Slashdot Mirror
MIT No Longer Owns 18.0.0.0/8 (ttias.be)
An anonymous reader shares: MIT no longer owns 18.0.0.0/8. That's a very big block of scarce IPv4 addresses that have become available again. One block inside this /8, more specifically 18.145.0.0/16, was transferred to Amazon.
I did it! I read the whole article. And so did everybody who read the summary.
Nobody envisioned more than a few thousand computers on the Internet. The notion that hundreds of millions would be limiting was ridiculous.
Technically, just the MIT LCS lab, not even the whole school, had the A block.
DEC also had a class A address block for a while. HP got this, plus their own, when they bought DEC. At one point, HP had twice as many IP addresses than China.
The Internet grew way more than any of the founders thought. 4 billion addresses seemed huge back then. Look at the List of assigned A blocks and how A blocks were thrown around in the early days.
Someone traded 10.0.0.0/8 for it. MIT got a deal, because like /. UIDs, lower numbered ones are better!
"National Security is the chief cause of national insecurity." - Celine's First Law
In the same regard, Xerox no longer owns all of 13.0.0.0/8 either. Amazon has a piece of that pie, too.
https://whois.arin.net/rest/or...
Needs an "M" in there for "misleading". MIT hasn't released the entire /8 back to ARIN; AFAICT from whois queries they've transfered a whole bunch of /16s (20+) directly over to Amazon, all of which are above the 18.145.0.0 line. Given the highly non-contiguous allocations across the upper half of the /8 range the most likely cause is that they've received chunk of cash for giving Amazon all the /16s that they were not currently actively using.
UNIX? They're not even circumcised! Savages!
MIT claimed all of 18.x.x.x early on and just held onto them. When I was there in 2007, I believe they let my frat have full control of all of 18.236.x.x, no subnets required, for 40 guys.
For those interested, Wikipedia has an amusing list of original A level IP assignments: https://en.wikipedia.org/wiki/...
Because IPv4 was designed as a limited proof of concept, with IPv6 being the properly designed replacement.
IPv4 was supposed to be deader than a can of SPAM by Y2K, and as historical as stacks of punch cards at this point.
-- Sometimes you have to turn the lights off in order to see.
Ford still seems to own 19...
Halliburton, Eli Lilly, U Michigan, Prudential, Merck are some of the more notable assignees.
Some of these must be subnetted and farmed out, but IPv4 is destined for obscurity, so why bother?
Still, reading RFCs and seeing Jon Postel's name makes me want to tear up. Miss him.
deleting the extra space after periods so i can stay relevant, yeah.
https://gist.github.com/simons...
"Fourteen million of these IPv4 addresses have not been used, and we have concluded that at least eight million are excess and can be sold without impacting our current or future needs, up to the point when IPv6 becomes universal and address scarcity is no longer an issue. The Institute holds a block of 20 times 10^30 (20 nonillion) IPv6 addresses.
"As part of our upgrade to IPv6, we will be consolidating our in-use IPv4 address space to facilitate the sale of MIT’s excess IPv4 capacity. Net proceeds from the sale will cover our network upgrade costs, and the remainder will provide a source of endowed funding for the Institute to use in furthering its academic and research mission.
--
There is a lot of expensive real-estate tied up in these "8-blocks"
https://en.wikipedia.org/wiki/...
HP, by virtue of their acquisition of the assets of DEC, has 2 8-blocks, which is probably worth a small fortune in real money. 33 million IP4 addresses.
Most (all?) of these were reserved in the great IP address land grab back in the early 90s.
there are 3 kinds of people:
* those who can count
* those who can't
The internet back then was mostly dialup. Even most schools would exchange email/news with each other via automated dialup at night when phone rates were lower. Consequently, most of the Internet traffic was store-and-forward. You sent an email, your mail server dialed up your school's computer and delivered the mail to them. The school's computer would hold it until night, when it would dial out to the main university in the area and deliver your mail. The university computer, being a minor hub would dial out more frequently, so after say an hour it would dial out to the man hub in the region and deliver your mail.
The main hubs were the ones with always-on dedicated links to other major hubs. They were the ones which got the class A subnets. It made sense because then they could then parcel out the IP addresses to the minor hubs and spokes as they saw fit, and thus DNS resolution could always be handled locally (and thus immediately). For those of you who weren't on the Internet back then, because data was mostly being transmitted as store-and-forward, email typically was only slightly faster than postal mail (usually took a few hours to days to reach someone on another continent), and DNS changes could take up to a week to propagate through the entire Internet. So being able to resolve DNS changes locally quickly was a big deal.
LOL at HP.
15.0.0.0/7
There are two types of people in the world: Those who crave closure
Oblig xkcd https://xkcd.com/865/
Mod up (+1, Facial)
FTFY.
Central Ohio Home Theater Installation - The Theater People
The internet wasn't even public. If they had known Al Gore and the other "Atari Democrats" in Congress were going to come along and spend money opening it to the world, they would have started with something like IPv6. We got IPv4 because it was only for institutional communication and research. MIT is big on both of those, they do lots of work for government, for industry, and in collaboration with other institutions, in addition to their world-renowned research programs.
In that environment, anybody important enough to even plug in could get a /8. People these days still don't understand what the internet is, no wonder we're out of IPv4 addresses! If the average idiot they let plug in now understood the resource, we wouldn't even be running out of IPv4 addresses for years. We have to switch to IPv6, because the commons is a tragedy.
Here's the changelog from the ARIN list if anyone's interested:
http://lists.arin.net/pipermail/arin-issued/2017-April/003050.html
Or, let's just all migrate to IPv6 and be done with this.
14.0.0.0/8 and 15.0.0.0/8 could be combined to 14.0.0.0/7 (or 15.0.0.0/7 if you prefer). 15 and 16 can't be combined. Do it in binary and it will be more obvious.
I have, like, TONS of 192.168.x.x addresses and I only use a few. How can I sell the rest?
Dear Slashdot: next time you want to mess with the site, add a rich-text editor for comments.
15.0.0.0 is not /7-aligned. 14.0.0.0/7 and 16.0.0.0/7 are valid, but 15.0.0.0/7 is not.
Even if 15.0.0.0/7 were valid CIDR notation, it would include 14.0.0.0 - 15.255.255.255, not 15.0.0.0 - 16.255.255.255 as was intended.
Back in the good old days when the Internet was used mainly for colleges and government agencies. These colleges will reserve large blocks for themselves. I went to a small college and they had a Class B and two Class C IP address ranges (Back in the time Class B was X.X.0.0 and Class C was X.X.X.0) giving the colleges more than enough IP address for the who institution. Hack in the late 1990's every PC was connected to an unprotected Static IP address for their PCs.
If something is so important that you feel the need to post it on the internet... It probably isn't that important.
Actually, this is entirely on ARIN rather than ICANN these days, and they absolutely allow transfer of IPv4 space for money (subject to a few criteria) and have done so for some time as part of their approach to dealing with IPv4 exhaustion. There's also nothing to say that these IPs have never been used by MIT - for all we know they were previously in use but have been freed up as part of MIT's IPv6 rollout - and since Amazon needs IPv4 space for their growing cloud platforms and can clearly afford this many IPs in one go it makes sense for MIT and Amazon to do a deal rather than parcel them out piecemeal to multiple users.
IPv4 space has been a resource with a sell by date for some time; at some point (probably still some way off) IPv6 will gain critical mass and the value of IPv4 space will plummet, but until then its basically a game of chicken against that unpredicatable deadline. You can sell now, and maybe get $10/IP (for suitably large allocations), or you can wait a bit longer and gamble on either making more money for your IP space as people get more desperate, or wiping out because IPv6 has finally taken off and demand for IPv4 space has dropped. MIT could easily have held on to the IPs for a few more years, and would likely have make a lot more as a result, but by doing a deal now they've actually helped Amazon grow their cloud and put the IPs into productive use again. Sure, MIT likely made a lot of money on the deal, but that's still better than having the IP space sitting around doing nothing at all.
UNIX? They're not even circumcised! Savages!