Hacker News new | past | comments | ask | show | jobs | submit login
[dupe] Google is shutting down its goo.gl URL shortening service (goo.gl)
81 points by modinfo on Mar 31, 2018 | hide | past | web | favorite | 56 comments

"While most features of goo.gl will eventually sunset, all existing links will continue to redirect to the intended destination." https://developers.googleblog.com/2018/03/transitioning-goog...

I hope this doesn't change too. Felt I had to use it to link a humongous wikidata SPARQL query in a (software) research paper recently.

I think in content intented for long term it is always better to use the canonical address, possibly in a footnote if it is too long.

Better yet also include an alternative link to an archive like https://archive.li/ or similar.

I hope you added the entire link in an appendix, otherwise it’s not Google’s fault for ruining your paper.

The query itself is available in a screenshot. Also a pretty periferal part of the paper.

Looked/thought hard about options but did not come up with better at the time.

But wanted to highlight how important it is for parties starting such services to have a plan for longevity. Hopefully Goog will keep old links going for a long time.

Please consider using DOI’s for this http://www.doi.org/factsheets/DOIKeyFacts.html

Agreed. Figshare is a good service for this. In addition all the publishing infrastructure is in place.

Great idea! I'm using figshare but neither me nor colleagues got to think about that.

I'm not sure how you guys feel about this, but personally, I've been avoiding Google services for the last couple of years because I just assume they'll eventually get shut down.

Same here. Only thing I'm still strongly dependent on is gmail. I don't really think that's going away anytime soon, but out of principle I'd like to move away from it. Takes a lot of energy unfortunately and you probably need to keep a redirect in place for a long time just in case...

Is there a self-hosted web mail service w ux like gmail, that you can migrate to smoothly and pay a license for?

I would consider this move. I just assume no one has ux even close. It would take me at least a year to fully transition from one gmail address.

Not self-hosted, but I am very happy with Zoho [0]. They target companies and their free plan includes lots of features like support for your own domain. The web UI is comparable to GMail (indeed in parts I prefer it) and they offer a very good Android app as well.

[0]: https://www.zoho.com/mail/

Zoho recently removed IMAP and email forwarding from their free tier (though existing users were grandfathered in), so stay away if you value email portability.

Rainloop is very usable. It just pulls everything in via IMAP. I have scripts in place to handle any filtering and organising I need to do.

No afaik. I accepted the loss of usability and moved away anyways.

Yes. I've migrated to other services for just about everything they offer except gmail. I'm still looking but have not found a comparable service in terms of quality and pricing.

If anyone has suggestions I would be really interested in hearing them.

FastMail. It’s great. Comparable interface to Gmail - I like it better, even - great customer service, true and reliable IMAP, easy to use your own domain and easy to migrate from Gmail. Also has Calendar and Contact support.

Isn't this something that you should keep in mind for all services? After all they are basically applications you have no code for, running in computers that you do not control, funded by means you individually cannot affect and owned by companies/businesses making decisions you have no say over. Any element of control you cede to others can be used in ways you may not like or want. So if you want to ensure something will continue to exist for as long as you want it, the only way to do it is to have control over that something and either control all of its dependencies or rely only on dependencies that are widely available commodities that you work with through as many open standards as possible.

For example, for web-based services, if you don't want something to go away (or be abused), then your best bet is to run a fully open source implementation of that version on a VPS (VPSes are on computers owned by others, but they are cheap commodities that you can easily replace, replicate and, if necessary, move to dedicated servers or your own machines). Personally i'd also only use simple applications with little to no (preferably no) external dependencies in case i need to pick up the development myself or make modifications.

(of course having said that, i generally avoid web-based applications altogether and instead opt for desktop-based solutions whenever possible - one major reason being control)

Yes it is, but I think - perhaps counter intuitively to some, that the larger the company the larger the risk of being left without the product or service you grow to like. In the case of Google, Microsoft and some others I'm now wary of even trialling things.

Being owned by a major is bad for life expectancy. Being bought out by one is worse. Google are quite probably the pack leader for shuttering services for no obvious reason (Reader etc) and also buying startups to shutter them or turn them into something unrecognisable.

I try to apply this view when choosing all products and services - I'll favour startup and family owned over multinational brand for everything from SaaS to food and clothing and often pay extra to do so. I'd like to think there's still a chance of individual customers mattering to them.

I also favour desktop where possible.

> I'm not sure how you guys feel about this,

Here's my main problems with google and why I've been trying to get away from them as much as possible.

1. Very poor quality customer service. They might have better customer service for things like adwords (I wouldn't know) but as far as everything I have experience with (business services, google cloud) my experience has always been horrible - good luck getting in touch with a human.

2. (as you mention) A high level of insecurity about whether or not a google service will actually be sustained or whether they will drop it and leave all their users hanging to go try out their next big idea.

3. IMHO unethical business practices, mainly when it comes maintaining their monopoly and violating copyright laws [1].

4. The fact that they control so much of an ever more centralized internet. Regardless of however benevolent one may think they are that much power concentrated in the hands of any one entity is likely a dangerous road to tread upon and one I'd rather not risk going down.

1. https://www.quora.com/What-are-examples-of-Google-acting-une...

I think thats good advice for anything, I would however put the caveat on things that use open protocols. I have no issue using AWS SES (Simple Email Service) because thats just SMTP, I would however never use something like Firebase. Just measure the time it would take you to get rid of it in a project.

I would also never use the sentiment shared on HN much for anything, half of the people here have a stake in their own little proprietary technology they want you to sell :P

They lost my trust after how they treated the 2013 Nexus 7. Great tablet in spite of the company behind it.

I've recently came across a use case for an URL shortener (Entering long API calls on mobile devices) and was looking for a self-hosted solution. Any recommendations?

(No public shortening service, only admins can create new shortcuts.)

Yourls is pretty good, I use it with Sharex (and Nextcloud) to quickly upload images to the internet.

2nd for Yourls; easy to install and maintain.

the following should cover your requirements: https://github.com/adfinis-sygroup/node-url-urlshortener self-hosted and only admin can create new shortcuts. Hope it helps :)

Why not just have a cloud-synced document, from which you can tap/copy the real URLs on the mobile device?

I'm looking for a URL shortener that tells me how many users clicked the link.

Is that your only requirement? In that case even my very simple one would be enough: https://github.com/ptman/urlredir

Other requirements would be that it is actually hosted somewhere :) and doesn't ever shut down.

Bitly does this. Just add a + to the end of any bitly URL to view stats, e.g.:


The last time I used an URL shortener was in the old days when Twitter still counted every character of the URL against the 140 char limit. I'm curious what use cases URL shortening services have nowadays?

A lot of people use them for tracking. For example, you can use one underlying URL but have several shortened ones for various promotional places (email, Twitter, etc) and then use the analytics for the shortener to see which place is actually sending you traffic.

There are a lot of places where having a compact url is handy. One page I often use them is when putting links in code comments, when I've used a blog post or StackOverflow answer as the basis for some tricky solution to a problem.

I used one for an IRC chat bot, because long URLs in terminal clients suck.

It is used heavily by spammers to hide url's. Often, 5 or more shorteners (including less-than-reputable ones) are used in a chain.

I use them in the title / banner box in hipchat as a sticky resources finder for relevant info for that chat room. Display space is limited.

- Tracking

- Sharing (Looks better when written out [like often social media posts], is easier to remember)

Archived copy, which works with JS disabled:


I used to register and use domain go.gl for my photography business in early 2000’s

Was thinking about registering goo.gl just for kicks (it was available back then) but turned off this idea because Greenland would charge $50+ for registering. And it had to be done over the FAX!

If you want a free OSS URL shortener for your own domain name, I built a simple one at http://xluh.co/repo. Will be expanding on it in the future.

I really would appreciate a link shortening service automatically creating IPFS mirror!

This would not be a problem if websites themselves provided shortened URLs for all paths within their domain name:

  foo.com/2018/03/31/the-title-of-an-article => foo.com/a3ef33

I do not understand. Please say what you mean in your message instead of just dropping some urls:s. For example now, your url:s goes to this submission and an unknown (!?). If you had written what you meant at least I could understand your point.

They are examples of a long "short URL" following your specification.

might shorten to

which is not very short given the length of the hostname.

A good alternative: https://lc.cx

I guess I was just a little prescient back in 2011, when I wrote of URL-shortening services in a blog post, "let's not forget that these entities have a nasty tendency to vanish". Not that being prescient about the Big G end-of-lifing services is particularly visionary or difficult, tbh.

The post[1] was to publicise a small shortener I'd written for myself (Java/JSP solution) -- in many ways deficient and lacking all sorts of features (stats, UI) but it worked well enough for me for a time until I got lazy, got ill, and got rid of the server it was running on. But I was bang on-target about the likelihood of such things disappearing under the waves.

I guess it's time to dust it off[2], bring the codebase up to date and get it working again. I'm a bit doubtful about hosting it on, say... Google Compute free-tier or something... for who knows when /that/ service will get taken down.

[1] http://onemikro2nd.blogspot.co.za/2011/04/shrtn-url-shortene...

[2] https://bitbucket.org/mikro2nd/shrtn

> it worked well enough for me for a time until I got lazy, got ill, and got rid of the server it was running on. But I was bang on-target about the likelihood of such things disappearing...

Oh, the irony...

Kinda the opposite of irony. They said these things disappear. Had they said, "but this one won't," then we would be approaching irony.

My impression was that building his own URL shortenner was undertaken as a mitigation of the risk of other URL shortenners disappearing.

It has been obvious that they were bad for a while. I called it years before: http://joshua.schachter.org/2009/04/on-url-shorteners

And no, the answer is not running more shorteners. You are just making the problem worse.

all existing links will continue to redirect to the intended destination.

It would be a PR nightmare for Google if they stopped redirecting existing links, but they won't do that as they're a huge user of the web graph as well.

Indeed, I saw that redirection will continue to work. It wasn't clear to me, though, whether any of the other stuff -- stats, particularly -- will continue to work.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact