much missing.GliderMaven (talk) 00:00, 3 November 2017 (UTC)

@GliderMaven: The material I removed is pretty much a copy of what's in Network topology. If you think it is essential that all this material remain in Computer network, perhaps we don't need Network topology at all. Or perhaps we should put all this material in a template so it can appear both places and be maintained in one. Or perhaps we can figure out a way to make WP:SUMMARY work in this case so that we don't have so much repetition. That's what I was going for. Would you be willing to try again and add back summaries of the stuff you identified as crucial (network link, node types, ...). ~Kvng (talk) 14:39, 5 November 2017 (UTC)

I had a look at trying to summarise it down in computer network, but it seemed fairly difficult. I'm leaning more towards not needing Network topology, unless more material can be found for it. Computer Network seems pretty gutted without it.GliderMaven (talk) 16:31, 5 November 2017 (UTC)

I think I am at this point desiring maintaining computer network and network topology as distinct articles. As I see it there are two main practices that are concerned with electronic communication typologies, where I expect many editors see only one. The one obvious practice is the enterprise level where technologies are oriented with the Internet. Because there is internetworking at this level, it is appropriate to discuss the different topologies that are used in that practice -- I think this is the appropriate level for computer network. But the practice where I am more involved is at the fieldbus (the general class not the specific protocol) level and lower. Largely, the same topologies are discussed, but in the sense of defining the protocol (CAN vs Ethernet vs ARINC 429). Here we can discuss dual, lockstep redundant star hubs for either Ethernet or CAN PHYs, but no Internet technology at all. Usually, a low-level protocol has one topology (or at most two). Understanding the protocol depends in part on understanding how its topology (different from other protocols) enabled or restricts its capabilities. For this purpose, comparing topologies, I would prefer to refer to an article that only discussed or listed various topologies, rather than refer to an article that (IMO) is really about internetworking, in which topology is but one consideration in the multi-layer interconnect model. Yes, the topic of "computer network" is technically inclusive of two microcontrollers communicating over SPI (and don't think such applications get to escape the questions of Presentation and Session even if these are usually ignored), but I do not think that is what past editors of computer network had in mind.

The present sections of computer network, Network topology (should be renamed?) and Communication protocols, make it clear that that article services internetworking, and so is of little service to those working with "non-internetworking", perhaps "peripheral", networks. IveGoneAway (talk) 15:52, 14 December 2017 (UTC) 15:59, 14 December 2017 (UTC) 16:13, 14 December 2017 (UTC)

There are individual articles on each of the topologies (Bus network, Star network...) so you can link to these if you want to point a reader to the topology relevant to a particular network without drowning them in Computer network. ~Kvng (talk) 14:41, 17 December 2017 (UTC)

Sure, and when I want to link to the broad topic/network parameter of topology, I likewise also do not want to drown the reader in Computer network (most, but no all, of those specific topology articles link back to Network topology, there is presently some inconsistency).

I started long enough ago to have just caught the tail end of experiencing equipment that used digital communication protocols (e.g., RS-232) that might have only discrete components on either end (no computers!). I have wondered if early ARINC 429 was used on some computerless networks. Probably, there is no point in addressing that in Network topologies. But neither would I want to burden the Computer network article with non-internetworking communication in the vehicle or microcontroller peripheral communication space. IveGoneAway (talk) 15:55, 18 December 2017 (UTC) 18:09, 18 December 2017 (UTC)

I think were we stand is that GliderMaven objects to removing or thinning network topology content from Computer network and IveGoneAway feels we need to retain Network topology. I have tried to solicit additional opinions but I don't think that has been successful. It seems like there is not consensus to make major reader-facing organizational changes to this material. If what we really want is the current organization where we basically have a copy of Network topology included in Computer network, it would make sense to put this material in a template so that the content only needs to be maintained in one place. ~Kvng (talk) 17:15, 21 December 2017 (UTC)

Re: Template-ing (shared) Network Topology (content): Oppose

I think templating will ossify the present problems indefinitely, namely the present computer network is really about computer internetworking and presently provides no real room for discussing the many non-internetworking protocols I have worked with. The present sections are duplicates and they should not be. Yes, there is a LOT of overlap, and it is not like we never discuss repeaters, switches, hubs, etc. in respect to CAN or RS-232. But the present computer network is not, er, neutral? Even the Links and Node sections that are duplicated between the two articles are particularly selective of Internet technology examples.

My sense of the issue is a matter of unintentional systemic bias; I conjecture that the population of Wikipedia editors have much more exposure to Internet technologies than they do to the Automotive, Aviation, Industrial, and peripheral non-internet technologies (IoT not withstanding), so that's what they write about. Sure, one of my present projects has an Ethernet switch, but it is under ModbusTP, SCPI, and some proprietary protocol. There are billions of CAN networks out there and the various high-layer protocols on top of them are interesting (and critical to public safety), but I speculate whether editors of computer network would classify them as "computer networks" (No criticism intended).

Sure, there should be nodes and links in both contexts, but the technologies in computer network should really be more about concrete technologies of computer networks in the Internet context since IMO that is where that article is already, while the nodes, links, and topologies in network topology should be broader and more abstract than they are now. If I was available to fix it, I would see as the network topology as the Broad Topic relative to more specific sections in computer network. E.g., Nodes and Links there are too specific now, the concept of node is much broader than that given under Network Interfaces; NIC and MAC address, as described, are alien to many other network standards. IveGoneAway (talk) 02:52, 22 December 2017 (UTC)

And, network topology should be the appropriate content for what Physical layer links to as Physical network topology. Considering the technologies listed under Physical layer, I think it would not work to expand the section in computer network for non-internet-related PHYs. IveGoneAway (talk) 03:16, 22 December 2017 (UTC)

Network topology is a heading under Template:Telecommunications

As broad of a topic as it is, computer network is a subtopic to Telecommunications; likewise the Topology content in computer network, is a subtopic to network topology.

IMO, Network topology content and Topology content in computer network should diverge, not converge, it never should have converged to the present state in the first place.

Nodes and Links are the two components of a network, but the present Nodes and Links duplicated between computer network and network topology is too specific to computer network to be valid for broad topic network topology. Nodes and Links in network topology should be more definition and theory of network topologies, than limited to examples used commonly used in networking computers. IveGoneAway (talk) 18:48, 23 December 2017 (UTC)

@IveGoneAway:, you've made hand-wavy proposals for reorganization but have said you're unable to help do the work. This is not actionable discussion. I hear that you object to my proposal to cover topologies in a template. If that is heeded, the only remaining option is to leave things as they are. I don't see anyone making an argument that things are fine as they are. ~Kvng (talk) 15:57, 26 December 2017 (UTC)

@Kvng: It has taken a too long to catch up on what your intent and original issue was in September 2016. I also see that you have put several years into these articles. In that same time, I was using some of the same articles but in a different practice. But I was not working with IT, rather with what mostly would fall under a broad definition of fieldbus (Level 1 to 2 plus incomplete Level 7). Viewing Network topology purely as an attribute of Level 1, I would have never missed discussing routers and firewalls. Physical Topology is a critical difference between PHYs, to most of which, switches, routers, gateways, and firewalls are alien. Yes, networks have nodes, and on some types of network PHYs, some nodes can be switches, routers, gateways, but many non-internetworking network PHYs have none of those.

Some specific places where I think the content you moved from Computer networks to Network topology was appropriate for the former but not for the latter:

• "...basic system building blocks, such as network interface controllers (NICs), repeaters, hubs, bridges, switches, routers, modems, and firewalls."' True, these are computer internetwork building blocks, but most network standards don't have much more than analogs to NICs. The following paragraphs are fine for the Computer networks context, but for the Physical Network topology context one of two things should happen, 1) the paragraphs should be deleted since they are already covered in Computer networks or 2) the paragraphs should be expanded to cover non-Ethernet and non-inter-networking examples (bloat?). Either way the links in the lead sentence should be checked for narrowness; e.g., Ethernet hub isn't the only kind of networking hub.
• The CAN silicon on a microcontroller attached to a CAN transceiver is clearly a network Node (two, technically), but would IT recognize that as a network interface controller?
• "A repeater with multiple ports is known as an Ethernet hub." Ethernet hub isn't the only kind of networking hub.
• "Hubs and repeaters in LANs have been mostly obsoleted by modern switches." I would think that statement is probably fine in the Computer networks context, but for the Physical Network topology context I would consider tagging it: in which networks have hubs been replaced by switches? I realize the qualifier is "LAN", and USB 3.0 is not a LAN PHY, but it is a PHY that has a topology that has hubs that have not been replaced by modern switches and has this article as a main topic (through degrees of separation).
• "Bridges": IMO, a topology point about Bridges is that not mentioned is that Bridges can be the connection between PHYs with different topologies. The example on my desk is a USB Hub to USB/RS-232 bridge to RS-232/ARINC 429 bridge (bus to star to point-to-point to bus [logical star]).
• "Routers" and Gateway: Routers are particularly associated with Internet, but that does not prevent the concept being adopted by non-Internet internetworks, e.g., AUTOSAR. And I would have expected gateway to be on the list, but then, I am not IT? So, for Network topology, I would really expect conversation to focus on the Level 1-2 connections; all the level 3-7 have little to do with definitions of PHY topology, they are really about connecting topologies. AFDX has an Ethernet PHY and switches, but all routing is virtual (static); above that, it works like ARINC 429.
• Links: IMO, the listing of media is appropriate to Transmission medium; but, if it is important to have a list of media that is specific to computer (inter)networking, then such a list is be appropriate for Computer networks.
• As I understand it, the attribute of topology was considered separately from the medium of the transmission, one topology can be realized in multiple media, one media can be used in multiple topologies (with restrictions).
• The present link media list does not include media for 1-Wire, SPI, I²C, etc, none of which are specifically twisted pair or coaxial, although Airbus literally has stretched the point.

IveGoneAway (talk) 02:24, 27 December 2017 (UTC) 14:57, 3 January 2018 (UTC)

@IveGoneAway and GliderMaven: Things have been quiet here for over a year. We still have a lot of material duplicated in the the two articles. Are we happy with this situation? Has anyone come up with any new ideas? Should I ask for more input at WP:NETWORK? ~Kvng (talk) 15:00, 10 March 2019 (UTC)

@Kvng: I saw your post above days ago and have been looking for time to steal to respond, so I am stealing some time now. I thought I got less hand wavy above in my above comments and in my 12 February 2018 edit. At present, you have not responded whether you consider network topology a subject only of computer internetworking, rather than a topic inclusive of fieldbusses and interchip networks, for example, but to be fair, no one else has either. That seems a consensus for the status quo until I or someone else work more PHYs into Network topology. To restate; for my purposes, I would be pleased with Network Topology being further broadened, while the topology section of Computer network of computer network narrowed (but on a superficial examination, the topology section of Computer network seems sufficiently scoped to Computer internetworking). IveGoneAway (talk) 13:32, 20 March 2019 (UTC)

Resolved

Computer network § Network topology has been thinned considerably since the version current at the time of this dicussion. ~Kvng (talk) 18:42, 13 August 2021 (UTC)