What stopping websites from directly linking to the ip address then? You don’t need a dns if the browser directly points to htttps://231.157.68.8 or whatever. domain names are just branding for fucks sake
This doesn’t work for a ton of sites. You can host multiple domains from one public IP, so if you go to the IP directly, the servers do not know which domain to use. Reverse proxy servers take advantage of this in home labs.
Further this would make it so webmasters have to keep the same static IP forever, lest they alienate their oldest users’ bookmarks. Imagine a business being unable to change datacenters unless they’re willing to walk away from their customer base.
Plus this gets even more difficult to read with IPv6 addresses. Would you want to bookmark something like https://[2a04:4e42:200::67]/ instead of http://xkcd.com/ ?
Maybe you’re getting the joke and I’m not, or maybe you just don’t know, but every computer has a hosts file that it checks before checking DNS (though some browsers skip it by default, behavior you can turn off). If you have a domain in that hosts file, it will always point you to the IP you set for it. So, for instance, if you type 127.0.0.1 reddit.com into this file, then you’ll be sent to your localhost every time you try to go to Reddit, essentially blocking the domain on your computer.
Or, more applicable here, you could find the IP of a page you want to visit that doesn’t have a domain name and assign it to whatever string of characters you want–the hosts file doesn’t even care about DNS TLD rules, so you could have it send you to your self-hosted image storage whenever you type in just “photos”, or to some dark web baguette forger’s site when you type in “carbs.gimmecarbs”.
I know that. Thats not the point. This isn’t about what a ton of sites want, this is about sharing how to connect to web servers that are actively being censored by governments.
You wouldn’t? Bookmarks exist. Websites that are being censored aren’t trying to be remembered anyway. Anyone who would be clicking on an ip only website should know better than to type anything important into them anyway
a “site” is just one computer sharing information with another. limited to the hyper text markup language. Why would a computer that is connected by its IP address need a host name to send html to your computer? It wouldn’t.
Websites are potentially dozens of virtual computers running on a half dozen physical computers, spread across several regions or countries.
There are potentially dozens or even hundreds of IP addresses involved with just the one host name, which all of these components rely on to function.
In a simpler setup, I in one case have a computer with one IP address that hosts 20 different websites. If you send a request to the IP address your request will be dropped because it won’t know which of those 20 websites you’re trying to reach.
1 website being 1 computer reachable on 1 IP address isn’t something you see very often anymore.
sure, if you’re trying to host a modern website that is doing all it can to scoop up peoples private information and compete with the modern web. If you’re trying to just BE ONLINE, you only need an IP address. These sites that are being censored by france don’t need to be 20 websites behind 1 ip. they don’t need to be dozens of virtual computers. If the goal is to not be censored, a single ip address is fine. If the goal is to scrape as much personal data as possible to turn around and sell to corporations and governments well then keep doing what you’re doing.
Subdomain.domain.com might point to a different IP address. You might have systems in mission critical deployments that can’t be updated with new IP addresses but will still work if they point to the domain.
That is if you’re trying to host a modern day website. These sites that are being censored by france probably don’t have any “mission critical” shit like that. You’re thinking like a modern day web developer which has been poisoned by corporations and ‘cyber security’
Nobody is going to memorize IP addresses. If.you want people to shop at your dildo store, it is a.lot.easier to remember davesFatDildos.com vs an IP address.
If created a nice table of names mapped to IP addresses, then you just recreated dns, but worse.
Also, no DNS means no fault tolerance, cert problems, no effective load balancing, very hard to change providers, maintenance of.taking a host down is all but impossible.
There is a reason DNS exists. Thinking we could abandon it just demonstrates a lack of understanding regarding how it works.
Sure if you want to be slow about it. And Google still shows urls so you know which page is legit.
And I don’t know if I’m going to be able to differentiate between 146.238.212.76 and 146.238.73.55 to know which is the correct server to connect to and which one is a phishing scam. Whereas omniva.ee vs omniva-ee.scamdomain.co is pretty easy.
if a website is actively being censored by a government, then people shouldn’t be typing any important information into any IP address only site. IP only links wouldn’t be for the majority of websites, they would be for those that are facing censorship by governments.
Nobody is going to memorize IP addresses. If.you want people to shop at your dildo store, it is a.lot.easier to remember davesFatDildos.com vs an IP address.
Hence my point that domain names are for branding.
taking a host down is all but impossible.
Yeah, that is my point. You can’t ‘take down’ an ip address (except through the obvious end point internet service provider)
The internet is just connecting IP addresses together. Everything you’re talking about has been added onto this fact since the first .com bubble build up in the 90s, which is a bunch of CORPORATE BULLSHIT meant to control the information of simply connecting peoples computers together.
Many websites and web apps these days have hostname filtering at the webserver level. Even if the IP address from the client is correct, it will not permit a connection that way. The req needs to come with one of the valid hostnames in the header, inc the correct protocol. At least, that’s what I do with mine. The req must come with https://whatever/ or https://www.whatever/ or the req gets ignored completely.
If an intermediary such as CloudFlare is used, there’s likely to be another layer of req filtering at the CloudFlare level.
IP certs are quite hard to get, used to be basically impossible not too long ago.
Edit: Not so hard anymore. Letsencrypt added them this year (finally) with the normal http challenge and 6 day lifetime. Which should be fine in normal automated setups. Still not supported accross major acme providers but this’ll make it easy enough to do.
So TLS on bare IPs should be just fine now.
What stopping websites from directly linking to the ip address then? You don’t need a dns if the browser directly points to htttps://231.157.68.8 or whatever. domain names are just branding for fucks sake
This doesn’t work for a ton of sites. You can host multiple domains from one public IP, so if you go to the IP directly, the servers do not know which domain to use. Reverse proxy servers take advantage of this in home labs.
Further this would make it so webmasters have to keep the same static IP forever, lest they alienate their oldest users’ bookmarks. Imagine a business being unable to change datacenters unless they’re willing to walk away from their customer base.
Plus this gets even more difficult to read with IPv6 addresses. Would you want to bookmark something like https://[2a04:4e42:200::67]/ instead of http://xkcd.com/ ?
There is an easy solution. We can just write all the IPs in a text file with easy to remember names!
Like a card catalog? Honestly, it would be nice if we could get copies of individual entries over a dedicated port - like 53.
Maybe you’re getting the joke and I’m not, or maybe you just don’t know, but every computer has a
hostsfile that it checks before checking DNS (though some browsers skip it by default, behavior you can turn off). If you have a domain in that hosts file, it will always point you to the IP you set for it. So, for instance, if you type127.0.0.1 reddit.cominto this file, then you’ll be sent to your localhost every time you try to go to Reddit, essentially blocking the domain on your computer.Or, more applicable here, you could find the IP of a page you want to visit that doesn’t have a domain name and assign it to whatever string of characters you want–the hosts file doesn’t even care about DNS TLD rules, so you could have it send you to your self-hosted image storage whenever you type in just “photos”, or to some dark web baguette forger’s site when you type in “carbs.gimmecarbs”.
I know that. Thats not the point. This isn’t about what a ton of sites want, this is about sharing how to connect to web servers that are actively being censored by governments.
Domain names are not just branding. They’re critical to the infrastructure behind most sites.
Also, how the fuck will you remember all those IPs? And even worse, domain names can let you detect phishing looking at the URL.
You wouldn’t? Bookmarks exist. Websites that are being censored aren’t trying to be remembered anyway. Anyone who would be clicking on an ip only website should know better than to type anything important into them anyway
deleted by creator
a “site” is just one computer sharing information with another. limited to the hyper text markup language. Why would a computer that is connected by its IP address need a host name to send html to your computer? It wouldn’t.
because you don’t understand how it works. it is not only “html”, and often not “just one computer”.
that entirely depends on what kind of site you’re looking to share. Sites that are being censored aren’t trying to act like modern webpages.
A lot has changed since 1999 friend.
Websites are potentially dozens of virtual computers running on a half dozen physical computers, spread across several regions or countries.
There are potentially dozens or even hundreds of IP addresses involved with just the one host name, which all of these components rely on to function.
In a simpler setup, I in one case have a computer with one IP address that hosts 20 different websites. If you send a request to the IP address your request will be dropped because it won’t know which of those 20 websites you’re trying to reach.
1 website being 1 computer reachable on 1 IP address isn’t something you see very often anymore.
sure, if you’re trying to host a modern website that is doing all it can to scoop up peoples private information and compete with the modern web. If you’re trying to just BE ONLINE, you only need an IP address. These sites that are being censored by france don’t need to be 20 websites behind 1 ip. they don’t need to be dozens of virtual computers. If the goal is to not be censored, a single ip address is fine. If the goal is to scrape as much personal data as possible to turn around and sell to corporations and governments well then keep doing what you’re doing.
Alright buddy time to hop back in your Time Machine and head home. We get it, you don’t know how the web works.
Subdomain.domain.com might point to a different IP address. You might have systems in mission critical deployments that can’t be updated with new IP addresses but will still work if they point to the domain.
That is if you’re trying to host a modern day website. These sites that are being censored by france probably don’t have any “mission critical” shit like that. You’re thinking like a modern day web developer which has been poisoned by corporations and ‘cyber security’
Nobody is going to memorize IP addresses. If.you want people to shop at your dildo store, it is a.lot.easier to remember davesFatDildos.com vs an IP address.
If created a nice table of names mapped to IP addresses, then you just recreated dns, but worse.
Also, no DNS means no fault tolerance, cert problems, no effective load balancing, very hard to change providers, maintenance of.taking a host down is all but impossible.
There is a reason DNS exists. Thinking we could abandon it just demonstrates a lack of understanding regarding how it works.
Noone types a url anymore. They go to google and rype what they want
Sure if you want to be slow about it. And Google still shows urls so you know which page is legit.
And I don’t know if I’m going to be able to differentiate between 146.238.212.76 and 146.238.73.55 to know which is the correct server to connect to and which one is a phishing scam. Whereas omniva.ee vs omniva-ee.scamdomain.co is pretty easy.
if a website is actively being censored by a government, then people shouldn’t be typing any important information into any IP address only site. IP only links wouldn’t be for the majority of websites, they would be for those that are facing censorship by governments.
Hence my point that domain names are for branding.
Yeah, that is my point. You can’t ‘take down’ an ip address (except through the obvious end point internet service provider)
The internet is just connecting IP addresses together. Everything you’re talking about has been added onto this fact since the first .com bubble build up in the 90s, which is a bunch of CORPORATE BULLSHIT meant to control the information of simply connecting peoples computers together.
No, even worse, domain names can let you detect phishing looking at the URL. Domains are not accesories. Go to s psychiatrist please. By yesterday.
no u
This idea is so insane that it never even occurred to me. LIKE WDYM Just use IP addresses instead of DNS 😭😭😭🤣
Many websites and web apps these days have hostname filtering at the webserver level. Even if the IP address from the client is correct, it will not permit a connection that way. The req needs to come with one of the valid hostnames in the header, inc the correct protocol. At least, that’s what I do with mine. The req must come with https://whatever/ or https://www.whatever/ or the req gets ignored completely.
If an intermediary such as CloudFlare is used, there’s likely to be another layer of req filtering at the CloudFlare level.
I imagine you could edit your hosts file, or connect to a nonstandard dns to get the correct host name behavior at least
Does this not break TLS?
Only if they don’t have the ip in the cert
IP certs are quite hard to get, used to be basically impossible not too long ago.
Edit: Not so hard anymore. Letsencrypt added them this year (finally) with the normal http challenge and 6 day lifetime. Which should be fine in normal automated setups. Still not supported accross major acme providers but this’ll make it easy enough to do.
So TLS on bare IPs should be just fine now.
They’re just uncommon not used often but if they’re included a lot for load balancing/cdn certs. Especially aws/Microsoft stuff
ISPs could block access to those, too.