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

91–95 of 95 posts

Re: The OSI Deprogrammer

#91

Post by RealityVoid »

robertgraham wrote: post #90
RealityVoid wrote: post #89

I think this might be the favourite thing I read on hacker news. Thank you so much Robert Graham for writing this, I always considered the OSI model fishy, and although I mostly worked with stuff like CAN (that was also mentioned) I was irked when people tried to shoehorn OSI model _even there_ ! If I ever meet you, robertgraham, I am definitely going to try buying you a beer. One thing I am a bit confused though is…

Ethernet doesn't use TCP/IP. Ethernet is it's own network. It has nothing to do with TCP/IP. Other things use Ethernet. Routers, when connected to each other, often use a local Ethernet network to communicate. Think of the TCP/IP Internet as it's own network, ignoring how routers physically talk to each other. Sometimes it's a directly link, a wire. Sometimes it's carrier pigeons. Sometimes it's WiFi. Sometimes it's…

I think my confusion arises that, when I access some other PC on the local network, I connect to an IP and port. That's not on the wide internet, but it uses IP to communicate. Therefore... where's the ethernet's role in that?

Re: The OSI Deprogrammer

#92

Post by jdougan »

onetimeuse92304 wrote: post #15
mannyv wrote: post #14

I always thought this model was pretty worthless. I remember when they tried to apply it to the internet. It didn't quite fit, but they needed a model because intellectuals.

It depends on when it was. Back in 1999 it fit pretty well. Even if not perfect it still was a good teaching tool to explain "how it all works" on some level. Then you got to know the hairy details. Nowadays "Internet" is mostly HTTP 2 for most people and then this model doesn't even remotely explain the structure of it.

When I was first exposed to the OSI model in the late 80s, it wasn't very good then either.

Re: The OSI Deprogrammer

#93

Post by jdougan »

idlewords wrote: post #75

This document is great but could use some tough-love editing. The author drives home the same points repeatedly, and in many cases redundantly, in a way that does no favors to the substantive content of the argument. Sometimes less ranting makes for a more devastating critique.

I think the history parts need a serious rewrite. Still lots of people alive who lived through this.

Re: The OSI Deprogrammer

#94

Post by brookst »

fanf2 wrote: post #83
brookst wrote: post #77

Earlier quoted context omitted.

The key concept in routing is hierarchy, which Ethernet does not have. It’s an extra dimension that changes the way the protocol and applications work at that layer. You can say a cube and a square are the same thing, but I don’t think that’s particularly useful pragmatically. Ditto for L2 and L3; if you abstract out the difference, they are indeed the same. But that’s not useful. Now, you could say that time has bro…

In my experience working alongside network engineers, they use “layer 2” as a synonym for Ethernet, “layer 3” as a synonym for IP (v4 and v6), “layer 4” as a synonym for TCP and UDP, and “layer 7” as a synonym for application protocols. They don’t talk about the other layers. It’s just jargon, without any deeper meaning.

Well yes, those are the network layers, so network engineers think about them the most.

Admittedly layers 5 and 6 are kind of blurry, but so what? Doesn’t mean the model isn’t useful, just that it isn’t a perfect map=territory.

Re: The OSI Deprogrammer

#95

Post by carlosm3011 »

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.

I don't think that Bob's essay goes against layering per-se. It's about the OSI layering specifcially not being suited to describe the current Internet anymore.

Layering is very much alive, but I tend to agree that we need a richer and better suited layering model.

Post reply on HN