i read in an earlier thread for this on HN - "this is a classic example of data driven product decision" aka we can reduce costs by $x if we just stopped goo.gl links. Instead of actually wondering how this would impact the customers.
Also helps that they are in a culture which does not mind killing services on a whim.
The Google URL shortener stopped accepting new links around 2018. It has been deprecated for a long time.
I doubt it was a cost-driven decision on the basis of running the servers. My guess would be that it was a security and maintenance burden that nobody wanted.
They also might have wanted to use the domain for something else.
The nature of something like this is that the cost to run it naturally goes down over time. Old links get clicked less so the hardware costs would be basically nothing.
As for the actual software security, it's a URL shortener. They could rewrite the entire thing in almost no time with just a single dev. Especially since it's strictly hosting static links at this point.
It probably took them more time and money to find inactive links than it'd take to keep the entire thing running for a couple of years.
> I doubt it was a cost-driven decision on the basis of running the servers. My guess would be that it was a security and maintenance burden that nobody wanted.
Yeah I can't imagine it being a huge cost saver? But guessing that the people who developed it long moved on, and it stopped being a cool project. And depending on the culture inside Google it just doesn't pay career-wise to maintain someone else's project.
I think the problem with URL shorteners like Google’s that includes the company name is that to the layperson there is possibly an implied level of safety.
Here is a service that basically makes Google $0 and confuses a non-zero amount of non-technical users when it sends them to a scam website.
Also, in the age of OCR on every device they make basically no sense. You can take a picture of a long URL on a piece of paper then just copy and paste the text instantly. The URL shortener no longer serves a discernible purpose.
Goo.gl didn't have customers, it had users. Customers pay, either with money or their personal data, now or the future. Goo.gl did not make any money or have a plan to do so in the future.
The monetary value of the goodwill and mindshare generated by such a free service is hard to calculate, but definitely significant. I wouldn't be surprised if it was more than it costs to run.
Which raises the obvious question -- why make a service that you know will eventually be shut down because of said economics. Especially one that (by design) will render many documents unusable when it is shut down.
While I generally find the "killed by Google" thing insanely short-sighted, this borders on straight-up negligence.
I always figured most of the real value of these url hashing services was as an marketing tracking metric. That is, sort of equivalent to the "share with" widgets provided that conveniently also dump tons of analytics to the services.
I will be honest I was never in an environment that would benefit from link shortening, so I don't really know if any end users actually wanted them (my guess twitter mainly) and always viewed these hashed links with extreme suspicion.
One of the complaints about Google is that it's difficult to launch products due to bureaucracy. I'm starting to thing that's not a bad thing. If they'd done a careful analysis of the cost of jumping into this url-shortener bandwagon, we wouldn't be here. Maybe it's not a bad thing they move slower now.
I would bet that the salaries paid to the product managers behind shutting this down, during the time they worked on shutting it down, outweigh the annual cost of running the service by an order of magnitude.
At this point, anyone depending on Google for anything deserves to get burned.
I don't know how much more clearly they could tell their users that Google has absolutely no respect for users without drone shipping boxes of excrement.
If companies can spend billions on AI and not have anything in return and be okay with that in the ways of giving free stuff (okay, I'll admit not completely free since you are the product but still free)
Then they should also be okay for keeping the goo.gl links honestly.
Sounds kinda bad for some good will but this is literally google, the one thing google is notorious for is killing their products.
This is basically modern SV business. This old data is costing us about a million a year to hold onto. KILL IT NOW WITH FIRE.
Hey lets also dump 100 Billion dollars into this AI thing without any business plan or ideas to back it up this year. HOW FAST CAN YOU ACCEPT MY CHECK!
Aurornis|7 months ago
I doubt it was a cost-driven decision on the basis of running the servers. My guess would be that it was a security and maintenance burden that nobody wanted.
They also might have wanted to use the domain for something else.
cogman10|7 months ago
The nature of something like this is that the cost to run it naturally goes down over time. Old links get clicked less so the hardware costs would be basically nothing.
As for the actual software security, it's a URL shortener. They could rewrite the entire thing in almost no time with just a single dev. Especially since it's strictly hosting static links at this point.
It probably took them more time and money to find inactive links than it'd take to keep the entire thing running for a couple of years.
JimDabell|7 months ago
Cloudflare offered to run it and Google turned them down:
https://x.com/elithrar/status/1948451254780526609
rdtsc|7 months ago
Yeah I can't imagine it being a huge cost saver? But guessing that the people who developed it long moved on, and it stopped being a cool project. And depending on the culture inside Google it just doesn't pay career-wise to maintain someone else's project.
1over137|7 months ago
It's a strange thing to consider 'since 2018' "a long time". Only in tech circles is this so, not in normal life.
rany_|7 months ago
Edit: nevermind, I had no idea Dynamic Links is deprecated and will be shutting down.
mort96|7 months ago
dangus|7 months ago
Here is a service that basically makes Google $0 and confuses a non-zero amount of non-technical users when it sends them to a scam website.
Also, in the age of OCR on every device they make basically no sense. You can take a picture of a long URL on a piece of paper then just copy and paste the text instantly. The URL shortener no longer serves a discernible purpose.
EGreg|7 months ago
resize2996|7 months ago
jerlam|7 months ago
CydeWeys|7 months ago
That way they'll make money, and they can fund the service not having to shut down, and there isn't any linkrot.
franga2000|7 months ago
cyphar|7 months ago
While I generally find the "killed by Google" thing insanely short-sighted, this borders on straight-up negligence.
somat|7 months ago
I will be honest I was never in an environment that would benefit from link shortening, so I don't really know if any end users actually wanted them (my guess twitter mainly) and always viewed these hashed links with extreme suspicion.
thevillagechief|7 months ago
margalabargala|7 months ago
observationist|7 months ago
Imustaskforhelp|7 months ago
Then they should also be okay for keeping the goo.gl links honestly.
Sounds kinda bad for some good will but this is literally google, the one thing google is notorious for is killing their products.
citizenpaul|7 months ago
Hey lets also dump 100 Billion dollars into this AI thing without any business plan or ideas to back it up this year. HOW FAST CAN YOU ACCEPT MY CHECK!
manquer|7 months ago
For company running GCP and giving things like Colab TPUs free the costs of running a URL service would be trivial rounding number at best
b112|7 months ago
I've handled far more traffic on single machines 20 years ago.
no_wizard|7 months ago
j45|7 months ago
Like other things spun down there must not be value in the links.
miohtama|7 months ago