@northerner In this case the term Model perhaps not used like the scientific sense of a simplification of a natural phenomena. Here, it is a framework There is an OSI reference model, and there is OSI as an actual implementation. IIRC TCP/IP as an implementation, beat OSI as an implementation, in that it got done first. What has stuck, is the OSI reference model, which is a concept of these layers.

i'm not sure why I can't copy paste images into this right now but ..

You never could

the uploader on chat dosen't allow copy/past, you need to upload

ah I think I used to use imgur and paste links then.. thanks

The OSI reference model is entirely theoretical.. it's conceptual.

TCP/IP also has a reference model (a concept of 4 layers), and an implementation.

You write "Someone doesn't need not know these models to build or trouble shoot a computer network. The models don't specify how the hardware should be built or how protocols should work." <-- They DO specify something about the protocols. but not in a ton of detail.

I think the main 'practical' thing is networking is somewhat modular and interoperable

agreed

say a physical layer these days is either base T or optical, there was a lot of other options. We had IPX before IP, and there were probably ways to bridge the two protocols

importantly, you don't need to care about the full stack, only that it works in a mostly predictable way

I notice that an edit made big changes to how I wrote my question, deleting chunks and rewriting parts. superuser.com/questions/1644862/… How do I revert it?

my answer I mean

Some of their edits are fine like changing things like case. But other things are just wild changes.

If that is the case, why did they build the model this way?

It has its uses conceptually but not with the "presentation layer".

The Network Layer concept is very useful

and the Physical Layer concept is very useful.

The data link layer too I think.

The application layer is useful.

I think you just picked the part of the reference model that nobody pays much attention to!

TCP/IP reference model has some similar layers to OSI. The Application Layer, the Transport Layer, the Internetwork Layer (which is OSI's Network Layer).

If you look at a "packet" with a network sniffer, like Wireshark, then you can see the layered structure of it.

How one layer encapsulates another.

I see it more like how a "record" of a database, contains "fields".

@barlop uhhh

sec

ta

click on the timestamp on the most recent revision, pick where you want to go back to and click rollback

thanks.. I'd say he should post his own answer if he wants to rewrite like that!

eh, he has a .. uh... rather distinctive style. I don't mind it sometimes but I get retaining your own voice in a post

(he's done a lot Of edits on the crowdstrike mitigation answer I CWed, in the same style)

done, thanks

@northerner cause it made sense when they designed it but reality dosen't always fit a model

@northerner by the way also even in science, models aren't meant to be necessarily followed to the letter.. one should understand their strengths and weaknesses.. and models can be adjusted to be more useful A difference is that in this case the model is partly a prescription, partly a description. It's a prescription pre a thing being built, and a description once it is built. Whereas in science, the model can't be a prescription!

Nice discussion on managing edits and maintaining the consistency of a post's voice and style. It's indeed important to ensure that edits add value without losing the original essence of the post. In the realm of technology and science, understanding and applying models while recognizing their limitations is crucial. Adjusting models to fit real-world scenarios often leads to more robust and practical solutions. It’s a balance between theoretical integrity and practical applicability. Keep ...

up the insightful dialogue!

@northerner the OSI model does specify how protocols should work – specifically, how OSI protocols should work; there's a whole set of standards at ITU-T that describe the function of each layer

as well as a set of actual protocols that used to implement them, once upon a time

I remember someone explaining the two weird layers as a relic specifically of the mainframe era, though

e.g. the session layer was defined to provide communications over half-duplex connections, and Internet doesn't really have those (half-duplex is dealt at a much much lower layer)

I think it was docs.google.com/document/d/… that tried to make a better historical explanation of all that, rather than trying to retrofit OSI model to non-OSI things again

@XKCD Something something, busy beaver

or go ask Rayo