Using .tel domains for URL shortening could be an interesting solution, but I'm not sure it's such a good idea. .tel's purpose is to disseminate contact information through the DNS. Do short URL hacks really fit sit well with this objective, or would it just be a distraction? I'm not sure.
On the other hand, it could boost .tel adoption, which would be a good thing :)
Andy Chantrill
· 4 months ago
Using .tel domains for URL shortening could be an interesting solution, but I'm not sure it's such a good idea. .tel's purpose is to disseminate contact information through the DNS. Do short URL hacks really fit sit well with this objective, or would it just be a distraction? I'm not sure.
On the other hand, it could boost .tel adoption, which would be a good thing :)
Timothy Post
· 4 months ago
Andy, Thanks for the comment.
I'm not necessarily "married" to the .tel doman itself. Rather, what I'd like to see is that web users have the opportunity to connect a domain, which they own to the various web services.
I have written in the past about OpenID and how it's my opinion that the service will realy catch fire when OpenID Providers offers users the opportunity to create a sub-domain for their central OpenID social hub, as well as, use that sub-domain for their universal login.
I thought of the .tel domain because it does seek to be a domain, which people can use as their central web hub and, I believe, it offers some additional add-on features.
Regarding short URLs, I should point-out that su.pr (StumbleUpon's entrant in the short URL world) has enabled a feature, which enables users to create a short URL using their websites domain for links that point to that website, itself.
into Twitter, I could, with su.pr short URL service, paste something like:
htpp://timothypost.com/hsrn instead.
While this su.pr feature is nice it doesn't, imo, go far enough. I would like to see su.pr enable users to shorten ANY website's URL with their own domain.
Do you know of any technical issues, which would make this new "custom" short URL impossible?
Andy Chantrill
· 4 months ago
OpenID delegation is somewhere on the roadmap for .tel.
Currently you can't redirect .tel CNAME's to external URLs, but it might be a feature worth considering. The problem is that we'd lose the predictability of the HTTP TelProxy always returning a uniform representation of records.
How does the su.pr service work? Do you have some kind of rewrite rule that forwards requests to the su.pr server, or did they provide you with a standalone script to use on your server? In any case I can't see any technical reason why its target URLs should be restricted to your domain.
Timothy Post
· 4 months ago
Andy:
While I am most certainly a "layman" when it comes to understanding domains, DNS, top level domains, etc., I can give some perspective from the average "power user's" viewpoint.
Couple thoughts:
1. Are you aware of Card.ly? It's an analogous service which let's users aggregate their personal information, web affiliations, and contact info in one place. They're using the "freeium" model. You can host your profile as on card.ly top level domain for free. For $25 a year a user can host their profile on their own domain and they have access to "premium" skins.
The option to skin your profile (you can customize it with your own CSS) is a very powerful selling point, imo. I would like to see .tel offer, if it doesn't already, the option of skinning one's profile. As we have seen with Wordpress and Rapidweaver, such an option encourages 3rd party developers to quickly offer the public a lot of great options.
What I don't know for sure is whether you can create a CNAME so that you could create a sub-domain that points to your card.ly profile. I assume its possible but I'm not sure.
2. I would argue that OpenID delegation should be the #1 on the .tel roadmap. Currently, there is a huge opportunity for an OpenID provider to offer DNS verification for sub-domains. The mistake that the current group of OpenID Providers are making is that they view OpenID as an opportunity to drive traffic to their own domains and they hope to monetize that traffic in the future.
Therefore, they have all made the classic mistake of trying to "tie down" people's account to their own domains. While they all say that you can create an OpenID domain on one which you already own, I have not seen how this is possible and trust me, I have looked long and hard.
Basically, these guys don't get the new "open web" and they are basically cutting off their noses to spite their faces.
Almost all the attention of OpenID has been focused on the universal login capability. That's a mistake. Google and Facebook has effectively killed OpenID's universal login growth by offering the same service for their users.
Therefore, telnic should seize the opportunity to become an OpenID Provider itself and enable folks to create their "social web hub" on their own .tel domains. This window of opportunity will probably close, in my opinion, shortly as Google "marries Google Apps, Google Profile, and Google Connect.
The key take-away from the OpenID environment is that people "in the know" do not want to be beholden to a web service, as Robert Scoble detailed with tr.im The other point is that people primary domain is not a good location for their social web hub because in most cases that domain is already being used for other purposes (e.g blog, website, etc.).
Telnic is uniquely positioned because it is already offering folks a top level domain as their social web hub. People will, I suspect, not be opposed to having a .com (or whatever top level domain) for their main website and a .tel top level domain for their single point of contact.
What .tel, imho, should focus on is not redirecting the primary .tel domain to external URLs but rather, redirecting people's .tel sub-domain to external URLs. This solution, of leaving the users primary .tel domain (say http://bobsmith.tel) alone and only letting users redirect SUB-DOMAINS would therefore, mean that you would not "lose the predictability of the HTTP TelProxy always returning a uniform representation of records."
I should note that currently I know of only a couple web services who offer this option of redirect and masking. Wordpress.com, Squarespace, and Google Apps. I have to admit that I was frankly very surprised that Facebook didn't offer this option to its users. The current FB username option is still weak because your username lives on FB's domain.
Facebook should have given people the option of redirecting and masking either their primary domain (i.e. timothypost.com) or a sub-domain (facebook.timothypost.com).
If Telnic moves quickly and smartly it could partner with dozens and dozens of web services to offer sub-domain redirects to the user account pages and masking with the user's own domain.
Ultimately, I think we will begin to see a service like Google Apps but for many, many web services. I have set-up my Google Apps account so that I can access my email account at http://email.timothypost.com, my calendar at http://calendar.timothypost.com, etc.
I want to do this for ALL my web services. One twist would be for me, the user, to name the sub-domain for the service, not the company providing the service. For instance, instead of http://linkedin.timothypost.com, I would name this sub-domain http://resume.timothypost.com. This would give me, the user, ultimate flexibility if, in the future, another better "resume" service came around.
Telnic could start by offering its users the option of redirecting only. It's my understanding that to truly mask the external URL one needs the cooperation of that company. Telnic could promote its developer API so new web services, say competing with LinkedIn, would offer masking as a competitive advantage.
This sub-domain redirect and masking topic is very important for two reasons:
1. User fatigue. One of the biggest hurdles facing web servcies today is that users are absolutely sick and tired of creating new accounts and then having to, once again, fill in their data.
Telnic could be the central place for people to enter all their personal, work, hobby, interests, etc. information with industry standardized Fields and when a user linked their .tel account with a new web service it would auto populate that information. People could "curate" their information at their .tel account accounts instead of in 37 different places.
A new technology from Google Pub Sub Hubbub would mean that when folks update their information at .tel it would immediately disseminate to their other accounts.
2. "Walled Gardens" not only are people sick of having to reenter data over and over but they are frustrated when that data is hard to export out of a web service. This solution with .tel would solve this issue because people's data would live in their .tel accounts.
This point raises another pertinent topic of today, namely where people's data is stored. This point was also raised in Robert Scoble's blog post.
Ideally, what I would love to see .tel do is a partnership with one of the companies offering a front-end interface to Amazon's S3 storage. In a perfect world, I would upload files, photos, etc. into my .tel account account and they would be stored at Amazon S3. Then, when I sign-up for a web service, such as Flickr.com, the web service company would access them at my S3 server. I already serve some photos to my blog using the S3 "Copy Path" option.
Obviously, many users would not want this option because they would have to pay for the storage. Nonetheless, Telnic should be offering this service to its power users who are also the thought leaders on the internet. Guys like Dave Winer would love this option and so would many corporations and organizations.
Telnic could actually create a revenue stream by doing a partnerhsip agreement with Amazon where it gets a commission for folks who access S3 storage through through .tel accounts.
Here's the URL for the su.pr set-up instructions for custom domains:
I have a suspicion that su.pr may be saving the custom domain short URL to ALL external URLs as a paid premium offering. They are not saying yet but in their Help Forums there's been some speculation able this possiblity.
So, to wrap up, I am excited for .tel and the possibilities before it to truly reshape the web and to give the power back to the people.
PS: I'm curious, I left Network Solutions and shifted all my domains over to GoDaddy but I read that GoDaddy is not one of the registrars able to offer .tel. Do you have any idea on why that is so. I would think that .tel would want to establish itself as the ultimate personal domain and thus, GoDaddy scale and reach would be very important?
Great to chat with you. I have been meaning to put my thoughts down as a blog post and thanks to you, I've just done that.
Henri Asseily
· 4 months ago
Hi Timothy, I really like the way you're going with the OAuth thinking. Let's discuss OAuth options beyond 3rd-party app authentication when you've got time. There are a few .tel features you're missing that could make the opportunity even more interesting. I'm of course at henri.tel...
Speaking of Card.ly, it's a web-based service that displays your contact info. It's got a nice UI but doesn't solve the problems of speed of access (http is slow), consistency (what if they change their html?), resilience (centralized web service) and persistence (what if Card.ly goes under?).
Timothy Post
· 4 months ago
Henri, I'd love to chat with you about OAuth and .tel. Drop me a direct message at Twitter and let me know when's the best time to reach you.
iWantMyName
· 4 months ago
Interesting thoughts and I think you're spot on. It's exactly the vision of our domain management service iWantMyName. Earlier this year we've launched "Domains For Apps" that lets you easily configure domains for popular web apps and services such as Google Apps, Zoho Business, Tumblr, Posterous, Squarespace, Moonfruit and many others.
We also wrote about the issue with URL shorteners in a blog article earlier this week and list as well hosted services as self-hosted options to run an URL shortener on your own domain:
Hey Tim, great thoughts on the .twt domain, I like it, but it's all up to the Twitter client software to decide what has to live and what has to die. You know I've been experimenting with short links lately and I wrote a plugin for WordPress which is doing quite good - Twitter Friendly Links. All your posts, no matter how long the permalink is, get shortlinks like yourdomain.com/7x1 or whatever. I've been using it for quite some time now on my own blog, but people are used to copying the link from their browser, into their twitter client (seesmic, tweetdeck or whatever) which is automatically shortened via bil.ly.
There's some microformats stuff going on with these shortlinks (the shortlink meta tag) but none of the Twitter client software work with those. It seems that there's not enough technical detail on what they have to do, and not enough software to gently handle the redirects with the proper headers and blah blah blah.
WordPress.com users recently got http://wp.me shortened links for each of their blog posts. It's nice to see that at least they care. Nevertheless people still post their username.wordpress.com/long-post-name-blah-blah on Twitter =)
All Comments
On the other hand, it could boost .tel adoption, which would be a good thing :)
On the other hand, it could boost .tel adoption, which would be a good thing :)
I'm not necessarily "married" to the .tel doman itself. Rather, what I'd like to see is that web users have the opportunity to connect a domain, which they own to the various web services.
For example, I would much rather have the opportunity to create a CNAME for my Facebook account so that my account at FB could be accessed at http://facebook.timothypost rather than at http://facebook.com/timothypost.
I have written in the past about OpenID and how it's my opinion that the service will realy catch fire when OpenID Providers offers users the opportunity to create a sub-domain for their central OpenID social hub, as well as, use that sub-domain for their universal login.
I thought of the .tel domain because it does seek to be a domain, which people can use as their central web hub and, I believe, it offers some additional add-on features.
Regarding short URLs, I should point-out that su.pr (StumbleUpon's entrant in the short URL world) has enabled a feature, which enables users to create a short URL using their websites domain for links that point to that website, itself.
What this means is that instead on me pasting:
http://www.timothypost.com/2009/08/10/i-want-a-...
into Twitter, I could, with su.pr short URL service, paste something like:
htpp://timothypost.com/hsrn instead.
While this su.pr feature is nice it doesn't, imo, go far enough. I would like to see su.pr enable users to shorten ANY website's URL with their own domain.
Do you know of any technical issues, which would make this new "custom" short URL impossible?
Currently you can't redirect .tel CNAME's to external URLs, but it might be a feature worth considering. The problem is that we'd lose the predictability of the HTTP TelProxy always returning a uniform representation of records.
How does the su.pr service work? Do you have some kind of rewrite rule that forwards requests to the su.pr server, or did they provide you with a standalone script to use on your server? In any case I can't see any technical reason why its target URLs should be restricted to your domain.
While I am most certainly a "layman" when it comes to understanding domains, DNS, top level domains, etc., I can give some perspective from the average "power user's" viewpoint.
Couple thoughts:
1. Are you aware of Card.ly? It's an analogous service which let's users aggregate their personal information, web affiliations, and contact info in one place. They're using the "freeium" model. You can host your profile as on card.ly top level domain for free. For $25 a year a user can host their profile on their own domain and they have access to "premium" skins.
The option to skin your profile (you can customize it with your own CSS) is a very powerful selling point, imo. I would like to see .tel offer, if it doesn't already, the option of skinning one's profile. As we have seen with Wordpress and Rapidweaver, such an option encourages 3rd party developers to quickly offer the public a lot of great options.
What I don't know for sure is whether you can create a CNAME so that you could create a sub-domain that points to your card.ly profile. I assume its possible but I'm not sure.
2. I would argue that OpenID delegation should be the #1 on the .tel roadmap. Currently, there is a huge opportunity for an OpenID provider to offer DNS verification for sub-domains. The mistake that the current group of OpenID Providers are making is that they view OpenID as an opportunity to drive traffic to their own domains and they hope to monetize that traffic in the future.
Therefore, they have all made the classic mistake of trying to "tie down" people's account to their own domains. While they all say that you can create an OpenID domain on one which you already own, I have not seen how this is possible and trust me, I have looked long and hard.
Basically, these guys don't get the new "open web" and they are basically cutting off their noses to spite their faces.
Almost all the attention of OpenID has been focused on the universal login capability. That's a mistake. Google and Facebook has effectively killed OpenID's universal login growth by offering the same service for their users.
Therefore, telnic should seize the opportunity to become an OpenID Provider itself and enable folks to create their "social web hub" on their own .tel domains. This window of opportunity will probably close, in my opinion, shortly as Google "marries Google Apps, Google Profile, and Google Connect.
The key take-away from the OpenID environment is that people "in the know" do not want to be beholden to a web service, as Robert Scoble detailed with tr.im The other point is that people primary domain is not a good location for their social web hub because in most cases that domain is already being used for other purposes (e.g blog, website, etc.).
Telnic is uniquely positioned because it is already offering folks a top level domain as their social web hub. People will, I suspect, not be opposed to having a .com (or whatever top level domain) for their main website and a .tel top level domain for their single point of contact.
What .tel, imho, should focus on is not redirecting the primary .tel domain to external URLs but rather, redirecting people's .tel sub-domain to external URLs. This solution, of leaving the users primary .tel domain (say http://bobsmith.tel) alone and only letting users redirect SUB-DOMAINS would therefore, mean that you would not "lose the predictability of the HTTP TelProxy always returning a uniform representation of records."
I should note that currently I know of only a couple web services who offer this option of redirect and masking. Wordpress.com, Squarespace, and Google Apps. I have to admit that I was frankly very surprised that Facebook didn't offer this option to its users. The current FB username option is still weak because your username lives on FB's domain.
Facebook should have given people the option of redirecting and masking either their primary domain (i.e. timothypost.com) or a sub-domain (facebook.timothypost.com).
If Telnic moves quickly and smartly it could partner with dozens and dozens of web services to offer sub-domain redirects to the user account pages and masking with the user's own domain.
Ultimately, I think we will begin to see a service like Google Apps but for many, many web services. I have set-up my Google Apps account so that I can access my email account at http://email.timothypost.com, my calendar at http://calendar.timothypost.com, etc.
I want to do this for ALL my web services. One twist would be for me, the user, to name the sub-domain for the service, not the company providing the service. For instance, instead of http://linkedin.timothypost.com, I would name this sub-domain http://resume.timothypost.com. This would give me, the user, ultimate flexibility if, in the future, another better "resume" service came around.
Telnic could start by offering its users the option of redirecting only. It's my understanding that to truly mask the external URL one needs the cooperation of that company. Telnic could promote its developer API so new web services, say competing with LinkedIn, would offer masking as a competitive advantage.
This sub-domain redirect and masking topic is very important for two reasons:
1. User fatigue. One of the biggest hurdles facing web servcies today is that users are absolutely sick and tired of creating new accounts and then having to, once again, fill in their data.
Telnic could be the central place for people to enter all their personal, work, hobby, interests, etc. information with industry standardized Fields and when a user linked their .tel account with a new web service it would auto populate that information. People could "curate" their information at their .tel account accounts instead of in 37 different places.
A new technology from Google Pub Sub Hubbub would mean that when folks update their information at .tel it would immediately disseminate to their other accounts.
2. "Walled Gardens" not only are people sick of having to reenter data over and over but they are frustrated when that data is hard to export out of a web service. This solution with .tel would solve this issue because people's data would live in their .tel accounts.
This point raises another pertinent topic of today, namely where people's data is stored. This point was also raised in Robert Scoble's blog post.
Ideally, what I would love to see .tel do is a partnership with one of the companies offering a front-end interface to Amazon's S3 storage. In a perfect world, I would upload files, photos, etc. into my .tel account account and they would be stored at Amazon S3. Then, when I sign-up for a web service, such as Flickr.com, the web service company would access them at my S3 server. I already serve some photos to my blog using the S3 "Copy Path" option.
Obviously, many users would not want this option because they would have to pay for the storage. Nonetheless, Telnic should be offering this service to its power users who are also the thought leaders on the internet. Guys like Dave Winer would love this option and so would many corporations and organizations.
Telnic could actually create a revenue stream by doing a partnerhsip agreement with Amazon where it gets a commission for folks who access S3 storage through through .tel accounts.
Here's the URL for the su.pr set-up instructions for custom domains:
http://www.stumbleupon.com/developers/Supr:Shor...
I have a suspicion that su.pr may be saving the custom domain short URL to ALL external URLs as a paid premium offering. They are not saying yet but in their Help Forums there's been some speculation able this possiblity.
So, to wrap up, I am excited for .tel and the possibilities before it to truly reshape the web and to give the power back to the people.
PS: I'm curious, I left Network Solutions and shifted all my domains over to GoDaddy but I read that GoDaddy is not one of the registrars able to offer .tel. Do you have any idea on why that is so. I would think that .tel would want to establish itself as the ultimate personal domain and thus, GoDaddy scale and reach would be very important?
Great to chat with you. I have been meaning to put my thoughts down as a blog post and thanks to you, I've just done that.
I really like the way you're going with the OAuth thinking. Let's discuss OAuth options beyond 3rd-party app authentication when you've got time. There are a few .tel features you're missing that could make the opportunity even more interesting. I'm of course at henri.tel...
Speaking of Card.ly, it's a web-based service that displays your contact info. It's got a nice UI but doesn't solve the problems of speed of access (http is slow), consistency (what if they change their html?), resilience (centralized web service) and persistence (what if Card.ly goes under?).
We also wrote about the issue with URL shorteners in a blog article earlier this week and list as well hosted services as self-hosted options to run an URL shortener on your own domain:
http://iwantmyname.com/blog/2009/08/10-tools-to...
There's some microformats stuff going on with these shortlinks (the shortlink meta tag) but none of the Twitter client software work with those. It seems that there's not enough technical detail on what they have to do, and not enough software to gently handle the redirects with the proper headers and blah blah blah.
WordPress.com users recently got http://wp.me shortened links for each of their blog posts. It's nice to see that at least they care. Nevertheless people still post their username.wordpress.com/long-post-name-blah-blah on Twitter =)
RIP tr.im, I never liked you :-P
~ @kovshenin