Perhaps you’ve been following the Tr.im fiasco. If not, basically the URL shortening service shut down and said all its links would cease to work by the end of the year, dealing a severe blow to users of any URL shortening service. Tr.im has since recanted its decision (if only to make it easier to sell), but the problem is still a very real one: What happens if your favorite URL-shortener just shuts down? 301works hopes to solve that.
Perhaps you heard about 301works in one of our recent pieces about how Bit.ly was attempting to salvage the Tr.im wreckage. The idea was the 301works would be a centralized hub for all shortened URLs, not run by any one URL-shortener. Tr.im balked at the idea of joining, but plenty of others are, including Bit.ly, Awe.sm, Adjix, betaworks, Cligs, and URLizer. All of them are teaming up with Gnip to launch this project.
One of the holdups in Tr.im’s participation was that it didn’t want one company ruling all of this data. And while Gnip will be handling it at first, to get the project off the ground, the plan is still to find a non-profit group to manage 301works. All the members are clear that they want this to be an open-source project that sets users’ minds at ease about using URL-shorteners.
The service will launch sometime in the next few weeks, after the participating companies have a chance to tell their users that they will be backing up their links on 301works. While most are unlikely to have a problem with that, some might, so they’re giving them some time to opt-out.
So how will 301works actually work? Well here are the key points for how companies will be able to back up their links:
- URL shortening services decide the frequency that they will make updates.
- URL shortening services decide how their updates can be made available to the public. Some services will provide regular uploads and downloads (hourly, daily, weekly, etc) and some will opt for a pure archival approach.
- Gnip is providing the infrastructure service to support aggregating data from URL shortening services. Gnip will provide the infrastructure service to compress the data into pre-defined download options for end users.
- Companies will be able to submit data via a REST API using HTTP POST over SSL. In addition, Gnip can provide other approaches on request.
All of this sounds great on paper, but the question of just how well this system works remains to be seen. Still, it’s promising that we’re seeing a bunch of companies take action on this so quickly after an incident that left a lot of people concerned about the future of URL-shorteners.
I see a problem with this though; the value in the majority of URL shortening services – besides one like awe.sm that provide something different – is the data that they hold, if they were handing over the data to another service for whatever reason it means their data is no longer unique of valuable. Take tr.im for example, anyone with an ounce of PHP understanding could clone it in a day, but because tr.im holds loads of URLs and stats it’s “better” than the clone, if the clone can get this data where is the value in tr.im? There isn’t any.
URL shorteners rely on the data they hold, if anyone can access it then where is the value?
Companies like awe.sm, bit.ly and cli.gs are all providing more than just a 301 redirect. They’re building value based upon the data surrounding pages, people and sharing. That metadata isn’t getting pushed into 301works.
Eric Marcoullier
CEO, Gnip
So I guess the common practice which Flickr, Techcrunch does is pretty well – providing their own shortened uri (yes, uri) should be the best way to do so. I don’t have to shorten techcrunch’s because it is already as short as http://tcrn.ch/5lF8.
this doesn’t solve the problem…web pages that still link to tr.im/whatever would still die when the service shuts down or compromised.
the problem is not the data, the problem is the links that are in place.
sure, if you end up with the domain name that formerly belonged to a shortener and participated in the 301works backup, you could resurrect the service. that’s a pretty unique situation. link rot isn’t.
Completely agree that it’s not a perfect solution, but it’s the first time that the leading shorteners are working together to solve the problem.
As for your specific use case, imagine that all of bit.ly’s mappings are in the public domain. If bit.ly were to suddenly shut down, then the leading applications (Twitter, Tweetdeck, etc) could do their own inline resolving.
Eric Marcoullier
CEO, Gnip
Great idea, and I see all the problems that you face.
I don’t know if this is what you plan to do, but this is what i would do, and as I just brought Youific.me and .com i see this is something I would be interested in.
I think that all shortURL services should share the following information
1)URL shortner
2)URL address
for e.g. if say I have a URL shortner of http://youific.me/tcrunch
and the link is
http://techcrunch.com
Then any site like tr.im or bit.ly would be able to access that information
so they would also be able to have tr.im/tcrunch or bit.ly/tcrunch.
Let me know your thoughts and Eric Marcoullier please contact me on twitter russellharrower.
@Samuel You are correct and you are wrong. Url shorteners *are* valuable because of the data they have, but it’s not the short->url mapping that is valuable, it’s the click data.
My understanding is that click data is not part of the transfer, the 301works project is there to ensure that links never die.
Now, there is some value in looking at the links that are created, but it’s not that useful if you don’t also know how often any of those links are clicked.
Correct, the only data that is part of the transfer is the following
name of shortener service
the short URL
the original URL>
In addition, the shortening service has the option to include the original creation time of the short URL as the timestamp element.
That is all that is necessary to provide the backup for mappings. It is also this level of simplicity that makes it possible for providers to join as they can make the web more open while still running their own individual businesses.
Cheers,
Shane
VP, Products
Gnip, Inc.
As the operator of tr.im the 301works.org project is an interesting idea for maintaining links going forward. I will keep an eye on it and see how it develops.
It would be really nice if the people at TC would make the small effort to actually contact me to collect additional facts for their articles about the tr.im fiasco, rather that publish assumptions over and over. bit.ly’s “offer” this past Monday to “help” was little more than a PR stunt to inject themselves into the conversation as the supposed saviour.
However, if GNIP has become involved it is a more credible option now that can actually be considered.
if there were not lots of odd assumptions and rushed delivery, it just wouldn’t be techcrunch
The ‘obvious’ independent recipient/manager of this data would seem to be the Internet Archive project, http://www.archive.org.
(But they might consider their plate already quite full.)
In the title, you might want to check your word usage. “Insure” generally applies to something covered by a financial backing. “Ensure” applies to something that is guaranteed to work.
sure, considered that, but i wouldn’t take it so literally
was going to work in ‘insurance’ but it was too long.
This is a nice idea, but the premise that tr.im isn’t working is incorrect. I’m using it through the API and interactively and there are no problems. It’s no more or less reliable than the other URL shorteners.
DW — glad you like 301.
About tr.im – wish Eric only well. Offered to help last Sunday when the announcement happened, offered again on monday. Very glad tr.im is back — the world is more interesting for it, look forward to seeing what you and he develop.
The URLs I put through a shortening server are only ever transient. There for throwing on twitter or IMing someone, so I really couldn’t give a damn if that service went awayt… I really couldn’t.
I wish urljar.com could join this alliance, they claimed to be the best with many unprecedented features. Their youtube demo is quite impressive http://www.yout...h?v=I53SevFBP0I
I’m trying to understand how this is to work. If urlshortener.com dies, 301works.org will be a directory containing all of its old mapping urls that i can search urlshortener.com/urlhash and it’ll turn up actualurl.com/actualdir/actualfile.html ?
how common will 301works be known to regular users? for those that will know who’s going to take the extra step of going to 301works to find out where the link actually goes?
i mean it sounds like a neat idea, but how relevant is it to users at large?
it is possible that the 301works idea is one to put a positive spin on the so-called url shortening business after tr.im rightfully realized that it was not worth staying in the game. by tr.im temporarily preventing users from making more tr.im urls (unless via API), and the response it garnered brought a ton of attention on this flawed business.
seems like bit.ly made their public offer to tr.im to “save” them as a message to all that bit.ly holds the crown and will not contribute to link rot and will help to fix other services link rot when they collectively shut down over time. a marketing tactic to fix the image of the url shortener/tracker market with emphasis of course on announcing themselves as king (thanks to twitter and shared investors).
301works is not a terrible thing to have out there if the idea of short urls taking precedence over the actual longer permalinks is going to stick around for awhile. but like another commenter pointed out… having a backup of all the links will not prevent link rot unless the domain is transferred to 301works or at least it would need to point to a new server so the links resolve. the former will never happen and the latter would happen at best as a temporary agreement until the domain is re-purposed and/or sold.
without the domain, applications themselves would have to handle an internal resolver which is possible with a small bit of code (if domain = broken short url domain, domain = working backup short url domain). feasible, sloppy, annoying, vulnerable.
this entire topic is almost funny and if urls were handled outside of the 140-160 char micromessage as metadata, short urls would just be a novelty/vanity thing.
I got few good information about all url shortening services. Though I don’t use them it is helpful information. Mostly I use 301 redirection.
tr.im should be where bit.ly is. i have been using it exclusively for trackbacks to blogs and other links alike.
if it’s tr.im’s time, then i guess it is.