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

The New Internet

tailscale.com

Post reply on HN
Print view View original

281–290 of 315 posts

Re: The New Internet

#281

Post by DyslexicAtheist »

jeroenhd wrote: post #266
DyslexicAtheist wrote: post #90

Earlier quoted context omitted.

I never thought of this. Forces me to rethink every negative post people made against DNSSEC which shaped my opinion. I still think that IPv6 and DNSSEC do more harm in practice than what they solve. Maybe the SCW podcast can do a deepdive on this together with somebody who is militantly-pro DNSSEC. edit: maybe even invite 2 or 3 DNSSEC advocates @tptacek :)

I may be in favour of DNSSEC, but I admit that it's time for a v2 of the RFC that removes the stupid "encryption can't be done at the endpoint" restriction. In practice you can just turn on validation on many computers and gain its benefits, especially if used in the manner as described here where you can just block connections to unprotected hostnames to work around the most glaring issue, but the whole spec is writ…

thanks!

Re: The New Internet

#282

Post by teddyh »

tptacek wrote: post #195
DyslexicAtheist wrote: post #90

Earlier quoted context omitted.

I never thought of this. Forces me to rethink every negative post people made against DNSSEC which shaped my opinion. I still think that IPv6 and DNSSEC do more harm in practice than what they solve. Maybe the SCW podcast can do a deepdive on this together with somebody who is militantly-pro DNSSEC. edit: maybe even invite 2 or 3 DNSSEC advocates @tptacek :)

I don't think the analysis upthread should make you rethink DNSSEC, since it, too, is a centralized system; rather than being controlled by Avery Pennarun (you could do worse), it's controlled by an unholy alliance of world governments and companies like Verisign. If we could find a credible DNSSEC advocate (for our audience; that is: a cryptography engineer, vulnerability researcher, or an engineering leader at a ma…

The title of that article which you link to is “Calling time on DNSSEC?”, and Betteridge's law of headlines applies to it. Here’s the final paragraphs from that article:

“I guess the question we should be asking is — if we want a secured namespace what aspects should we change about the way DNSSEC is used to make it simpler, faster, and more robust?”

Re: The New Internet

#283

Post by aborsy »

I used to use Wireguard. It connects a peer to a peer, but stops there. I have since replaced it with Tailscale. It goes much beyond Wireguard, and connects everything to everything. A lot of my networking problems went away. After using Taildrop for several months, I feel the post is right about it. It’s a frictionless one click peer to peer file transfer tool that is very useful. It should have been built into the internet.

The idea of entirely decentralized internet is wishful thinking. You always need servers. Even with IP6, you have to run a STUN or DDNS server, since ip addresses change. Do you want to run them at home? I don’t.

I do think Tailscale is on path to different networking.

Re: The New Internet

#284

Post by ragall »

Esras wrote: post #273
rsync wrote: post #199

Earlier quoted context omitted.

"An incredibly long ramp up ..." Agreed. We would all do well to learn about, and begin implementing, "Iceberg Articles": https://john.kozubik.com/pub/IcebergArticle/tip.html

This feels like an overly-complex treatment of the Inverted Pyramid in journalism: https://en.wikipedia.org/wiki/Inverted_pyramid_(journalism) , or Bottom Line; Up Front: https://en.wikipedia.org/wiki/BLUF_(communication) . Start with the important statements, then expand. Doesn't have to be the "Tell you what I'm telling you, tell you, tell you what I told you" format that many (American) students were taught, but s…

BLUF is bad, it's precisely a technique borne in the the world of newspaper publishing for writing catchy articles (what is now called clickbait). Classical philosophical writing is the exact opposite: start with some problems, elaborate in high detail and finish with a conclusion (the name says it all).

Re: The New Internet

#285

Post by ragall »

mrkeen wrote: post #256
j2kun wrote: post #179

Earlier quoted context omitted.

This rests on the incorrect assumption, pointed out in the post, that most applications need the kind of scale that warrants quickly scaling to more servers.

The article tried to make k8s all about scale, when that's not the whole story. On other socials, a screenshot of the 'Not scaling' section is getting responses of "Those idiot developers think they need k8s scaling for their 1 req/s sites, ha ha." The author brags about being able to (skip testing, CI/CD pipelines and just) edit their perl scripts (in prod,) really quickly. What uptime is associated with that practi…

In an enterprise environment, I'd like a networking solution that allows me to run an app on my own office workstation and expose it as a service to some part of the company at an SLO that can be reasonably be guaranteed with a workstation: 99.9%. That would allow to cut so much time in "productionizing" stuff that doesn't need CI/CD pipelines or deploy to a datacenter: just me editing a Python file and restarting it.

Re: The New Internet

#286

Post by ragall »

Borg3 wrote: post #99
wmf wrote: post #24

Earlier quoted context omitted.

Apenwarr is kind of an IPv6 hater. He thinks it's not going to happen.

Because IPv6 is mistake. Thats why market does NOT want it. Unfortunately, we all start to feel the heat of IPv4 exhaustion. Anyway, remember IPv4 classes? Then they made it classless. IPv6 is not 128bit, its just 64bit with 64bit host address. So, first mistake. IPsec mandatory? pure stupidity. Crypto moves fast, every 10 years many protocols are obsoleted. How you will provide E2E connectivity with that? In 1997 IP…

> Thats why market does NOT want it.

What market are you talking about ?

Re: The New Internet

#287

Post by tptacek »

teddyh wrote: post #282
tptacek wrote: post #195

Earlier quoted context omitted.

I don't think the analysis upthread should make you rethink DNSSEC, since it, too, is a centralized system; rather than being controlled by Avery Pennarun (you could do worse), it's controlled by an unholy alliance of world governments and companies like Verisign. If we could find a credible DNSSEC advocate (for our audience; that is: a cryptography engineer, vulnerability researcher, or an engineering leader at a ma…

The title of that article which you link to is “Calling time on DNSSEC?”, and Betteridge's law of headlines applies to it. Here’s the final paragraphs from that article: “ I guess the question we should be asking is — if we want a secured namespace what aspects should we change about the way DNSSEC is used to make it simpler, faster, and more robust? ”

I'm happy just to see more people reading it. People can make their own call about it.

Re: The New Internet

#288

Post by RulerOf »

gorgoiler wrote: post #261
api wrote: post #161

Earlier quoted context omitted.

There is another reason: the addresses are long and impossible to remember and hard to type. I always bring this up and it’s always dismissed because tech people continue to dismiss usability concerns. Even “small” usability differences can have a huge effect on adoption.

In practice I haven’t ever had a problem memorising IPv6 addresses. The significant proportions of any address you might type manually are 48 bits long at one end and a few bits at the other. An example IPv4 address is 8 to 12 digits: 10.30.115.5 A memorable IPv6 address at a /56 site — the prefix and then one or two digits — isn’t much longer: 2001:db8:404:14::42 If you’re with a reasonably clued in ISP you probably…

>the prefix and then one or two digits — isn’t much longer

I'd argue it's just enough to make the difference though.

The problem is that people got used to being able to rely on memorizing IP addresses. IPv6 does its best at making IP addresses both harder to memorize, and completely dynamic, going so far as to change the IP on a fairly regular basis. It's antithetical to some very core qualities that an IP address is supposed to have in the minds of many.

Re: The New Internet

#289

Post by tempcommenttt »

gorgoiler wrote: post #261
api wrote: post #161

Earlier quoted context omitted.

There is another reason: the addresses are long and impossible to remember and hard to type. I always bring this up and it’s always dismissed because tech people continue to dismiss usability concerns. Even “small” usability differences can have a huge effect on adoption.

In practice I haven’t ever had a problem memorising IPv6 addresses. The significant proportions of any address you might type manually are 48 bits long at one end and a few bits at the other. An example IPv4 address is 8 to 12 digits: 10.30.115.5 A memorable IPv6 address at a /56 site — the prefix and then one or two digits — isn’t much longer: 2001:db8:404:14::42 If you’re with a reasonably clued in ISP you probably…

This is the first time I read about someone actually trying to remember IP6 addresses, maybe I should try that, because it’s really easy to remember IP4. For me, the problem is that there’s hex numbers, which are harder to remember and missing zeros, so you need remember the colons. If IP6 would just be 6 decimal numbers and this would be the default way of writing them, this would not be a problem. But it feels to me that the cryptic way IP6 is written is to make it hard for humans to remember it.

Re: The New Internet

#290

Post by tempcommenttt »

wolfendin wrote: post #242
api wrote: post #161

Earlier quoted context omitted.

There is another reason: the addresses are long and impossible to remember and hard to type. I always bring this up and it’s always dismissed because tech people continue to dismiss usability concerns. Even “small” usability differences can have a huge effect on adoption.

That’s because nobody normal— anyone who isn’t a tech person— remembers IP addresses. Hell I can’t get tech people I work with to give me their public IP.

Maybe young people can’t, but older folks can easily store phone numbers in their brains, so IP4 is easy.
Post reply on HN