The million dollar mystery behind Milk.com | Hacker News Reader

The million dollar mystery behind Milk.com

npr.org

26 pointsdanso50 comments

  Back in the internet's early days, it was easy to get a domain name. They were cheap or even free.

That is not exactly how I remember things. Everyone bought domains from one registrar (was it Network Solutions?) and the cheapest domain started at around $100. There were all sorts of short and trademarked dot-coms available though, that is true.

I'm going by ancient memory here and would love it if someone corrects me.

Plus you could only interact with NS by filling in fields in a text file and emailing it to them, and hoping you did it right so their script made the changes you intended. Even at the time, this was annoying.

You could get free domains in the 90s still, a few that were totally free because they wanted to sell you hosting, many of them ad supported with an iframe however (although you could easily break out of the iframe and present your content without it).

Ads on the internet were mostly prohibited during the early-internet period due to the NSFNET AUP.

I used NameZero for johnedgar.com in 98, and it had an iframe ad, Freeservers, FortuneCity, and Xoom also all gave free domains in exchange for ads.

Yes, but that was about five or ten years after the end of the early internet, well into the dot-com bubble. The internet proper had been operating for 20 years at that point, and the first ARPANET packets were 29 years earlier. The NSFNET AUP had ceased to be a factor 3 years earlier: https://www.merit.edu/wp-content/uploads/2025/01/Retiring-th... All of those companies were founded in the com-priv era.

Apparently we have different definitions of "the early internet" - I consider it anything before 2000. If the argument is the 90s are not the early internet therefore I am wrong, I'll accept that perspective.

Apparently, and there is no objectively correct answer. But the year 2000, or even 01998, is closer to today than it is to RFC 20, the oldest surviving RFC.

From my point of view at the time, companies like NameZero and Xoom were Johnny-come-lately scumbags in danger of destroying the internet through their flagrant disregard for its culture and traditions. Looters. And, from today's perspective, I'm not sure I was wrong.

That's interesting to hear - I presume you're a little older than me I'm 40. I never met anyone else in person in the 90s who also used the internet (I grew up in rural uk) - everyone I knew who used the internet was on the internet, so for me it felt extremely undiscovered. FWIW I probably would never have gotten into the web in the way I did had it not been for NameCheap, having a domain and being able to show off on IRC was a lot of what drove me forward so personally I'm incredibly grateful of them. :)

Yeah, I was in favor of commercialization precisely because it was necessary for inclusion, but horrified by how much of that commercialization turned out. I started using it in 01992, at which point there were no ISPs anywhere within thousands of kilometers, so the only way to get access to the internet was to know people in person.

NameZero was from 01999 evidently: https://www.crunchbase.com/organization/namezero

The leading zero that only applies to pre Y2K years sets my teeth on edge, so mission accomplished there

You're a troll. Not everything in the universe is about you. Fuck off. Nobody should have to put up with people like you.

My comment was totally in jest, but in any event, I can't think of a comment that is more against the HN guidelines than yours, and I've been here longer and more frequently than I'm willing to admit IRL

https://news.ycombinator.com/newsguidelines.html

Oh, well, then, so was mine! Got you, ha ha!

  I never met anyone else in person in the 90s who also used the internet

Quite a few young people were on the internet in the early '90s, but almost entirely university students. I noticed quite a few from GB on Usenet. I remember perusing threads about English 'raves' since the rave phenomenon had barely reached America.

Yeah! It was amazing!

The Internet was no longer early after it was front page news in mainstream magazines and more importantly, certainly not after Microsoft acquired FrontPage from Vermeer due to enough of a market for web publishing.

When I joined a couple other guys working to build the southeast US's fastest growing ISP in early mid-90s, enough mom and pops, small businesses, and government agencies -- from backyard bird feeder webcams to US Navy recruiting -- were already staking their brochureware claims that it was worth marketing ourselves as offering FrontPage hosting.

Windows 95 already took you online fine, while Windows 98SE bundled IE 5 which included the first version of XMLHttpRequest, giving birth to Ajax, enabling single page apps... and of course e-commerce in ad banners.

By the time you are calling the early Internet, everyone called it Web 2.0, and most things are still that.

Interestingly, I met tons of teens from rural UK in early 90s online, in MUDs and MOOs and especially Talkers such as Surfers: https://muds.fandom.com/wiki/Surfers (Small world, several went on to found or lead top web 2.0 digital agencies.). As a sibling comment notes, most were at uni.

By that time, the early 90s, it was then considered the end of the early Internet, beginning of mainstreaming. US law changed to allow making money online, Web 1.0 got its footing, and most of all, it was the start of Eternal September: https://en.wikipedia.org/wiki/Eternal_September

>That is not exactly how I remember things. Everyone bought domains from one registrar (was it Network Solutions?) and the cheapest domain started at around $100

By the mid to late 90s (my own earliest domain) yes. But a quick whois on milk.com shows a creation date of 1985-01-01, and that dates back to basically the foundation of DNS at all, RFC 882 & 883 were 1983. Initially it was just the Stanford Research Institute manually maintaining HOSTS.TXT for arpanet. IIRC you like, literally called up Elizbeth Feinler during business hours and just asked and that was it. This is all second hand because I was just a wee lad at that point and wouldn't even discover BBS until well after that heady era!

As things got formalized yeah unfortunately some players were able to start grabbing at monopolies, and NetSol began operating the registry for DISA in I think 91 or 92? It was still initially free, but then pretty quickly they started commercializing. But if someone got in truly on the ground floor it was pretty informal, academic and free. Getting a block of IPv4 was pretty easy/free for a long while into the 90s as well.

That explains it, thanks! I wasn't on the net till the '90s

That wasn't the early days of the internet. It was free when I registered ear.org in probably 01994, 24 years after the ARPANET started and 16 years after the first TCP/IP internet nodes. You just had to have a nameserver set up.

Shortly after that they started charging, and I hadn't carried out my plans with it, so I let it go, figuring that someone else would make better use of it.

They didn't.

I googled ear.org. Complete waste tbh. It's just blank

> (was it Network Solutions?)

rs.internic.net, if memory serves. It stopped redirecting to anything some years ago.

If he’s sitting on a domain that’s actually worth millions and has a well-paying job, maybe he’d get a bite for “I’ll accept less than 10 if you match it 1:1 with a donation to <charity>.”

It’s his and he can do what he wants. In fact I think it’s cool that there’s still a Classic Web holdout like this. But there’s latent joy locked away in that url.

There are a lot of people that saw domain names as real estate in the 90s. :D

I had a lot of offers for wicker.com. In 98? 99? some guy asked what I wanted, and I said give me something interesting so I have a good story to tell.

He offered a 1544 Dante, which I accepted. His gig was auctioning estates and his wife imported and sold wicker.

I found out about milk.com when I was thinking about how to make an Android app completely from scratch (assembling DEX bytes from zero, kind of like writing an assembler for the Dalvik VM). That’s when I came across the author of the DEX format, Dan Bornstein — and I was surprised he actually owns a domain like that.

In a weird way, when someone mentions something about milk with a domain name, it always remind me of the Milk Table.

https://milk.dk

Not to be confused with a milking table. (Not a term I'd recommend googling at work, by the way.)

I don't dislike the design, but this (500€!) aquarium is criminally small, and there's two fish in there!

https://milk.dk/wp-content/uploads/2021/09/Milk_aquarium-102...

I thought this was going to be about Kevin Rose.¹

¹ https://techcrunch.com/2011/04/26/milk-completes-1-5-million...

purple.com was a similar thing. Originally the site had just a FAQ page about how they didn't want to sell the domain:

https://web.archive.org/web/20090226000820/http://www.purple...

https://web.archive.org/web/20090225184613/http://www.purple...

I just checked now and it's one of those mattress drop-shipping sites :-(

I use http://neverssl.com as its replacement when checking internet connectivity.

I guess everyone has a price

I hope the original owner made a fortune from it and not had it taken via a trademark dispute.

I believe this was one of, if not the first site to implement a barcode/UPC generator. Current iteration here:

https://milk.com/barcode/

The guy pretends he doesn't want to sell the website, all while hinting exactly at who might want to buy it and for how much... lol

This guy worked at Danger! Man I loved my Hiptop/Sidekick

https://milk.com/value/

I think he’s aiming a bit high with 10 milly. I’d take 1 or 5.

I'd love to have a job that I liked so much I'd keep doing it rather than cash out a domain name and retire.

If you look through his LinkedIn [1], he was at Google from 2005-2011, as the tech lead for Dalvik, and spent some time as a Senior Staff Engineer at Slack. I would guess he can probably retire anytime he likes.

Agree that $10M is too high, but you have to drop your anchor somewhere.

[1] https://www.linkedin.com/in/danfuzz/details/experience/

If the milk marketing budget is truly 100 million a year (a cursory check shows it's likely closer to 200 million now) as he mentions, 10 million doesn't seem very high at all.

So they've had a 100 million per year marketing budget for at least 10 years. Let's say they spent 1% per year towards the domain... The opportunity cost for not having something as obvious and useful as milk.com seems significant. To me it seems like they've already made a mistake by not paying up despite having the means, and now they've dug their foot in. They've spent hundreds of millions on billboards, yet they can't even put "milk.com" on their billboards next to "got milk". Imo but 5% less billboards for a few years and buy the domain...

Idk it seems like everyone knows about milk at this point, is milk.com vital? We buy our books at Amazon.com not books.com. And you wouldn’t even buy milk at milk.com it would just be info about milk.

I agree it sort of seems like a waste of money, but this is what they are doing year after year - spending money to promote milk. May as well stretch to have milk.com at that point if you plan on spending 100 mil a year for the foreseeable future on promotion. Then again, they've already blown through the entire old-school .com age without getting the domain, maybe they are better off just riding it out until web addresses are entirely irrelevant rather than just mostly irrelevant.

The value isn't to the milk cartel, it's to alternative milk producers. Imagine an ad for a new oat milk with "milk.com" as the domain. It's got a ton of media value.

In some countries you're not allowed to call them milk, for example, in the UK a judge ruled that they must be called something like "oat drink" instead.

Which actually makes sense. If it's not coming from a mammal's mammary glands, it's not milk!

Calling it milk IS deceptive marketing.

they'd not be calling their product milk, it would just be the name of their domain. you don't buy amazons at amazon.com

It’s affected by how Google currently ranks the plain domain name in a search. It used to be higher than it is now, but it still has got a lot of weight

If you listen to the podcast, if I recall correctly, he said he’d consider numbers higher than 1m.

This is why you're not a CEO. You would fail to extract maximal shareholder value.

Reply on news.ycombinator.com