Link shorteners that are outsourced are in my humble opinion an order of magnitude worse for the internet than they would be if in-house.
The way they accrete URLs over time, and then time passes, and anything can happen (your account gets deactivated, shortener company goes out of business / experiences catastrophic data loss / etc) is just asking for permanently broken links which may be unfixable.
Especially when you consider how these services can, unless they're spammy enough to inject gross interstitial ad pages, make 0 revenue from you if you aren't currently paying for a premium service, it's easy to see how they can go under.
Note: Clearly with the kind of ephemeral crap these 2011 tweets were, I'm assuming the 'attacker' publicized "Hey guys look at these funny tweets" because obviously who cares about 15 year old McDonalds marketing messages. But in general it's sad to see these still being used. To be fair to bitly, they do offer the ability to use your own domain at least, but sadly, with a service targeted mainly at marketing departments, usually they don't have the ability to work with the in-house tech team(s) to even set up DNS.
Yeah, but the alternative is everyone having a short domain, which is unfortunately difficult due to the fundamental mathematical principle that a low amount of characters has a very small number of options available.
Everything with finite space is expensive, the more finite the more expensive.
So, while I am in favour of what you say, and I’m in favour of self-governance as much as possible, I understand the challenges.
>Yeah, but the alternative is everyone having a short domain, which is unfortunately difficult due to the fundamental mathematical principle that a low amount of characters has a very small number of options available.
I get where you're coming from but I'm wondering if that's undersells the value-add of link shorteners as literally just shortening the link.
I remember that I liked using link shorteners because they gave me analytics on who clicked my link and where from. I believe paid QR Code link shorteners operate on a similar principle of offering analytics on where and how QR Codes are discovered.
And while in theory, this should be handled by first-party analytics, the users of link shorteners don't necessarily always have control over that analytics platform (e.g., student researchers driving traffic to a survey on an institutional domain).
The original mass market use of link shorteners was to minimize the number of characters needed to link from a Tweet.
Now that Tweets aren't restricted to 140 bytes anymore, there is less need for every single byte. Maybe the difference between sending 1 SMS per message and 2...
I think the problem has always been the shit that gets appended in the query string, like UTM parameters for analytics or other random stuff that has no real impact on the rendering of the page itself.
The other problem was the dawn of the slugified URL where the title of the page was converted into a URL friendly format, and then the page's actual ID appended in it somewhere. The slug was just for SEO because it looked more 'readable' but the only thing responsible for fetching the page was the ID attached to it.
I think the walled-gardening of the internet was more responsible for this than individual webmasters. Half of that tech is useless unless you are running the hosting platform for bloggers or tweeters or whoever, and it was those companies playing the SEO arms race with Google.
The issue wasn't necessarily the stable URLs underneath, but the query parameters being long with characters in them that often broke or confused older systems.
Remember too that we were often dealing with 1024 x 768 or even 800 x 600... 1280 × 1024 was the emerging laptop standard. Long URLs without formatting would often wrap the text display panel.
Those that didn't reformat would have formatting issues. Those that did attempt to wrap excessively long lines on punctuation and hyphens would break the URL.
One very long url in a chat panel on the side of a web game would cause everyone's chat panel to be messed up until that scrolled off.
URL shorteners were (for me) more useful for ephemeral chat where it will scroll off the screen in an hour or so.
They lost their utility when message boards would take really long lines and reformat them (like HN does) so that they fit or would allow for [title](link) style so that the ugly url could be hidden behind a nice title.
> but the alternative is everyone having a short domain
Or just not bother.
Practically no one types a URL like that any more aside from a few biggies like BBC.co.uk, Amazon.com, etc, and even those are usually cached in URL history so they only type the first couple of characters. When were you last given a URL by voice, rather than being given a suggestion to “search for [something-or-other]”? The 140 char limit on Twitter¹ is now 280 making a really short URL less important there, those foolish enough to pay for the site can post much longer missives in one go, and I can't think of another popular site with a stringent artificial post length limit. People looking for your stuff by name will search², not type your address in.
I use a deliberately long name for humour value³. Though I host nothing on there that isn't just for me and friends so this is maybe a special case, I doubt that for many sites a short name makes much real difference these days. People who see my stuff do so by clicking a link, not typing anything in.
--------
[1] the social network desperately trying to be known as “X”
[2] or ask AI which amounts to the same thing usually
[3] subdirectories or subdomains (depeding on my mood at the time) of doubleyoudoubledoubleewe.www.dash.deeohhtee.dash-dash.enneeetee.net
I think there's some signaling and marketing value in a reasonable URL. If I see a really odd name with an unusual TLD that's not somehow relevant or clever I just think "wow, this scammer couldn't even afford a real TLD".
>Practically no one types a URL like that any more aside from a few biggies like BBC.co.uk, Amazon.com, etc, and even those are usually cached in URL
<
in most cases, on my browser history I have to type only the first or first-two letters to get usually directly where I want to go - so you can see, Im a monotonic surfer :-D
Also, my company at the time saw an uptick in account takeovers when HotMail (a popular free email provider) changed their policy to allowing usernames to be recycled. Because email account ownership is basically equivalent to identity proof for most websites, HotMail's policy change effectively cost us a massive customer service workload. (I think the mass liquidity of virtual items out of accounts was more of the direct issue, not so much that users were still active on our site while being inactive on their email account for a year)
I don't think it goes either way in general – it's just whether you're trust the stability of the company creating the link or the third-party (in one case) providing the link service more isn't it?
Bit.ly et al. (even some that have disappeared) must have outlived many a startup that used their services, leaving working links behind (that didn't necessarily go down with the startup if archived, on 'social media' rather than their own site, etc.).
That’s an argument against shorteners though, because if Startup A linked to Website B using Shortener S, simply posting a link to B gives the highest likelihood that the URL will still be valid many years later, as it doesn’t require any party besides B to survive.
If using a shortener at A.com or S.com then yes, depends on whether you think A or S will survive longer (or if S purges old URLs when A abandons the account or stops paying).
Have been trying to advocate for an in-house URL shortener since I joined, but it's not been anything anyone other than myself cares about. These recycled bit.ly IDs might help reignite the discussions.
The idea of link shorteners are fundamentally stupid. We already have a thing where an arbitrary short word when clicked takes you to another website with arbitrary long URL. It’s called hyperlinks. It’s the first thing invented for the Web ever.
We have that in hypertext (and other rich text flavors) but we don't have that in plaintext. That said, offering a URL within a plaintext scenario is less and less of a thing as time marches on.
Yes and you want to personalise the arbitrarily long URL, which means programmatically making short links. And that's a function not everyone wants to program for themselves, ergo link shorteners.
'Fundamentally stupid' is such a god-awful take. Have you ever seen the difference between a QR code where the URL is 15 characters, versus one where it's 45?
Yup and the most common failure mode I see of QRs is that the publisher who printed the code used some sketchy website found by googling “make QR code” that injected its own “shortener” in an attempt to extort them into some stupid subscription service (which has tons of analytics and other value adds but which they probably never wanted). So when I scan a code it goes to some shitty site that says the URL was suspended or whatever.
You can sell the click through data of different links, so theres a bunch of ways for that to be valuable to marketing teams posting to systems they can't track from.
That said it was always obvious this trend was a bad idea that only existed because of a fake limitation, but that's how you can make money from it.
Short anecdote: Famous author (with huge fanbase!) used a Bit.ly link in one of their books (the narrative needed a fake tweet with fake link). Editor didn't notice/care and never registered it. I noticed on release date, caught it and redirected it to a website that referenced the book, a few people tweet about it thinking the author used the link on purpose!
I think the biggest surprise to me is that a URL shortener makes enough money to have a Trust and Safety team. I assumed it was just a side project a person or two ran.
My take is that any shortener without one would voluntarily close doors after being inundated with the most illegal, scammy, and horrid use-cases, and subsequent requests from law enforcement that would far outweigh any ROI.
While abuse of internet facilities does not come as a surprise, I wonder why they would want to shorten those urls? To hide them, probably, but from what? Do you have any insight?
Many average users do not take the time to check where the link is actually going. This could lead to users being targeted for malware, which is very common with shortener redirects.
A common requirement for distributing malware is obtaining domains with a good reputation so that your payloads aren’t auto filtered by security software.
That won't necessarily protect you. Many of the top 10K domains allow user content I assume, so if you don't have your own 24/7 safety department you are relying on theirs.
If you aren't already, make sure you check your whitelist of domains against the PSL¹ to make sure you know where having some trust in a domain definitely doesn't mean that a subdomain of it can be similarly trusted.
Most importantly many of the top 10k domains contain an open redirector, so the abusers will simply give you a 'trusted' top10k domain link like www.example.com/openredirect?target=http://somethinghorible
Google had a popular link-shortening service for a long time before trust and safety started consuming too much staff time, and they killed it. Considering how often link shorteners are used for scams and illegal stuff, you either need a trust and safety team or you're just waiting for the FBI to knock and do it for you.
There is a QR code site that lets you create free QR codes. Unbeknownst to the user, all URLs route through their domain to the destination site.
At some point, it stops redirecting and shows a paywall. If you want that QR code you created and put on physical objects to continue working you have to pay a subscription fee.
Bit.ly isn’t that…but you have understand that the business model is mostly people paying you because they feel they have to and not because they want to.
There are many of these. They are, in my opinion, 100% scams and Google/etc shouldn’t be listing these.
The “some point” tends to be after the unknowing victim has ordered physical media/prints with the code, and the “subscription” is $30/month, annual payment required.
You would be surprised. I've worked with many non-technical people who don't seem to understand that a QR code for their poster etc is just a URL, maybe because they don't fully understand what a URL is.
They invariably turn to these freemium or paid services in search results that inject tracking, ads, or limit the number of referrals before payment is required. In fact this happens so often I set up an internal page for them to make their own QR codes with some customization.
In the longterm, photobucket pulled the same crap. Broke a lot of 10+ year old “Bob’s Ultimate Guide to 1987 Nissan Sentra brake job” guides/howtos on automotive forums.
I believe a lot of their money comes from managing QR codes for non techie companies. then providing them analytics. when and where something gets scanned
The URL shortener in question raised $91.5 million in funding before being acquired by a private equity company for $63 million in 2017. URL shorteners were (are?) big business.
The business is actually quite simple: ads on the clicker side and analytics on the business side. And in terms of compliance it's mostly dealing with issues like fraud and malware links. Very much winner take all type of business. Most link shortener side projects go out of business within 6 months, you can find tons of examples on Reddit.
unfortunately absolutely anything with a publicly available user-generated content - even a stupid link shortener - needs a Trust and Safety team, otherwise it's overwhelmed with... all kind of bad stuff.
There's a lot of abandoned url ownership things on the internet. We need something like a bureau that keeps track of what it was for. We do have archive.is or .org to save the content at a url (side note - are these archivers legal or in some country that allows this?).
I do often come across an article occasionally where someone is writing about a narrow topic that I'm interesting in, but I can't find them any more - maybe they had a vauge user name that isn't resolvable.
You have sites trying to block spammers, but you have over-active groups like Reddit that seem to delete accounts aggressively with no information about what you did wrong. I want a meta id that I own, that people can check I'm the real "foo-bar1418", and that can't be deleted or taken over.
It feels like there's still room for a super-crosssite-ID. But there are also spammers, haters etc that sites have to deal with.
archive.org is a 501(c)(3) non-profit run publicly and generally follows the law.
archive.is is run anonymously, has engaged in ddos campaigns in violation of criminal laws just about everywhere, and has used visitors computers in those campaigns making even visiting the site (when you know of their conduct) a way to violate criminal hacking statutes in most countries.
You might think I’m exaggerating with that URL; I wish I was.
By contrast, HTTPS://QR.EXAMPLE/XYZABC can be the smallest and clearest size, version 1 (21×21) with low error correction (which is normally acceptable at that size).
The short URL will scan more easily and reliably than the long one.
(I’m mildly sad that HTTPS://CHRISMORGAN.INFO/ already completely fills the 25 characters available when you use the alphanumeric set, so my home page is the only version 1 QR code I can use.)
the QR they offer (free today, then you need to pay) is a total scam: but so many professionals cannot spot it until is too late (saw in printed magazines, books ...)
I found it so crazy when people I thought had technical sense do this. I'm always like, you know, QR codes can have the actual URL, right? It's not a magical thing that needs any back end.
Same. It's been marketing and other departments that have claimed to need them for ease of use, but then also that same department wants to shove a lot of long text in the url for SEO...
I've warned several times over the years that these links will rot and our going to be out of our control, but... that's what they want!
These days you can buy a short domain easily and create your own short links.
Right now there are around 540 TLDs that anyone can use (not brands or country restrictions) and most are really cheap and haven't been bought out yet.
> Right now there are around 540 TLDs that anyone can use
Be careful if you care about visits from those using corporate networks, many arbitrarily block uncommon TLDs¹ unless users explicitly ask for the restriction to be dropped².
--------
[1] The reason given being that links on them are often phishing and related crap as they are real useful links
[2] and even then an assessment is made, the drop is not automatic and your site alone is unlikely to be deemed sufficient reason
I'm surprised there isn't a DNS standard or standardized API for link shorteners. You'd think we'd have a SHORTN record type that works like SRV or TXT with a standard format to verify where the short link is supposed to point. For example:
short.name. 300 IN SHORTN 0 5 /foo https://some.long.name/whatever/foo
This reminded me of the fact that bit.ly links used to be (maybe still are?) case sensitive. I got a bill in the mail that suggested to pay I could got to https://bit.ly/SomeCompanyPayBillOnline so naturally I tried registering bit.ly/somecompanypaybillonline and sure enough, those were considered separate shortlinks. I suspect most that when typing in a URL, most people assume it's case insensitive, so this was probably pretty effective.
They probably should never register links in all lowercase, all uppercase or capitalized, but being case sensitive is a major reason of their shortness
This sort of thing is not rare; link rot is how a lot of websites get compromised. I recently had to clean up the Shopify site for a client that's been on the platform for quite a few years. The site was occasionally redirecting visitors to all sorts of shady domains.
The cause was an old Shopify app, which pulled resources from an external domain that had expired and been reregistered.
I’ve seen it on WordPress plenty of times, but I kind of thought that Shopify apps were held to a higher security standard.
Globally unique human friendly names are more trouble than they're worth. Sooner or later their referential integrity falls to corruption or neglect. We'd be better of with cryptographic hashes on the wire and user configurable pet names in the UI.
And there's not a particularly human-friendly aspect to shortened URLs. They're short, so they fit into human- and machine-related length constraints, but the short arbitrary string is hardly human-friendly compared to a good vanity URL consisting of actual words that are slightly longer.
They shouldn't be but most people will just paste a link in their message without using markdown ()[] or equivalent formatting. Big links mess with the flow of paragraphs
Maybe we need smarter text handling so that all links are short links on view, but the full data is preserved whenever the data moves (e.g. copy/paste)
This is a very interesting idea. Sounds a bit like word-wrapping logic: the line breaks do not exist in the storage/transmission/editorial sense, but they exist at the presentation layer. And just as you can quite easily discover soft wrapping (by adjusting the size of the window, in some cases), the user should be able to easily discover the full URL (maybe hover / long press).
Slack essentially does this nowadays, hiding the scheme and truncating the rest. Even address bars do it a little bit. So why not bake it into OS toolkits (the way word-wrapping is) to eliminate the need for each app to reinvent this wheel?
I think it could be extended to do quite a bit more than just shorten. For instance, if I suggest that you download and install QmdfTbBqBPQ7VNxZEYEj14VmRuZBkqFbiwReogJgS1zR1n (you'd see this not as a long string but as some OS-controlled visual for a reference to arbitrary data) and you've got peers which know that data to be harmful in some way... misinformation, malware, whatever... your OS could make it glow in a certain color and optionally present you with more info.
This would be a bit like the padlock that appears in a browser, but instead of validating merely that a server is not lying its domain name this would attempt to surface info about the trustworthiness of the data that you'll find on the other end of the link.
Rather than "don't click any links in emails" it could be "configure your client to trust IT's annotation feed, and we'll make known-to-be-safe links glow green, and known-harmful links glow red. So as soon as one person in the org flags a link as suspected-harmful, UI's everywhere now include a warning.
Works OK for URL's, but even better in a hash-linked content because your OS can verify the payload matches its hash before rendering it in any way. This would eliminate a very prominent category of skulduggery.
Somewhat related, I've been getting a lot of spam with links to urls of big, reputable, organizations, such as universities.
You'd think that the organization was hacked, but in all these case instead these were old domains that used CNAMEs.
The domain that the CNAME pointed to was once owned and used by the organization, but at some point it was let go, and it ended up being purchased by the spammers.
So, the spammers were able to use links that looked safe, and maybe that also helped with passing spam filters.
Thus, if you manage domains, please keep track of the ones with CNAMEs, recommend the users to tell you when they stop using them, and maybe check every once in a while (better if automatically) that the domain they're pointing to hasn't changed hands.
I'd be curious to know how the spammers find these old cname domains, by the way.
I had this happen to me. I had a CNAME pointing into a domain I let lapse, and some criminal bought up the domain name and used it for malware with my CNAME.
There were probably crawlable HTML links to the CNAME, in my case. It neve had a TLS cert, so it wouldn't have been in Certificate Transparency logs, and the DNS server doesn't allow AXFR to anybody but the actual secondary.
This is likely but not certainly on Bitly: They claim that their shortener behaves in the sensible way (links are permanent and URLs not reused), but it doesn't seem to.
The only way for this not to be primarily Bitly's fault would be if the links were in accounts that allow links to be edited (as I understand it, that only applies to paid accounts) and someone hijacked the account using stolen creds.
The "article" mentions "At 09:18 UTC another user described the method in a post that has been viewed more than 46,000 times." but doesn't link the post or explain it (unless I missed the explanation in the slop sea)...
if the original people had kept their accounts active they would still be the original URL. That's on BK, Trump, etc.
Obviously, them's the breaks though when you voluntarily rent URLs from random companies. Bitly could sell their whole domain to (insert whatever organization you trust the least) and every URL you've ever made either breaks or is redirected to whatever that entity wants.
My company use a shortener to reduce how much data gets encoded into QR codes that we use, thus making them work better in marginal lighting conditions with worse phones. However, it is entirely self-hosted (301/302 redirect is not a difficult service to build) to protect us from this exact kind of problem.
Traditionally they were useful at the height of Twitter's heyday, because it used to limit Tweets to 140 characters. SMS is a similar use case, but I disliked it there even more, because it normalised getting official messages from random shortcodes with shortened links -- just like a phishing campaign.
You certainly can, I'd happily set one up for my company if they'd let me, but the marketing organizations that generally use these things are 99% of the time only distant cousins from people like me in the orgchart family tree. They just go sign up accounts, use a marketing credit card, and do whatever, and they only talk to people in the tech side of the company after some mess or disaster is made (e.g. "Oh no, we used a URL shortener that went out of business/sold to a porn site/etc")
I don't need shorter links per-se but I need a way to track the effectivness of the links I send out to the public for marketing purposes. I could put tracking ids and such but it doesn't look as good.
I also have links to my Google reviews, Google Maps etc and it's better show bitly.com/business-google-reviews and bitly.com/business-google/map than https://maps.app.goo.gl/sGa1KAU2131212cqU7
You want a way, and I can assure you that normal links look better, especially if the tracking part is easier to spot and can be removed by those who want to (who will be a minority, so your statistics won't be affected too much).
The other replies have largely covered most use cases they can also be handy for less tech savvy people to quickly generate links with unique analytics for marketing purposes.
Before Twitter, they were used to put URLs on physical things like business cards. They increased in popularity once Twitter imposed a 140 character limit.
People then saw that there were lots of secondary benefits. As another comment mentioned, hiding affiliate forwarders / query strings were a huge use back when everyone was trying to score those kickbacks. And then getting metrics on every clickthrough.
Back in the day every useful link you found on forums used to be shortened with an advertisment proxy service that promised a few pennies per click in exchange for making the user sit through a minute of ads.
The tweet that McDonald's Sweden is proud of what they've achieved, redirected to the Wikipedia page for 9/11, is probably the most quintessentially 'Gen Z' joke to the point that dictionaries ought to use it as the example.
I'm guessing some combo of dark humor, and the fact that they don't remember 9/11 so they don't mind joking about it, plus ironically trolling big corporations.
Pete Davidson, millennial of my vintage, to Justin Bieber in 2015:
>"Justin, you know I lost my dad on 9/11, and I always regretted growing up without a dad... until I met your dad, Justin. Now I'm glad mine's dead."
>And now for the greatest transition in the history of comedy. Two people from the movie Soul Plane are here, right? Kevin, Snoop—Soul Plane was the worst experience of my life involving a plane.
tribute.wmv was made by a millennial in 2002 and trolling big corporations on the internet has been a thing as long big corporations have been on the internet to troll. None of those are even vaguely genz traits.
Basically, bit-rot-ly missed up by deleting old link names when deleting accounts; people now register old links and redirect them to wherever they want.
Isn't Twitter's original "feature" or short "content" how URL "shorteners" (dang technology making us say ridiculous non-words with a straight face) came about? With that framing, this is pretty funny, at least to me. Well ok, I find this very funny regardless of how we got here.
A bigger issue perhaps than some old links now redirecting to scams is large scale link rot. The scam problem does already exist for normal links when the domain is not renewed and the new owner points it to a shady website. But link rot is a wider problem: if one link shortener ceases operations, a huge number of links will just no longer work, and it will usually no longer be possible to find out what content they were linking to, because you can't guess it from the domain name and URL slug as you would from a normal dead URL.
The convenience can so a large extend be blamed on people not doing proper URLs. Typing bit.ly/qwerty is easier than typing company.com/pages/1234567/november_marketing_pre_black_friday?utm_page=123&utm_source=app but it also lose all context and no marketing department would want to go without adding a metric ton of UTM tags to that URL because internal tooling either doesn't exist or they don't know how to use it.
From previous jobs I can safely say that most people in marketing have zero knowledge about how the internet and the web works, and they also don't care.
There's nothing preventing a company from running a URL "shortener" on their website directly. It could be as simple as "company.com/campaign/qwerty/" which then redirects the same way bitly will do. That was our solution, just namespace everything, e.g. /product/<sku>, /search/?query= and /campaign/<id>. It works every bit as well, has the same features and was 100% within our control.
What is the convenience, really? In 2026, we're rarely speaking/reading URL's. Anything on a print ad can be a QR code. Where that doesn't work, you can say, "go to <company> dot com and search for <product>", like how people used to advertise AOL keywords. Or they can put a link to their active promotions above the fold on their home page.
The main convenience seems to be for pranks, plus it's just another vulnerability as described in the article. At best, they're just something else that has to be maintained forever.
There's a lower chance of a link being truncated if it's being forwarded, or copied/pasted or referred into some way.
Also, including named short links in a url is a friendly way to put a link on printed materials, email, etc.
There's lots of relevant uses for link shorteners. Maintaining it forever is for sure a consideration, it's why running one from your own domain lets you own the 301 redirect permanently.
Maybe this isn't the case for the majority of podcast listeners, but I've never had to hear and type a link from a podcast if that's what you mean. Don't most places you hear a podcast also have just enough UI to hold producer supplied text/show notes?
Some of the link shortener projects have some additional features that look useful, but still, running something directly on cloudfront/flare seems hard to overlook.
Obviously should probably never use this in the |bash situation, but it's `-L` -- in general a good one to use by default unless you're only using curl to inspect redirect headers or something.
> We tested the method on links nobody disputes. A link whose creator says he made it in 2009 shows a 2009 creation date. A custom link tweeted by a YouTube channel on 21 June 2017 shows 20 June 2017. Among the Trump links that are still original, 32 of 40 were created on the day of the tweet or the day before.
Link shorteners that are outsourced are in my humble opinion an order of magnitude worse for the internet than they would be if in-house.
The way they accrete URLs over time, and then time passes, and anything can happen (your account gets deactivated, shortener company goes out of business / experiences catastrophic data loss / etc) is just asking for permanently broken links which may be unfixable.
Especially when you consider how these services can, unless they're spammy enough to inject gross interstitial ad pages, make 0 revenue from you if you aren't currently paying for a premium service, it's easy to see how they can go under.
Note: Clearly with the kind of ephemeral crap these 2011 tweets were, I'm assuming the 'attacker' publicized "Hey guys look at these funny tweets" because obviously who cares about 15 year old McDonalds marketing messages. But in general it's sad to see these still being used. To be fair to bitly, they do offer the ability to use your own domain at least, but sadly, with a service targeted mainly at marketing departments, usually they don't have the ability to work with the in-house tech team(s) to even set up DNS.
Yeah, but the alternative is everyone having a short domain, which is unfortunately difficult due to the fundamental mathematical principle that a low amount of characters has a very small number of options available.
Everything with finite space is expensive, the more finite the more expensive.
So, while I am in favour of what you say, and I’m in favour of self-governance as much as possible, I understand the challenges.
>Yeah, but the alternative is everyone having a short domain, which is unfortunately difficult due to the fundamental mathematical principle that a low amount of characters has a very small number of options available.
I get where you're coming from but I'm wondering if that's undersells the value-add of link shorteners as literally just shortening the link.
I remember that I liked using link shorteners because they gave me analytics on who clicked my link and where from. I believe paid QR Code link shorteners operate on a similar principle of offering analytics on where and how QR Codes are discovered.
And while in theory, this should be handled by first-party analytics, the users of link shorteners don't necessarily always have control over that analytics platform (e.g., student researchers driving traffic to a survey on an institutional domain).
> but the alternative is everyone having a short domain
a typical domain is not that long (but a typical URL is), one can have a short-url service in the same domain https://example.com/r/Yx9FH -> https://example.com/a-typically-long-url
^agree. Yeah it’s not that important that the domain is shockingly short anymore. Just use something like go.(mycompany).com/slug
Your audience can type your whole domain name, it’ll be fine.
The alternative is typing or scanning the actual address.
The original mass market use of link shorteners was to minimize the number of characters needed to link from a Tweet.
Now that Tweets aren't restricted to 140 bytes anymore, there is less need for every single byte. Maybe the difference between sending 1 SMS per message and 2...
> The original mass market use of link shorteners was to minimize the number of characters needed to link from a Tweet.
That's an anachronism. URL shorteners predate Twitter by several years. (Even the Wikipedia article for URL shorteners is older than Twitter.)
Yes, they existed, but were they really that popular before that use?
They were also useful for links in forums and other similar sites where links could get broken by parsers or wrapped inadvertently.
Also the gold standard for sharing long, ugly URLs in IRC.
In my circles using a public url shortener was considered more rude than posting a long url.
Sites should be encouraged to have shorter, more stable URLs.
I think the problem has always been the shit that gets appended in the query string, like UTM parameters for analytics or other random stuff that has no real impact on the rendering of the page itself.
The other problem was the dawn of the slugified URL where the title of the page was converted into a URL friendly format, and then the page's actual ID appended in it somewhere. The slug was just for SEO because it looked more 'readable' but the only thing responsible for fetching the page was the ID attached to it.
I think the walled-gardening of the internet was more responsible for this than individual webmasters. Half of that tech is useless unless you are running the hosting platform for bloggers or tweeters or whoever, and it was those companies playing the SEO arms race with Google.
The issue wasn't necessarily the stable URLs underneath, but the query parameters being long with characters in them that often broke or confused older systems.
Remember too that we were often dealing with 1024 x 768 or even 800 x 600... 1280 × 1024 was the emerging laptop standard. Long URLs without formatting would often wrap the text display panel.
Those that didn't reformat would have formatting issues. Those that did attempt to wrap excessively long lines on punctuation and hyphens would break the URL.
One very long url in a chat panel on the side of a web game would cause everyone's chat panel to be messed up until that scrolled off.
URL shorteners were (for me) more useful for ephemeral chat where it will scroll off the screen in an hour or so.
They lost their utility when message boards would take really long lines and reformat them (like HN does) so that they fit or would allow for [title](link) style so that the ugly url could be hidden behind a nice title.
How about https://ha.mr/ ?
> but the alternative is everyone having a short domain
Or just not bother.
Practically no one types a URL like that any more aside from a few biggies like BBC.co.uk, Amazon.com, etc, and even those are usually cached in URL history so they only type the first couple of characters. When were you last given a URL by voice, rather than being given a suggestion to “search for [something-or-other]”? The 140 char limit on Twitter¹ is now 280 making a really short URL less important there, those foolish enough to pay for the site can post much longer missives in one go, and I can't think of another popular site with a stringent artificial post length limit. People looking for your stuff by name will search², not type your address in.
I use a deliberately long name for humour value³. Though I host nothing on there that isn't just for me and friends so this is maybe a special case, I doubt that for many sites a short name makes much real difference these days. People who see my stuff do so by clicking a link, not typing anything in.
--------
[1] the social network desperately trying to be known as “X”
[2] or ask AI which amounts to the same thing usually
[3] subdirectories or subdomains (depeding on my mood at the time) of doubleyoudoubledoubleewe.www.dash.deeohhtee.dash-dash.enneeetee.net
I think there's some signaling and marketing value in a reasonable URL. If I see a really odd name with an unusual TLD that's not somehow relevant or clever I just think "wow, this scammer couldn't even afford a real TLD".
>Practically no one types a URL like that any more aside from a few biggies like BBC.co.uk, Amazon.com, etc, and even those are usually cached in URL <
in most cases, on my browser history I have to type only the first or first-two letters to get usually directly where I want to go - so you can see, Im a monotonic surfer :-D
This comment is where the risk is.
Also, my company at the time saw an uptick in account takeovers when HotMail (a popular free email provider) changed their policy to allowing usernames to be recycled. Because email account ownership is basically equivalent to identity proof for most websites, HotMail's policy change effectively cost us a massive customer service workload. (I think the mass liquidity of virtual items out of accounts was more of the direct issue, not so much that users were still active on our site while being inactive on their email account for a year)
I don't think it goes either way in general – it's just whether you're trust the stability of the company creating the link or the third-party (in one case) providing the link service more isn't it?
Bit.ly et al. (even some that have disappeared) must have outlived many a startup that used their services, leaving working links behind (that didn't necessarily go down with the startup if archived, on 'social media' rather than their own site, etc.).
That’s an argument against shorteners though, because if Startup A linked to Website B using Shortener S, simply posting a link to B gives the highest likelihood that the URL will still be valid many years later, as it doesn’t require any party besides B to survive.
If using a shortener at A.com or S.com then yes, depends on whether you think A or S will survive longer (or if S purges old URLs when A abandons the account or stops paying).
Have been trying to advocate for an in-house URL shortener since I joined, but it's not been anything anyone other than myself cares about. These recycled bit.ly IDs might help reignite the discussions.
The idea of link shorteners are fundamentally stupid. We already have a thing where an arbitrary short word when clicked takes you to another website with arbitrary long URL. It’s called hyperlinks. It’s the first thing invented for the Web ever.
We have that in hypertext (and other rich text flavors) but we don't have that in plaintext. That said, offering a URL within a plaintext scenario is less and less of a thing as time marches on.
Yes and you want to personalise the arbitrarily long URL, which means programmatically making short links. And that's a function not everyone wants to program for themselves, ergo link shorteners.
'Fundamentally stupid' is such a god-awful take. Have you ever seen the difference between a QR code where the URL is 15 characters, versus one where it's 45?
If anyone is curious, it makes the QR-code go from 21x21 to 25x25 squares.
I struggle to think of any case where that would make any significant difference.
Yup and the most common failure mode I see of QRs is that the publisher who printed the code used some sketchy website found by googling “make QR code” that injected its own “shortener” in an attempt to extort them into some stupid subscription service (which has tons of analytics and other value adds but which they probably never wanted). So when I scan a code it goes to some shitty site that says the URL was suspended or whatever.
>paying for a premium service,
I was always questioning how any of them would make any money as I do not see any commercial usecase?
You can sell the click through data of different links, so theres a bunch of ways for that to be valuable to marketing teams posting to systems they can't track from.
That said it was always obvious this trend was a bad idea that only existed because of a fake limitation, but that's how you can make money from it.
Short anecdote: Famous author (with huge fanbase!) used a Bit.ly link in one of their books (the narrative needed a fake tweet with fake link). Editor didn't notice/care and never registered it. I noticed on release date, caught it and redirected it to a website that referenced the book, a few people tweet about it thinking the author used the link on purpose!
that's very kind of you! the wrong person noticing this could've done some really nasty stuff...
Good on you!
Just say the name of the author bro...
I think the biggest surprise to me is that a URL shortener makes enough money to have a Trust and Safety team. I assumed it was just a side project a person or two ran.
My take is that any shortener without one would voluntarily close doors after being inundated with the most illegal, scammy, and horrid use-cases, and subsequent requests from law enforcement that would far outweigh any ROI.
As someone who was running such a service for a brief period of time, parent is 100% correct.
While abuse of internet facilities does not come as a surprise, I wonder why they would want to shorten those urls? To hide them, probably, but from what? Do you have any insight?
Many average users do not take the time to check where the link is actually going. This could lead to users being targeted for malware, which is very common with shortener redirects.
A common requirement for distributing malware is obtaining domains with a good reputation so that your payloads aren’t auto filtered by security software.
As with the sibling comment, it doesn't surprise me. But what's the legal basis for liability for offering a service that simply expands links?
I am about to launch a shortener. And I only allow redirection to the top 10k domains.
That won't necessarily protect you. Many of the top 10K domains allow user content I assume, so if you don't have your own 24/7 safety department you are relying on theirs.
If you aren't already, make sure you check your whitelist of domains against the PSL¹ to make sure you know where having some trust in a domain definitely doesn't mean that a subdomain of it can be similarly trusted.
--------
[1] https://en.wikipedia.org/wiki/Public_Suffix_List
Most importantly many of the top 10k domains contain an open redirector, so the abusers will simply give you a 'trusted' top10k domain link like www.example.com/openredirect?target=http://somethinghorible
And of course PSL won’t fully protect you either —- sites like GitHub.com have tons of malware on their main domain
Google had a popular link-shortening service for a long time before trust and safety started consuming too much staff time, and they killed it. Considering how often link shorteners are used for scams and illegal stuff, you either need a trust and safety team or you're just waiting for the FBI to knock and do it for you.
There is a QR code site that lets you create free QR codes. Unbeknownst to the user, all URLs route through their domain to the destination site.
At some point, it stops redirecting and shows a paywall. If you want that QR code you created and put on physical objects to continue working you have to pay a subscription fee.
Bit.ly isn’t that…but you have understand that the business model is mostly people paying you because they feel they have to and not because they want to.
There are many of these. They are, in my opinion, 100% scams and Google/etc shouldn’t be listing these.
The “some point” tends to be after the unknowing victim has ordered physical media/prints with the code, and the “subscription” is $30/month, annual payment required.
Wow that is sketch.
I bet that they're doing going to do it is very carefully buried in the T&Cs. Probably in 1pt font.
That's so stupid. You can just right-click on a page in Chrome and click "Create QR code for this page" and Google will hand you one for free as well.
You would be surprised. I've worked with many non-technical people who don't seem to understand that a QR code for their poster etc is just a URL, maybe because they don't fully understand what a URL is.
They invariably turn to these freemium or paid services in search results that inject tracking, ads, or limit the number of referrals before payment is required. In fact this happens so often I set up an internal page for them to make their own QR codes with some customization.
QR codes have their own mythology for many, even though it’s just a URL (in most cases) somehow being QR makes it better.
I’ve even had people suggest a QR code instead of a link - people browsing can take out their phone! I made the image clickable.
You might want to look at a site named getqr [dot] com. And then look at the number of reviews they have on Trustpilot.
If you're making a lot of money with the scam you can spend a lot of money on SEO, ads etc. to acquire "users" (victims)
QR codes might be an antipattern for physical-online links given:
In the longterm, photobucket pulled the same crap. Broke a lot of 10+ year old “Bob’s Ultimate Guide to 1987 Nissan Sentra brake job” guides/howtos on automotive forums.
Bob can’t even login and pay because Bob is dead.
https://en.wikipedia.org/wiki/Photobucket
That's an incredibly brilliant business model, given that brilliance is solely and exclusively measured by profits generated.
name shame please
A popular online business card print shop does this.
bit.ly has 300-400 employees & its last public statement on revenue (2023) stated they had >$100M ARR.
Analytics, for marketing campaigns commissioned by businesses who do not know any better.
what? how do they make that much money
They’re used heavily by marketing and demand gen teams
I believe a lot of their money comes from managing QR codes for non techie companies. then providing them analytics. when and where something gets scanned
The URL shortener in question raised $91.5 million in funding before being acquired by a private equity company for $63 million in 2017. URL shorteners were (are?) big business.
Having emailed bitly a few dozen times, all of my replies were from one guy, most of them telling me to GFY.
The business is actually quite simple: ads on the clicker side and analytics on the business side. And in terms of compliance it's mostly dealing with issues like fraud and malware links. Very much winner take all type of business. Most link shortener side projects go out of business within 6 months, you can find tons of examples on Reddit.
They can also accumulate a fair amount of data to sell (IP x was interested in y at time z)
unfortunately absolutely anything with a publicly available user-generated content - even a stupid link shortener - needs a Trust and Safety team, otherwise it's overwhelmed with... all kind of bad stuff.
There's a lot of abandoned url ownership things on the internet. We need something like a bureau that keeps track of what it was for. We do have archive.is or .org to save the content at a url (side note - are these archivers legal or in some country that allows this?).
I do often come across an article occasionally where someone is writing about a narrow topic that I'm interesting in, but I can't find them any more - maybe they had a vauge user name that isn't resolvable.
You have sites trying to block spammers, but you have over-active groups like Reddit that seem to delete accounts aggressively with no information about what you did wrong. I want a meta id that I own, that people can check I'm the real "foo-bar1418", and that can't be deleted or taken over.
It feels like there's still room for a super-crosssite-ID. But there are also spammers, haters etc that sites have to deal with.
archive.org is a 501(c)(3) non-profit run publicly and generally follows the law.
archive.is is run anonymously, has engaged in ddos campaigns in violation of criminal laws just about everywhere, and has used visitors computers in those campaigns making even visiting the site (when you know of their conduct) a way to violate criminal hacking statutes in most countries.
and changed links to point to malware (https://news.ycombinator.com/item?id=50275373)
who the heck downvoted this twice?
this is why i have always been against link shorteners :)
I’m also against QR code services which don’t just generate a direct URL to something you own.
Unlike link shorteners, the URL length barely matters, and the codes often get permanently printed onto posters, tables, art, etc.
> the URL length barely matters
Consider https://www.somebusiness.example/multiple-long/parent-direct.... This will be a version 9 code (53×53) with medium error correction. (You could also make it version 8 (49×49) with low error correction, but that’s riskier.)
You might think I’m exaggerating with that URL; I wish I was.
By contrast, HTTPS://QR.EXAMPLE/XYZABC can be the smallest and clearest size, version 1 (21×21) with low error correction (which is normally acceptable at that size).
The short URL will scan more easily and reliably than the long one.
(I’m mildly sad that HTTPS://CHRISMORGAN.INFO/ already completely fills the 25 characters available when you use the alphanumeric set, so my home page is the only version 1 QR code I can use.)
yes
the QR they offer (free today, then you need to pay) is a total scam: but so many professionals cannot spot it until is too late (saw in printed magazines, books ...)
Have a look at trustpilot reviews
I found it so crazy when people I thought had technical sense do this. I'm always like, you know, QR codes can have the actual URL, right? It's not a magical thing that needs any back end.
Same. It's been marketing and other departments that have claimed to need them for ease of use, but then also that same department wants to shove a lot of long text in the url for SEO...
I've warned several times over the years that these links will rot and our going to be out of our control, but... that's what they want!
To be fair, the way some companies use single use / campaign specific domain names, they will rot too once it doesn't make sense to renew them.
These days you can buy a short domain easily and create your own short links.
Right now there are around 540 TLDs that anyone can use (not brands or country restrictions) and most are really cheap and haven't been bought out yet.
> Right now there are around 540 TLDs that anyone can use
Be careful if you care about visits from those using corporate networks, many arbitrarily block uncommon TLDs¹ unless users explicitly ask for the restriction to be dropped².
--------
[1] The reason given being that links on them are often phishing and related crap as they are real useful links
[2] and even then an assessment is made, the drop is not automatic and your site alone is unlikely to be deemed sufficient reason
I personally only use .com but that's a great point! I've been looking into uncommon TLDs for short referral links.
The classic discovery every startup with a cool domain runs into when they start selling to enterprises
I'm surprised there isn't a DNS standard or standardized API for link shorteners. You'd think we'd have a SHORTN record type that works like SRV or TXT with a standard format to verify where the short link is supposed to point. For example:
...could tell a browser/app/whatever that http(s)://short.name/foo is supposed to point to https://some.long.name/whatever/fooAt the very least, we could have a standard shortener API call that clients could use to verify before allowing the link redirect.
This is why I'm suddenly in favour of link shorteners.
Ah yes the duality of man
Indeed this should be a surprise to exactly nobody.
This reminded me of the fact that bit.ly links used to be (maybe still are?) case sensitive. I got a bill in the mail that suggested to pay I could got to https://bit.ly/SomeCompanyPayBillOnline so naturally I tried registering bit.ly/somecompanypaybillonline and sure enough, those were considered separate shortlinks. I suspect most that when typing in a URL, most people assume it's case insensitive, so this was probably pretty effective.
IIRC the spec says that the path part is case sensitive.
What the spec says is quite irrelevant for this because people are going to human.
They probably should never register links in all lowercase, all uppercase or capitalized, but being case sensitive is a major reason of their shortness
This sort of thing is not rare; link rot is how a lot of websites get compromised. I recently had to clean up the Shopify site for a client that's been on the platform for quite a few years. The site was occasionally redirecting visitors to all sorts of shady domains.
The cause was an old Shopify app, which pulled resources from an external domain that had expired and been reregistered.
I’ve seen it on WordPress plenty of times, but I kind of thought that Shopify apps were held to a higher security standard.
Mine was hijacked by bitly itself, I have a username "username" and they just decided to delete it without any notice and make it unavailable.
Globally unique human friendly names are more trouble than they're worth. Sooner or later their referential integrity falls to corruption or neglect. We'd be better of with cryptographic hashes on the wire and user configurable pet names in the UI.
And there's not a particularly human-friendly aspect to shortened URLs. They're short, so they fit into human- and machine-related length constraints, but the short arbitrary string is hardly human-friendly compared to a good vanity URL consisting of actual words that are slightly longer.
They shouldn't be but most people will just paste a link in their message without using markdown ()[] or equivalent formatting. Big links mess with the flow of paragraphs
Maybe we need smarter text handling so that all links are short links on view, but the full data is preserved whenever the data moves (e.g. copy/paste)
This is a very interesting idea. Sounds a bit like word-wrapping logic: the line breaks do not exist in the storage/transmission/editorial sense, but they exist at the presentation layer. And just as you can quite easily discover soft wrapping (by adjusting the size of the window, in some cases), the user should be able to easily discover the full URL (maybe hover / long press).
Slack essentially does this nowadays, hiding the scheme and truncating the rest. Even address bars do it a little bit. So why not bake it into OS toolkits (the way word-wrapping is) to eliminate the need for each app to reinvent this wheel?
I think it could be extended to do quite a bit more than just shorten. For instance, if I suggest that you download and install QmdfTbBqBPQ7VNxZEYEj14VmRuZBkqFbiwReogJgS1zR1n (you'd see this not as a long string but as some OS-controlled visual for a reference to arbitrary data) and you've got peers which know that data to be harmful in some way... misinformation, malware, whatever... your OS could make it glow in a certain color and optionally present you with more info.
This would be a bit like the padlock that appears in a browser, but instead of validating merely that a server is not lying its domain name this would attempt to surface info about the trustworthiness of the data that you'll find on the other end of the link.
Rather than "don't click any links in emails" it could be "configure your client to trust IT's annotation feed, and we'll make known-to-be-safe links glow green, and known-harmful links glow red. So as soon as one person in the org flags a link as suspected-harmful, UI's everywhere now include a warning.
Works OK for URL's, but even better in a hash-linked content because your OS can verify the payload matches its hash before rendering it in any way. This would eliminate a very prominent category of skulduggery.
Somewhat related, I've been getting a lot of spam with links to urls of big, reputable, organizations, such as universities.
You'd think that the organization was hacked, but in all these case instead these were old domains that used CNAMEs.
The domain that the CNAME pointed to was once owned and used by the organization, but at some point it was let go, and it ended up being purchased by the spammers.
So, the spammers were able to use links that looked safe, and maybe that also helped with passing spam filters.
Thus, if you manage domains, please keep track of the ones with CNAMEs, recommend the users to tell you when they stop using them, and maybe check every once in a while (better if automatically) that the domain they're pointing to hasn't changed hands.
I'd be curious to know how the spammers find these old cname domains, by the way.
I had this happen to me. I had a CNAME pointing into a domain I let lapse, and some criminal bought up the domain name and used it for malware with my CNAME.
There were probably crawlable HTML links to the CNAME, in my case. It neve had a TLS cert, so it wouldn't have been in Certificate Transparency logs, and the DNS server doesn't allow AXFR to anybody but the actual secondary.
Yeah, what struck me is that my spammers seem to have picked deliberately reputable domains
I guess that they either looked directly for them intentionally, or there's some list of them you can buy in spamners networks.
CT Logs is pretty useful to find things. https://subdomainfinder.c99.nl is a useful source. The attack is often called 'Dangling DNS'.
Oh that's a great service I didn't know
Although it's probably not what my spammers used, it doesn't find a subdomain they used today
This is likely but not certainly on Bitly: They claim that their shortener behaves in the sensible way (links are permanent and URLs not reused), but it doesn't seem to.
The only way for this not to be primarily Bitly's fault would be if the links were in accounts that allow links to be edited (as I understand it, that only applies to paid accounts) and someone hijacked the account using stolen creds.
The "article" mentions "At 09:18 UTC another user described the method in a post that has been viewed more than 46,000 times." but doesn't link the post or explain it (unless I missed the explanation in the slop sea)...
Isn't this just a subcase of people thinking all links are diamonds (permanent and forever)?
most examples have been taken down by now but here are two still left
https://x.com/realdonaldtrump/status/228935671061020673?a=87...
https://x.com/realdonaldtrump/status/350628162595590148?a=Jk...
these were also taken down, unfortunately
Unfortunately? Are part of the team hijacking the links? Taking them down is a good thing
^ The No Fun Patrol here to stand with the corporate machine against harmless hijinks.
No, I'm just tired of people ruining for the rest of us.
I remember when link shortening services were a thing. I built my own around 2008 due to Twitter content limits.
Thats really bad nowadays with the help of AI many big websites are taken down
>"and is disabling the affected links."
Wouldn't it make more sense to restore the original destination URL?
if the original people had kept their accounts active they would still be the original URL. That's on BK, Trump, etc.
Obviously, them's the breaks though when you voluntarily rent URLs from random companies. Bitly could sell their whole domain to (insert whatever organization you trust the least) and every URL you've ever made either breaks or is redirected to whatever that entity wants.
What's the value proposition of link shorteners? I mean, why do companies, or customers, use them?
My company use a shortener to reduce how much data gets encoded into QR codes that we use, thus making them work better in marginal lighting conditions with worse phones. However, it is entirely self-hosted (301/302 redirect is not a difficult service to build) to protect us from this exact kind of problem.
Traditionally they were useful at the height of Twitter's heyday, because it used to limit Tweets to 140 characters. SMS is a similar use case, but I disliked it there even more, because it normalised getting official messages from random shortcodes with shortened links -- just like a phishing campaign.
Can’t you just run that internally on a one character subdomain?
You certainly can, I'd happily set one up for my company if they'd let me, but the marketing organizations that generally use these things are 99% of the time only distant cousins from people like me in the orgchart family tree. They just go sign up accounts, use a marketing credit card, and do whatever, and they only talk to people in the tech side of the company after some mess or disaster is made (e.g. "Oh no, we used a URL shortener that went out of business/sold to a porn site/etc")
We do exactly that :).
Shorter links.
Or for historical context, this is all Twitter's fault since its character limit included links too.
Before twitter, it was forum signatures with a character limit :’)
Worth noting it doesn't anymore, all links are now worth the same by going through Twitter's own link shortener.
To make affiliate/tracker links harder to spot and work around.
It’s often more convenient to have shorter links. Both for human convenience and length-limited mediums.
And many of them provide analytics to see who’s clicking your link.
I always thought it was good for tracking eg you have different links for different advertisements and you can track which ones get used.
I don't need shorter links per-se but I need a way to track the effectivness of the links I send out to the public for marketing purposes. I could put tracking ids and such but it doesn't look as good.
I also have links to my Google reviews, Google Maps etc and it's better show bitly.com/business-google-reviews and bitly.com/business-google/map than https://maps.app.goo.gl/sGa1KAU2131212cqU7
You want a way, and I can assure you that normal links look better, especially if the tracking part is easier to spot and can be removed by those who want to (who will be a minority, so your statistics won't be affected too much).
The other replies have largely covered most use cases they can also be handy for less tech savvy people to quickly generate links with unique analytics for marketing purposes.
That’s like asking what’s the value proposition of TwitPic. You had to be there.
Short urls are nicer if someone's got to write them down, remember the url, and make for a less dense QR code too. That's what I use them for.
I host my own now, for free and for fun.
We use it in text messages. Shorter, easier to read.
Shorter links
While usually purported to be for the shortness, it's almost always to track their usage.
It's lovely when books use them, and by the time you read them they don't work anymore.
Before Twitter, they were used to put URLs on physical things like business cards. They increased in popularity once Twitter imposed a 140 character limit.
People then saw that there were lots of secondary benefits. As another comment mentioned, hiding affiliate forwarders / query strings were a huge use back when everyone was trying to score those kickbacks. And then getting metrics on every clickthrough.
Back in the day every useful link you found on forums used to be shortened with an advertisment proxy service that promised a few pennies per click in exchange for making the user sit through a minute of ads.
What? I don’t remember that ever being a thing. What are you talking about?
that's why you should use https://shorter.dev to find your own short link(:
shameless
The tweet that McDonald's Sweden is proud of what they've achieved, redirected to the Wikipedia page for 9/11, is probably the most quintessentially 'Gen Z' joke to the point that dictionaries ought to use it as the example.
What aspects make it so quintessentially "gen z"?
I'm guessing some combo of dark humor, and the fact that they don't remember 9/11 so they don't mind joking about it, plus ironically trolling big corporations.
Older millennials were joking about 9/11 on 9/11
Pete Davidson, millennial of my vintage, to Justin Bieber in 2015:
>"Justin, you know I lost my dad on 9/11, and I always regretted growing up without a dad... until I met your dad, Justin. Now I'm glad mine's dead."
>And now for the greatest transition in the history of comedy. Two people from the movie Soul Plane are here, right? Kevin, Snoop—Soul Plane was the worst experience of my life involving a plane.
Pete's father died in the towers.
tribute.wmv was made by a millennial in 2002 and trolling big corporations on the internet has been a thing as long big corporations have been on the internet to troll. None of those are even vaguely genz traits.
I remember that exact type of humor from the "good old internet" when schoolchildren introduced each other to Rotten and 4chan would close the pool.
In fact, that's prime 4chan humor.
Is there a version of this "article" not written by AI? Feels like this was just an AI scraping twitter and writing an article.
It really took ages to get anywhere. Even the section explaining how the hack happened left me non the wiser.
Basically, bit-rot-ly missed up by deleting old link names when deleting accounts; people now register old links and redirect them to wherever they want.
Isn't Twitter's original "feature" or short "content" how URL "shorteners" (dang technology making us say ridiculous non-words with a straight face) came about? With that framing, this is pretty funny, at least to me. Well ok, I find this very funny regardless of how we got here.
This "article" is just direct LLM output ... Not even from a prompt like "write an article"
i called this 17 years ago http://joshua.schachter.org/2009/04/on-url-shorteners
Oh good.. maybe they can take over some of those old Rick rolls, 2 girlz... and goatse meme links.
My god. Who could have seen this coming.
A bigger issue perhaps than some old links now redirecting to scams is large scale link rot. The scam problem does already exist for normal links when the domain is not renewed and the new owner points it to a shady website. But link rot is a wider problem: if one link shortener ceases operations, a huge number of links will just no longer work, and it will usually no longer be possible to find out what content they were linking to, because you can't guess it from the domain name and URL slug as you would from a normal dead URL.
Boring article about some pretty funny jokes.
This is a lesson for any entry point including usernames , subdomains, url slugs. Anything of value will be abused.
The convenience of link shorteners is one thing, running them on your own domain seems to still make the most sense.
The convenience can so a large extend be blamed on people not doing proper URLs. Typing bit.ly/qwerty is easier than typing company.com/pages/1234567/november_marketing_pre_black_friday?utm_page=123&utm_source=app but it also lose all context and no marketing department would want to go without adding a metric ton of UTM tags to that URL because internal tooling either doesn't exist or they don't know how to use it.
From previous jobs I can safely say that most people in marketing have zero knowledge about how the internet and the web works, and they also don't care.
There's nothing preventing a company from running a URL "shortener" on their website directly. It could be as simple as "company.com/campaign/qwerty/" which then redirects the same way bitly will do. That was our solution, just namespace everything, e.g. /product/<sku>, /search/?query= and /campaign/<id>. It works every bit as well, has the same features and was 100% within our control.
People can and do purposely delete all the context content from all long URLS. :)
Firefox has added a feature for this:
https://support.mozilla.org/en-US/kb/enhanced-tracking-prote...
What is the convenience, really? In 2026, we're rarely speaking/reading URL's. Anything on a print ad can be a QR code. Where that doesn't work, you can say, "go to <company> dot com and search for <product>", like how people used to advertise AOL keywords. Or they can put a link to their active promotions above the fold on their home page.
The main convenience seems to be for pranks, plus it's just another vulnerability as described in the article. At best, they're just something else that has to be maintained forever.
I’ve seen it used in letters when an address is the only form of contact you have.
And, actually, if your long url uses mostly meaningful words, it's often easier to remember and type than a gibberish shortened url
There's a lower chance of a link being truncated if it's being forwarded, or copied/pasted or referred into some way.
Also, including named short links in a url is a friendly way to put a link on printed materials, email, etc.
There's lots of relevant uses for link shorteners. Maintaining it forever is for sure a consideration, it's why running one from your own domain lets you own the 301 redirect permanently.
1. People really really hate QR codes - and I understand why.
2. Podcasts are a thing that exists
Maybe this isn't the case for the majority of podcast listeners, but I've never had to hear and type a link from a podcast if that's what you mean. Don't most places you hear a podcast also have just enough UI to hold producer supplied text/show notes?
Ya. I could see random individuals using them on twitter, but McDonalds couldn’t figure out a way to create shortened links on their own domain?
Seriously, I built a link shortener on nothing but Cloudfront in about 30 minutes zero to live.
Some of the link shortener projects have some additional features that look useful, but still, running something directly on cloudfront/flare seems hard to overlook.
We used to worry about link shorteners. We still do, but we used to, too.
Link shorteners are just waiting to get abused.
Imagine if someone is using curl | bash on a shortened URL because "it's too long" and you can't see where the link is going.
Weird tangent. curl|bash doesn't work with link shorteners anyway because you need an extra flag to handle redirects.
Obviously should probably never use this in the |bash situation, but it's `-L` -- in general a good one to use by default unless you're only using curl to inspect redirect headers or something.
Very often they'll add the -L in the instruction because the url is a redirect to the latest version. See for example UV: https://docs.astral.sh/uv/getting-started/installation/#inst...
> We tested the method on links nobody disputes. A link whose creator says he made it in 2009 shows a 2009 creation date. A custom link tweeted by a YouTube channel on 21 June 2017 shows 20 June 2017. Among the Trump links that are still original, 32 of 40 were created on the day of the tweet or the day before.
Fucking AI prose
The whole issue could be resolved if bit.ly just actually did what they said rather than lie.
Obligatory https://xkcd.com/1698/