Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The OSI Deprogrammer

docs.google.com

Post reply on HN
Print view View original

51–60 of 95 posts

Re: The OSI Deprogrammer

#51

Post by robertgraham »

kiitos wrote: post #31
lmm wrote: post #27

Earlier quoted context omitted.

It's far from ubiquitous though. It breaks down as soon as you enter the real world. It's simultaneously too vague to be useful for talking about specifics and too inaccurate to be useful for talking about generalities.

It's absolutely ubiquitous as a model. It's taught in every CS curriculum in every university in the world. It also doesn't break down in the real world, unless you're being extremely pedantic. Many bits of software don't map cleanly to the defined layers, but we can still talk about the distinction between the network layer vs. the session layer, for example, and get value out of that discussion.

You can't. Nobody knows what the session layer does. Most falsely believe that layer handles sessions.

Re: The OSI Deprogrammer

#52

Post by kiitos »

corethree wrote: post #35
kiitos wrote: post #32

Earlier quoted context omitted.

OSI is just a model. Like all models, it's wrong. But its ubiquity, and the fact that it's survived the test of time, clearly signal that it is a useful model. And that's as good as models get.

Not all models are truly wrong. But I get your point. I think though the author isn't just arguing that the OSI is like your typical "wrong" model. I think he's saying it's a shitty model. It's crap. It's like using Hippocrates four humors to model human health. Yeah that model is wrong just like every model. But unlike every model it's a garbage model. Religion has ALSO survived the test of time, but Jesus Christ co…

Look, just tautologically, if it was so shitty, we wouldn't be talking about it, because nobody would be using it.

You're free to use a definition of "shitty" that doesn't take into account usage or longevity, but that definition is not really relevant except at the theory level.

Re: The OSI Deprogrammer

#53

Post by robertgraham »

tptacek wrote: post #6
NBJack wrote: post #4

This is wonderful material. But admittedly, I am a bit lazy on reading to the part I want: what is a good model, or set of models, I should re-educate myself with to correct what I was taught?

Ethernet, IPv4, TCP, and HTTP; what functions do each provide, how is each coupled to its neighbor, and what are the viable alternatives in the design spaces of each? Part of the OSI layering model was I think meant to get people's brains wrapped around the dynamics of packet switching networks, which were once a novelty and are now just part of the air we breathe. Then, read: https://web.mit.edu/Saltzer/www/publicat…

It's wrong to think of them as different functions of the same network. Instead, they are differnet networks.

Ethernet and the Internet both provide the function of forwarding packets through a network to the target destination address. The MAC protocol and IP protocol provide the same theoretical function.

Where they differ is that Ethernet was designed to a be a local network, whereas the Internet is designed to be an internetwork. All it demand from local networks is that they get packets from one hop to the other. Ethernet does this for the Internet, but so do carrier pigeons.

The same is true for HTTP. Instead of thinking of it as a component of the network, think of it as something that rides independently on top of the network. In a hypothetical future whe we've replaced the Internet with some other technology, the web would still function.

Re: The OSI Deprogrammer

#54

Post by robertgraham »

derefr wrote: post #22
tptacek wrote: post #12

Earlier quoted context omitted.

I sort of agree that "OSI" is really weakening his argument, and that he could take clearer aim at the classical Internet layering model (network, transport, application) --- even that simple layering is obsolete now. But it'd still be a straw man! We even run DNS over HTTP at this point.

I think that the "OSI" he's speaking about isn't an engineering tool, a thing people would use to construct systems; rather, this "OSI" is a way of modelling systems that exist — systems that, of course, were not implemented by people thinking about OSI. It's an analytical modelling tool, a thing applied to a system retrospectively in an attempt to understand it, break it down along abstract rather than concrete line…

Yea, it's not about engineers constructing systems. I mean, engineers do frequently pretend their creations fit the OSI model, but they work backwards to make it appear to conform to orthodoxy.

The issue is about education. People teach the model, or some variation of it. It teaches misconceptsion, such as how Ethernet and the Internet are integrated in a single network stack rather than being independent networks. It leads to professionals in IT and cybersecurity who continue to hold this misconception.

Re: The OSI Deprogrammer

#55

Post by hughesjj »

culopatin wrote: post #2

This is hilarious. I happen to be preparing for a test centered around the OSI model and I don’t get why it’s set this way. Every few paragraphs the text says something like “remember that nowadays things a bit fuzzier and don’t fit within one layer” well then why are we even using this model?

You might want to read this for a bit of a primer

https://apenwarr.ca/log/20170810

> The point of all this is that ethernet and IP were getting further and further intertwined. They're nowadays almost inseparable. It's hard to imagine a network interface (except ppp0) without a 48-bit MAC address, and it's hard to imagine that network interface working without an IP address. You write your IP routing table using IP addresses, but of course you know you're lying when you name the router by IP address; you're just indirectly saying that you want to route via a MAC address. And you have ARP, which gets bridged but not really, and DHCP, which is an IP packet but is really an ethernet protocol, and so on.

Re: The OSI Deprogrammer

#56

Post by kiitos »

throwawaaarrgh wrote: post #8
tptacek wrote: post #5

I mean, this isn't bad or anything, and if Rob Graham is going to write about anything with authority, this is certainly it, but it's kind of beating up a straw man, isn't it? I feel like "layering violations" went out the window with HTTP, and certainly by the time we got WebSockets.

It's not great. It reads like some angsty nerd who really doesn't like the idea of something and has spent an inordinate amount of time trying to explain why it's wrong and why his solution is better. Sometimes that works, if your name is Locke or Marx or something, but this writing really doesn't say "genius" to me. Moreover, I have not heard anyone but myself utter the word "OSI" out loud in 8 or 10 years. Maybe it…

I definitely have used the OSI model in all of my corporate jobs for my entire career.

Re: The OSI Deprogrammer

#57

Post by robertgraham »

aidenn0 wrote: post #37

I'm really surprised that RFC-1122 isn't mentioned in the entire article; The IETF defines 3 layers for a TCP/IP network (link, IP, and Transport) and those layers are (IMO) much more useful for understanding the internet than the OSI model. In addition they don't pretend to be describing a platonic ideal of a network, but rather just specific parts of the network we are using today.

I use RFC 791, the original Internet model.

RFC 1122 is a retconned version of the Internet model that tried to change terminology to fit OSI.

Re: The OSI Deprogrammer

#58

Post by jdougan »

robertgraham wrote: post #44
tptacek wrote: post #5

I mean, this isn't bad or anything, and if Rob Graham is going to write about anything with authority, this is certainly it, but it's kind of beating up a straw man, isn't it? I feel like "layering violations" went out the window with HTTP, and certainly by the time we got WebSockets.

Well, as my footnote in the abstract says, it's swatting a fly with an anvil. It's not about people who already deal with this, like Nanog or systems programmers. Whatever model they've started with, they've developed a better one in their heads based upon their own expertise and experience. It's about educators who continue to teach out-dated concepts. Take Presentation Layer #6 as an example. It was created because…

Apparently Mike Padlipsky lived in vain [1].

I'm surprised you did not mention his work. Defining and critiquing the reference models in the case of the ARPAnet and OSI was his thing. Much of it was published in a (quite funny) book The Elements of Networking Style[2], partially taken from some of his RFCs.

[1] https://en.wikipedia.org/wiki/Michael_A._Padlipsky

[2] https://archive.org/details/elementsofnetwor00padl

Re: The OSI Deprogrammer

#59

Post by KirillPanov »

"All models are wrong, some are useful".

People refer to the OSI model because it's useful, and can be explained in about 1/30th of the text needed to produce this rant. Why not propose an equally-concise yet more useful model instead?

Re: The OSI Deprogrammer

#60

Post by KirillPanov »

aidenn0 wrote: post #37

I'm really surprised that RFC-1122 isn't mentioned in the entire article; The IETF defines 3 layers for a TCP/IP network (link, IP, and Transport) and those layers are (IMO) much more useful for understanding the internet than the OSI model. In addition they don't pretend to be describing a platonic ideal of a network, but rather just specific parts of the network we are using today.

Yeah; as somebody who was taught the OSI model at university, they basically straight up told us to ignore the OSI model above layer 4, and to treat the remaining stuff as a starting point rather than the whole story.

So criticizing layers 5/6/7 is not really helpful. Those are not actually taught anymore.

Post reply on HN