This article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.ComputingWikipedia:WikiProject ComputingTemplate:WikiProject ComputingComputing articles
It is my understanding that many of the Application Layer Protocols actually perform the services of the Application, Presentation and Session layers of the OSI model. This is why the DoD model lumps the layers together. Someone correct me if I'm wrong. Rsduhamel23:32, 18 September 2006 (UTC)[reply]
Sounds correct to me, layers 5,6 and 7 are somewhat blurred. For example SIP (Session Initiation Protocol) should arguably be represented at layer 5, not at layer 7 as it is here.
jsut wanted to make note of a factual error....there aren't seventeen layers in the OSI model...only seven. Just thought someone might want to correct that.
Following the link "application protocol" in the "Lightweight Directory Access Protocol" article redirected me to "Application layer". Am I to take that as a clear statement that "application protocol" is always synonymous with "application layer protocol"? If so, such a clear statement should be made explicitly. If not (for example, in reference to a protocol used by an application having nothing to do with communications), the redirect is simply wrong. Either way, simple redirection without explanation seems like the wrong way to go. Unfree13:13, 1 December 2007 (UTC)[reply]
7th layer OSI (application) vs. 5th TCP/IP (application)[edit]
Layer 7 in OSI is not layer 5 in TCP/IP!! Please, someone correct this.
One example: RDP (presentation layer 6 OSI / application layer 5 TCP/IP).
(also, I restored good ver. of page, just FYI)
G.
The article does not discuss Layer 7 protocols, but Application Layer protocols in general. Since the article is not model-specific to OSI or TCP/IP, most of these should be restored. Separate articles or separation into sections may be appropriate however. Kbrose (talk) 16:29, 29 January 2009 (UTC)[reply]
_sigh_ That's why I put it on the discussion page a week ago... Still DHCP and BOOTP are not on the application layer of either model, and still H.323 is a set of protocols of which only some are on AL.
Now, what can be done? split into separate entities like
Application layer protocols convert user info into data. DHCP and BOOTP do not do that, they are on the network layer of OSI and TCP/IP models. See e.g. Forouzan, Behrouz A (2007). Data Communications and Networking. McGraw-Hill. pp. 612–620. ISBN007-125442-0. --Pgallert (talk) 08:40, 3 February 2009 (UTC)[reply]
Unfortunately, your reference is wrong. This text's discussions has other defects as well and is not a reliable source (however, it is ok to quote 'wrong' sources and state the controversial nature of a topic) Nowhere is the application layer (at least in TCP/IP) defined as converting specific types of info into data. What do those terms mean anyways: "user info" and "data"?. The only distinction that OSI makes is the separation of data w/r/t to the functions of the presentation and session layers. But obviously this has caused much confusion already, because in many instances this distinction is very murky and controversial. The extensions of the original OSI model do consider a network management section in the network layer, but TCP/IP has nothing like that. There is nothing special about providing boot information to a system, it's an application and the delivered info is 'user data' like any other application. Kbrose (talk) 15:03, 3 February 2009 (UTC)[reply]
For the distinction of information and data, see the respective articles on WP. The difference is the abstraction levels. For a definition of the Application Layer in TCP/IP, see RfC1123, for the other layers RfC1122. And no, DHCP does not deliver user data. There is a difference between management protocols and application protocols. The management protocols do not easily fall into any layer, but if they must they are commonly sorted into the layer they manage, except by you. Have fun. The payload defines layering, not the transport mechanism.
It would of course have been nice had you come forward with any source to back your view of the matter instead of repeatedly reverting me because you think I'm wrong. Every text book has defects, and the one by Forouzan is (in my personal, utterly inferior view) not the worst of them all. But never mind, I will now step aside from this discussion as it seems I must first correct your misconceptions before I can correct the article itself, which is neither my intention nor apparently in my power. While I do not match his acumen, I now understand what frustrated Howard Berkowitz. --Pgallert (talk) 09:48, 4 February 2009 (UTC)[reply]
yes, you do have the source of definition of the TCP/IP layers correct, which are already quoted in almost all relevant articles here and which support this model. BOOTP is specifically mentioned and it is not placed in the Internet Layer, but grouped as a support protocol discussed in the same context as application protocols, otherwise it would have been in RFC 1122, and not in RFC 1123. There is no intent evident in any of the RFCs to support the concept of management layers or sublayers. DHCP and BOOTP do not manage any layer in particular, in fact they (DHCP in particular) go far beyond just providing IP addresses. TCP/IP defines the layers not by data or payload type, but by operating scope. Kbrose (talk) 15:58, 4 February 2009 (UTC)[reply]
Application Layer is working on the TCP. Its works to prepear data to show on human reading form when recving it presentaion Layer and change it on when give presantation Layer. — Preceding unsigned comment added by 180.151.3.178 (talk) 08:16, 2 June 2011 (UTC)[reply]
Why are the layers in capital letters? I know Standards Like to Cpaitalize Every Word But That Is Not Wikipedia Style. This certainly sounds like a general concept, since there are so many of them, and even different concepts in the ISO vs. TCP models, for example. W Nowicki (talk) 21:49, 27 September 2011 (UTC)[reply]