The Problem with Vanity TLDs (2011) | Hacker News Reader

The Problem with Vanity TLDs (2011)

templetons.com

The real issue with the vanity TLDs, and many of the newer ones in general, is that they have zero recognition. They are most worthless.

You can slap a joehardware.com on the back of your van, or a local TLD, and people will know that it's the address of your website. Now do the same with joehardware.builders, people have no glue as to what that might be. It doesn't even help to write www.joehardware.builders, that somehow more confusing.

No, the issue with the new/generic/vanity TLDs is that they've lost all meaning. They lack context.

Paint a QR code on the side of the van instead, I bet it would get more engagement even with a .com address.

QR code requires you to scan it in that moment. A human-readable (and -recognizable) URL stands a chance at being remembered at a later date.

> QR code requires you to scan it in that moment

Not really, anymore. Android permits you to take a picture of a QR code and use it's URI later by going to the saved picture and clicking "Lens" (which activates Google's ML to identify QR codes).

Sadly I have no way of tracking that, because I would make the opposite bet. My guess is that 90+% of QR codes, that people aren't forced to scan somehow, are never used.

Lots of restaurants are forcing this. I stop going to them. I don't go to a coffee shop to spend 3-10 minutes futzing with phones instead of talking to the person I'm meeting.

Right? I went up to a cashier the other day and she said, “You know you can order from your table using your phone, right?”

But why would I want to learn some god awful system when I could tell her what I want in a sentence and let her do the data entry for me?

You could write www.joehardware.builders to make it clearer.

I suggested that, but as I wrote, that's actually more confusing if you don't know that there's a .builders TLD (in fact there's also a .builder).

A remembering problem exists with the main TLDs though. Alice's Home Cleaning Service needs a website, but the one she wants is taken, so she starts doing weird things with the domain: alicehomecleaning.com alice-homecleaning.com alice-home-cleaning.com alicehc.com

These domains are also worthless. I can say I would find something like alicehome.services a lot more memorable.

And with advertising, it's more about continuous exposure. So I see the alicehome.services car around town for the third time and say, "oh there's that weird domain again".

> The real issue with the vanity TLDs, and many of the newer ones in general, is that they have zero recognition. They are most worthless.

I disagree. No one was born recognizing .com domains, and outside of the US non-.com domains are already quite mundane. In fact, FANGs like Amazon employ cc TLDs quite extensively to segregate marketplaces, and they do just fine.

> No, the issue with the new/generic/vanity TLDs is that they've lost all meaning.

This assertion has no bearing in reality, specially given domains like .io, .icu, .info, .site, and even .aws.

There's a legitimate engineering issue here too that bears mentioning.

Your marketing team will charge ahead with migrating all your product.business.com sites to just product.business

You'll get half a year into that migration before someone asks about shared domain cookies. Oh, login.business.com dropped an SSO cookie on business.com?

After that you'll get the lovely request - you work with the browser people, can't we just edit the standard to drop a cookie on a TLD?

Discussion on stack overflow... https://stackoverflow.com/questions/3342140/cross-domain-coo...

I would really like a better solution; but that appears to live solidly within a successor to current HTML pages, something designed from inception with security contexts in mind. Maybe they can fix login / logout / credential management too; I really hope they just use kerberos.

> It creates a three-tier world. The big boys who have TLDs, the cheaper boys who have .com

The prophecy didn't come true. Granted the big boys got their .googles and .amazons, but good old .coms are still a thing and not considered "cheaper".

Needs (2011):

http://web.archive.org/web/20110321001950/https://www.temple...

It might even be older according to the main page ("mostly written in 2001")[0]. Sure would be helpful if the author included a publish date. Many blog entries (and stack overflow type solutions) need time context (it may have been the solution/a reasonable position at the time)

https://www.templetons.com/brad/dns/

You're right, it could be much older. I thought 2011 seemed likely because that's the earliest the web archive has and that's the year that the new gTLD program really got rolling. However, it looks from some of his other content that he might have been referring to discussions that were started in 2000 and 2001:

http://archive.icann.org/en/tlds/

You can tell the writing is very old because it mentions .ibm as an example of super well-known and wealthy corporation. No mention of Google nor Amazon.

Given the reference "I wrote a satire of issues around this some years ago" for a piece posted in 2005, I suspect the 2010 / 2011 year is reasonably accurate.

https://ideas.4brad.com/node/221

https://www.templetons.com/brad/dns/sell-english.html

I would tolerate the ICANN TLD money-grabbing scheme, but the "new" TLDs are a little more than an indication of bad taste. I can't be the only one to dislike anything more than 3 characters in a TLD, like .cloud or .engineering.

What’s the problem with longer length tlds?

In practice URLs often cannot be more than 2K characters. It's also more work to type and bytes to transmit and harder to see on small screens. For many use cases these are irreverent, but they do matter to some.

Where is 2K a limit these days?

Chromium based browsers!

> The Google Chrome browser supports a maximum length of a web page URL of 2 MB (2048 characters) in size.

taken from [0]. Do note that this is a restriction created by the browser developers. In theory a URL can be of arbitrary length. Some browsers (like Firefox) do comply with this, though they might now show the full URL to the user, cutting it off after a certain number of characters.

[0]: https://mywebshosting.com/what-is-the-maximum-url-length-lim...

That would be 2KB (the error is in the reference material; you cited it faithfully).

What's interesting is how 2048 characters is 2 KB (a factor of 1000 difference) but this conversion of 2 MB = 2048 characters seems to have been copy-pasted around the web

2MB would be 2 million characters - if using ASCII characters, a bit less for special characters in UTF-8

Experimentally, you can easily verify that 2MB is the actual limit, not 2048 characters.

For example, the following domain renders just fine, despite being ~2076 characters long: https://postman-echo.com/get?foo1=aaaaaaaaaaaaaaaaaaaaaaaaaa...

See the upstream chromium docs: https://chromium.googlesource.com/chromium/src/+/refs/heads/...

I don't think adding ~10 or so characters to the TLD meaningfully impacts this limit.

> In practice URLs often cannot be more than 2K characters. It's also more work to type and bytes to transmit and harder to see on small screens.

The longest vanity TLD I could find was .travelersinsurance. Two dozen characters are nothing more than a rounding error in the 2k character limit. You routinely pass 10 times more characters in path and query parameters than you do in the domain name.

Also, how often do people type in URLs instead of clicking links?

Also, keep in mind that some browsers are starting to hide path and query parameters in the address bar and instead only show the domain name. I presume that makes domain names and TLDs more relevant in terms of name recognition.

As non-native speaker, I always have to google how to spell `engineering`. And you can't really auto-correct domain name.

I agree with this. I feel it takes away from the standardization or uniformity 2 and 3 char TLDs bring to domains. Also it does feel like a bait and switch- where for so many years three character tlds (and then 2 char tlds) were so valuable and sought after, only to have these >3 char tlds come along and change the landscape/market.

Additionally, I always found it interesting how the .xxx TLD never took off (was essentially DOA), even after they were so popular and valueable in the pre-sale auctions icann and others conducted b4 their introduction/release several years ago.

> I feel it takes away from the standardization or uniformity 2 and 3 char TLDs bring to domains.

One of the original TLDs is .arpa, and I don't see any of that uniformity in a world where second-level domains like .co.uk domains are pervasive.

> Additionally, I always found it interesting how the .xxx TLD never took off (was essentially DOA)

The usecase driving .xxx was not from site operators but from third-parties seeking an easier way to censor porn sites.

https://en.wikipedia.org/wiki/.xxx

The concept was flawed from the start given the same site can be served through multiple domains, and it relied on site operators located in any jurisdiction from any corner of the world to voluntarily serve their sites thorough that TLD.

I’m never going to remember a domain with a silly tld. I don’t necessarily think badly of the company when I see a link to one (I mostly think “fuck you ICANN”) but it is anti-memorable.

Oh please. A Cloud Guru's acloud.guru is awesome and easy to remember, and there are probably tons of other examples.

Notion.so is another example. The gTLD in this case could even be considered part or the brand, as notion.com points to it.

.so is a ccTLD, not a gTLD.

And a lot of the .com domain space is polluted either with people sitting on domains and wanting to sell them for thousands or just leaving it unregisterable. So you have this choice of either creating a really long .com, or a shorter .other

Granted, the domain name system is not ideal, but adding new TLDs, however long, is a better solution than dealing with having to name your company so you can get a .com that matches it.

> I’m never going to remember a domain with a silly tld.

I doubt this assertion is grounded in reality. Sites that resort to puns from cc TLDs like lobste.rs don't suffer from name recognition problems, and no one ever had any problem remembering twitch.tv.

ICANN has clearly strayed from its charter of nonprofit root trustee, and is acting like any other for profit entity and arbitraging away their assets. I'd love to see a root-restandardization effort that aims to take anything that isn't (com/net/org, the country code tlds, and whatever other traditional ones I'm forgetting) and sticks it under a new .icann or whatever. So ICANN can "sell .google", but most everyone would end up referring to it as ".google.icann". The sooner this is done the better, to make it clear that attempting to use the root pollution will end up being a painful experience.

Heck, let Big Tech take .google/.apple/.amazon/.comcast etc if that's what it takes to get this done. Point being to keep the root pollution to a finite amount rather than the ongoing sell off of any name imaginable.

And invalidate every URL and hostname in existence? That's an absolute nonstarter of an idea.

If you think through the implications of my proposal you'll see that it is only invalidating hostnames in these ICANN-giveaway new "TLDs". Traditional domains would remain unchanged. Trustees of traditional TLDs would still continue their corruption (as was narrowly avoided for .org a few months ago), but that's probably inevitable. Meanwhile we'd preserve the root namespace for the adoption of better technologies (eg .onion).

And honestly we need better mechanics for machine-intended references regardless. It's ridiculous that you resolve and load a webpage, only for that webpage to require you to resolve a bunch more human readable (ie non-decentralized) names for loading subresources. For example, going to a bookmarked page shouldn't result in any DNS queries for human readable names.

> If you think through the implications of my proposal you'll see that it is only invalidating hostnames in these ICANN-giveaway new "TLDs". Traditional domains would remain unchanged.

And why should that distinction be made? "New gTLDs" have been active since 2013. That's nearly 10 years of usage you're still arguing should be invalidated.

Because it was clear at the time the trustee had gone crazy, and combined with being ugly as hell, they never really got used for anything serious. Rather why should ICANN be able to convert a public resource that had been held in trust, and we just have to accept it?

To be clear I'm not picking on things like .biz etc. Even though those were obvious cash grabs, they're at least widely applicable and thus widely adopted. But things like .christmastrees .business .companyname .morebiglongwords etc. Over "nearly 10 years of usage" I can still count on my fingers the number of times I've seen things like this used.

I have to disagree. I mean, yes, the money-grabbing scheme is probably true, but I don't agree with the part about them being bad taste or too long. I will probably lose at most one hour of my total life having to spend extra time typing > 3 characters. And that's a purposeful overestimation. We have browser history and bookmarks. As far as bad taste, I don't know, I'm not really comfortable with creating this "legitimacy" for .com domains or any 3-letter domain over any of the others.

I think that as more and more companies and individuals get online there's absolutely a use for more TLDs, and having more descriptive ones doesn't hurt anyone. I do think TLDs like .sucks are a flagrant display of ICANN's greed and corruption.

No there just shouldn't be a thing as a "TLD". I should be able to register abc.ou8hurbaskjdbflubsaduf if I want to. Or just 8huasidvbsd with no domain. Only way to fight domain squatting and the stupidly overpriced domain hacks.

I think they're needed so your DNS resolver knows who to talk to. But I agree they're kind of useless from a user point of view.

Who’s the joker who bought http://bad.coffee?

I don't understand why people are interested in keeping the implementation details of the internet exposed. There was a reason we came up with TLDs and such, and you had to type in http vs ftp vs https etc. But there's no reason to be constrained by these detalis. If the computer can figure it out? good.

Whoever invented selling 1 row in a database for massive amounts of money per year was a genius

Modern Papal indulgences.

it's better than that: they're RENTING database rows.

I mean, all domains are really vanity domains. This just gives people more ways to be vain, which is all anyone really wants.

There is already http://microsoft.bing. See https://en.wikipedia.org/wiki/List_of_Internet_top-level_dom....

Which redirects to bing.com, proving that they really have no use for the .bing TLD other than to prevent someone else from grabbing it.

Imagine the clamour of people who want to be associated with Bing

The real problem is can I still type weirdfilename.exe into the browser bar and get to a search engine?

The Crosby estate?

.bing tld is theirs, is not available for the public

and even if it was available to register, they can reserve words for their own usage before launching it

and if you mean, someone else actually doing the paperwork to get the .bing tld for them, it wouldnt ever be approved.

I am reminded that Alphabet's Google's Charleston Road Registry acquired .foo with the promise of sharing it with the community, only to pull a 180 post-grant and keep it to themselves. FeelsBadMan

https://en.wikipedia.org/wiki/.sucks_(registry)

Sorry, but *looks at my nails* .com is a boomer TLD, it's passé. Do you really want a domain from last century? Or do you want an exotic .asia? A cool .club? A forward looking .future? Or even joining the ranks of the celebrity media with a stylish... .xn--45q11c?

The answer is clear, if you want to stay in the past, then go with .com, but the .future is in vanity TLDs. For more information, check out my website https://ok.boomer and considering buying a .boomer domain today!

.xn--45q11c is actually .八卦 (According to Google translate, it means Gossip)

Yes, a perfect TLD for the new Perez Hilton ;)

I bought a LOT of .biz domains, turns out carriers often block people from receiving SMS messages containing email addresses at these TLDs, without notifying me or them.

Probably because they're often used for spam? (I bought em because they were cheap.)

If they're cheap for you, they're cheap for spammers, too.

OMG what if someone buys .thinkdifferent!!!

Reply on news.ycombinator.com