Hacker News Reader

Comment by Animats | Hacker News Reader

The article misses a big point. The OSI model originally envisioned operating over X.25 connections operated by telcos. Then came local area networks. Most network traffic was, for a long time, local; in business settings, it often still is. TCP/IP worked well both for local and distant networking, and worked through gateway choke points between LANs and larger networks. X.25 over a LAN was a huge mismatch; X.25 was intended for slow links.

OSI TP4 eventually could do everything TCP/IP could, and Microsoft included it in NT (NT4 and Windows 2000, I think), so it did ship. But nobody cared much.

View on HN

It may have shipped, but the issue was that there was very little interoperability. Two independent implementations probably could not communicate, which seriously hindered adoption of the OSI model.

The TCP/IP community had Connectathons, where interoperability was verified between vendors.

The elephant in the corner of the room is that TCP/IP is a bad protocol. That's why hacking is on the front pages of the newspaper every day now. Sometimes it is better to let the grown-ups design things.

TCP/IP is perfectly fine. It's the security of endpoints (OSes, language runtimes, etc.) that sucks and is largely responsible for our perpetual security disaster. Why do we still have buffer overflows in 2015? OSI would not save us from that. If anything, the byzantine complexity of OSI would result in giant blobs of C code that contain even more vulnerabilities. Complexity is a very bad thing security-wise-- a linear increase in complexity always leads to an exponential increase in opportunities for bugs. Bugs in turn are fixed in linear time. Exponential effects always trump linear ones, hence our security nightmare.

If grown-ups know better they should show us by shipping things. If I can't use it it doesn't exist.

IMHO one of the more general anti-patterns in security is to try to secure the network not the device. It's a fool's errand for many reasons, most notably the fact that there is an intrinsic trade-off between the transparency / audit-ability / firewall-ability of network protocols and the security of the protocol itself. A secure protocol is utterly opaque and therefore cannot be inspected at the perimeter. Secure the device. While it's important to pay attention to your network, if your devices are not secure netsec will not save you.

By what stretch of the imagination is TCP/IP a bad protocol? It is a 30+ year old standard that is still in wide use today. It has managed to scale from a few networked computers to billions. From an engineering perspective, TCP/IP is a good protocol.

From a security perspective, most issues with modern computer security have absolutely nothing with the TCP/IP protocol. If you take a look at top 10 vulnerabilties (https://www.owasp.org/index.php/Top_10_2013-Table_of_Content...), none of them are directly related to TCP/IP.

It is always easier to look back in hindsight and criticize something knowing what we know today. Saying we should leave design to "grown-ups" is not a fair assessment of TCP/IP and its designers. Engineering is difficult and every design comes with trade-offs.

P.S. If you want to see truly bad protocols, look to ones designed by "professional", "grown-up", organizations. CORBA and anything WS-* are a few that come to mind.

I assume that if the Internet had been designed "correctly" with public-key authentication down to the individual user [1] (with $300 VeriSign certs, natch) then it would have been too expensive and brittle for universities, hobbyists, or home enthusiasts to use and thus a rebel hacker alliance would have created their own "insecure" network (with porn and worms! and rants about "the right to network"). Maybe it would have been some sort of real-time upgrade to FidoNet.

[1] This wouldn't stop attacks, of course, but it would provide strong enough attribution to create a deterrence effect (at least in The Man's mind).

Reply on news.ycombinator.com