Re:Dawn of a New Internet Era
by dreen on Wednesday October 23, 2013 @07:11PM
(#6602418)
Serious question here, what are the technical limitations in ditching gTLDs and allowing domain names to be any string?
[ Parent
| on HN ]
Someone has to have root servers that serve the children of the TLD...where would those be looked up? The closest you could do is assume a TLD (i.e. .com) if it's not typed, but then you harm the non-.com names. It needs to be a tree where roots are controlled.
We have this already. All TLDs are under the root domain, ".".
You could start here http://www.icann.org/en/news/announcements/announcement-27ju...
Someone has to maintain a list.
OK so the responses seem to boil down to this. My question is rather why that list cannot live in a globally distributed database inside thousands of DNS servers, everywhere, operated by ISPs and volunteers. Sort of like now but every server would have a near-complete list of all domains that have no syntax or schema.
I guess the infrastructure costs would be higher, but what are other downsides? It definitely wouldn't be possible at the time of creation of DNS, but I can't see why it wouldn't be now.
Re:Dawn of a New Internet Era
by dec0dedab0de on Wednesday October 23, 2013 @10:42PM
(#6603131)
Thats exactly how it works now. http://www.iana.org/domains/root/files
It doesnt because we have TLDs. Why is the hierarchy necessary? Why can't we have a one big list of strings matched with IP addresses?