Talk:IEEE 802.3

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Merge proposal[edit]

This article adds little over, and should be merged with, Ethernet, as should be obvious from the title of the next section. 18.26.0.18 05:09, 14 Jan 2005 (UTC)

I agree, so I've put mergeto and mergefrom templates into those pages. Guy Harris 02:33, 27 December 2005 (UTC)[reply]

I'm not too enthusiastic about that idea. The Ethernet article is already big, if not too big. The large standards listing table is much better placed here. I suggest we leave it the way it is and update it as necessary. The Power Over Ethernet stuff, OTOH, might be better placed in the main article. Wefa 05:33, 4 January 2006 (UTC)[reply]

So what about that table makes it apply to "802.3" rather than "Ethernet"? At this point, "Ethernet" and "802.3" are pretty much the same thing - as of 1997, "802.3" doesn't imply "length field, not type field", so that's no longer the distinction between "802.3" and "Ethernet". Should there be a split of the Ethernet page, with a top-level page, linking to subpages for items such as the list of PHY types, but with no distinction made between "802.3" and "Ethernet"? Guy Harris 07:09, 4 January 2006 (UTC)[reply]
I like the split of "technical stuff goes to Ethernet article, standardization history goes to 802.3 article". A history of the antics of the type/length field would belong in the 802.3 article, but really just confuses the "today facts" of the Ethernet article. --Alvestrand 08:00, 9 January 2006 (UTC)--Alvestrand 08:00, 9 January 2006 (UTC)[reply]
So the history section of the Ethernet page would just deal with Experimental Ethernet and DIX Ethernet and then put the rest of the history in the 802.3 article, with the 802.3 article handling the differences between DIX and 802.3 and the reconciliation in 802.3x-1997? Guy Harris 08:30, 9 January 2006 (UTC)[reply]
Not quite - if someone wants to chart the various technologies' market shares or years of dominance, that would (in my opinion) belong on the Ethernet page. But the DIX/802.3 war & reconciliation is (in my opinion) only interesting for poeple who enjoy standards history - it never significantly impacted the Real World. So that belongs on the 802.3 page. --Alvestrand 20:21, 9 January 2006 (UTC)[reply]
I oppose the merger. This page should talk about the standard more indepth. 132.205.45.148 17:35, 1 February 2006 (UTC)[reply]

Since there seems to be a consensus towards oppose, certainly on the Ethernet page (I also include myself in this opposition) I have removed the tag. The Ethernet article is too long to put a massive table of information in - there is no downside to keeping the two articles. QmunkE 18:41, 8 March 2006 (UTC)[reply]

Having reread the comments on Ethernet I realised this may actually only be slight majority view opposing merge, however since no action has been taken for over a month in terms of approving the merge I have removed the tag.

Power over Ethernet[edit]

My draft text for power over ethernet, so I don't lose it:

Power Over Ethernet technology describes any system that transmits electrical power, in addition to data, to remote devices over standard twisted-pair wire in an Ethernet network. This technology is useful for powering IP telephones, wireless LAN access points, webcams, hubs, and other appliances where it would be inconvenient or infeasible to supply power separately. The technology is comparable to POTS telephones, which also receive power and data (although analog) through the same cable. It works without requiring any modification to the existing Ethernet cabling infrastructure. Power over Ethernet is standardized in IEEE 802.3af. There are several ad-hoc standards for supplying power over Ethernet cable that predate IEEE 802.3af, however most future solutions will probably adhere to the IEEE standard.

See also[edit]

Rhobite 22:38, Jul 5, 2004 (UTC)


Ethernet layers[edit]

This article starts "IEEE 802.3 is a collection of IEEE standards defining the physical layer and transport layer of ... Ethernet".

Transport layer links to the Transport layer article.

Surely, it should be Data Link layer? (Maybe transport layer is used in a slightly different context here but if so then it shouldn't link to the layer 4 article).

- Leon

DIX Ethernet is thin?[edit]

Is DIX Ethernet v2.0 really thinnet, as the table here claims? v1.0 is thicknet. Guy Harris 06:51, 29 December 2005 (UTC)[reply]

No it is not. References in Ethernet#History make that clear. I have corrected the article. --Kvng (talk) 05:11, 7 May 2010 (UTC)[reply]

DIX vs II[edit]

This article says, "...DIX Ethernet, followed by Ethernet II," implying these are two different things. But Ethernet II framing says, "Ethernet II framing (also known as DIX Ethernet...," implying they are the same. I suggest this confusion be cleared up (in the article, not here on the talk page). Rees11 (talk) 16:59, 30 July 2009 (UTC)[reply]

DIX and Ethernet II are the same thing. I've added a reference to the Ethernet II article. I've removed the erroneous content from this article because it is covered correctly and in more detail in Ethernet#Ethernet frame types and the EtherType field. --Kvng (talk) 05:40, 7 May 2010 (UTC)[reply]
DIX and Ethernet II are, but Ethernet II and IEEE 802.3 are not. IEEE 802.3 framing differs a) the type of packet field is replaced by length (as is already noted) and b) the Ethernet II preamble is broken up into a shorter preamble and a start of frame delimiter. Finally I know for a fact that this machine accepts Ethernet II and 802.3 frames but only ever sends 802.3. It is not true to say that Ethernet II "is used on all forms of Ethernet by protocols in the Internet protocol suite." Crispmuncher (talk) 20:20, 26 October 2010 (UTC)[reply]
This is all discussed in Ethernet frame. I've removed the shards of meterial that were here. Please read Ethernet frame and make any corrections or comments there. Thanks. --Kvng (talk) 14:39, 27 October 2010 (UTC)[reply]

Working group[edit]

802.3 is probably better thought of as the name of the working group that produces Ethernet physical layer standards. This article should be changed to be more like 802.1. --Kvng (talk) 14:15, 20 July 2010 (UTC)[reply]

Conversion of bandwidths to megabytes/second[edit]

Do we really need these? Especially when they have been done using a simplistic divide-by-8 method rather than considering how data is encoding for transmission. For example, 10 and 100Mbit ethernet both use 8b10b coding: every byte is ten bits long on the wire meaning the theoretical maximums are 1.0 and 10.0 megabytes per second before we even consider framing overhead and so on. More generally, techniques such as bit stuffing make calculations impossible since the conversion factor depends on the data being sent. The IEEE do not quote these figures, they are highly ambiguous (coding and framing are internal to the standards and therefore externally visible bandwidth will always be lower), and they are of little practical use. I think they should simply be dropped. Crispmuncher (talk) 20:07, 26 October 2010 (UTC).[reply]

I agree. These are not needed. --Kvng (talk) 14:25, 27 October 2010 (UTC)[reply]

Layer muddle RFC[edit]

Is IEEE 802.3 a network layer technology? Tarian.liber claims that it is and prefers to discuss it at Template_talk:OSIstack#Network_.2F_Link_Layer_Muddle. Please visit there and help us sort this out. ~KvnG 03:32, 4 September 2013 (UTC)[reply]

Actually, he claims that IEEE 802.3 is a family of standards which defines a stack of protocols, going from the physical (OSI Layer 1) to the network (OSI Layer 3).
Tarian.liber (talk) 08:27, 4 September 2013 (UTC)[reply]

I'm having trouble seeing a clear improvement in these extensive anon edits. I have reverted. Let's discuss the purpose of these before restoring. ~Kvng (talk) 21:33, 22 October 2017 (UTC)[reply]

Regarding the change of "IEEE 802.3 standard" to "802.3-1983" specifically: this is plain wrong. There is no 802.3-1983. It should be 802.3-1985. This has been resolved by the #Date of First Version topic below. Rob Donnelly (talk) 18:16, 22 January 2024 (UTC)[reply]

Date of First Version[edit]

I'm having trouble finding a source that supports the 1983 date of the first version of the IEEE 802.3 standard. It seems that the first published version of the standard was actually 802.3-1985 (in 1985).

I'm basing this off of the following factors:

Factor 1: The IEEE page for 802.3-1985 does not specify that 802.3-1985 supersedes any other standard. Revisions include a section that specifies which version they supersede. See the IEEE page for 802.3-1998 for example.

Factor 2: The 802.3-1985 standard itself includes the following text (emphasis mine):

  • "Approved June 24, 1983 IEEE Standards Board"
  • "Approved December 21, 1984 American National Standards Institute"
  • "Copyright 1985 by The Institute of Electrical and Electronics Engineers, Inc"
  • "IEEE Std 802.3-1985 The first edition of the standard defines a 10 Mb/s baseband implementation of the Physical Layer using the CSMA/CD access method. It is anticipated that future editions of the standard may provide additional implementations of the physical layer to support different needs (for example, media, and data rates)."

The most important part here is that 802.3-1985 says that it is the first edition and while it was approved by the IEEE Standards Board in 1983, it wasn't published until 1985.

Factor 3: I am unable to find a 1983 version of the 802.3 standard.

Based on these factors, I recommend merging the entry for "IEEE 802.3 standard" and "802.3-1985" and setting the date to 1985. Rob Donnelly (talk) 23:43, 19 January 2024 (UTC)[reply]

I had been wondering about the discrepancies a while back as well and always wanted to research. From what I read from the 802.3 preface (any edition), the 1983 standard is 802.3-1985. It was approved by IEEE in 1983, by ANSI in December 1984, and published with the 1985 date. I think we should use the IEEE approval in the Date column to show that. --Zac67 (talk) 10:48, 20 January 2024 (UTC)[reply]
Agreed on using the IEEE approval date instead of the published date. I found several instances of IEEE using the approval date instead of the published date. Revisions of the standard include a "Historical participants" section which lists all 802.3 documents and their approval date(s).
Based on all of this, the changes I now propose are:
  • Remove the entry for 802.3-1985.
  • Rename the entry for "IEEE 802.3 standard" to "802.3-2015"
  • Prepend "First edition of the IEEE 802.3 standard." to the description for "802.3-2015"
  • Add a tooltip to the Date header cell that says "IEEE approval date"
Rob Donnelly (talk) 17:50, 22 January 2024 (UTC)[reply]
Actually, not all dates listed are IEEE approval dates (e.g. DIX). So instead of a tooltip I'm thinking adding clarification on the dates to the description. E.g. "First edition of the IEEE 802.3 standard. Approved by IEEE in 1983, approved by ANSI in 1984, and published in 1985. ..." Rob Donnelly (talk) 18:01, 22 January 2024 (UTC)[reply]
Changes made in https://en.wikipedia.org/w/index.php?title=IEEE_802.3&oldid=1197992851 Rob Donnelly (talk) 18:07, 22 January 2024 (UTC)[reply]
Excellent - I just added footnotes for non-IEEE standards and missing data from the 802.3 preface. --Zac67 (talk) 20:12, 22 January 2024 (UTC)[reply]