Google Reader: Anti-Social Software

I’ve been a longtime Google Reader user, and I recently decided to explore their “Shared items” capability. The idea of “Shared items” is that you can mark posts interest you come across in Google Reader, and share them with your friends; and, vice-versa, you can view items your friends have shared with you. Pick the right friends, and your social network becomes an effective news filter, minimizing the amount of RSS feeds you actually track and read.

It sounds like a great idea, until you try to use the feature. The first step in any social networking-type application is simple: add your friends. If you got no friends, the whole thing doesn’t work. It would seem reasonable, therefore, that the first and most important aspect of such an application would be to make adding friends easy. It is in this regard that Google Reader not only hops on the failcopter, but grabs control of the stick, and jams it into a steep descent. Into the side of a mountain.

To add friends in Google Reader, you have to add friends in…GTalk? It’s hardly an auspicious start to the user experience when using the web application requires the user to navigate to another web application. And of course, to use GTalk, you have to use Gmail. Fine, whatever, I already use Gmail. In fact, I’ve imported about 1000 contacts into my Gmail address book, so the rest should be simple, right?

Wrong. Here’s the UI to add a user to GTalk:

Worst. Interface. Ever.

That’s right, you have to add users manually. In addition, there’s no autocomplete capability either, which means you’ll have to type in all of your friends’ complete email addresses. Who thought this was a good idea? It’s like the application needs human suffering to provide sustenance. Does this application thrive on misery?

Seriously, Google, come on. I’ve given you my email contacts. You even have a Google Contacts API that allows third parties to use my Gmail contacts! What the heck is going on here? In fact, this UI shouldn’t even exist – it should be a list of my Gmail contacts, filtered by those that are already using GTalk, that allows me to easy select a number of contacts and make the request. Done.

The lack of integration between different web properties is not unique to Google. If you use Upcoming, you’ll note that adding a user is a painful manual process similar to the Google Reader experience.

It’s like they actually want these applications to fail. If these providers can’t even integrate their own APIs to simplify the exchange of data within their own company, what hope does the DataPortability movement have?

Is OpenID Doomed?

Zoho's LogoZoho made an interesting move today by adding support for using a Google or Yahoo! account to authenticate to their suite of online productivity tools. I wonder what the OpenID community will think of this?

I’ve been following the OpenID space for the last little while, and this marks a curious turn in the development. OpenID has been promoting the idea of decentralized system that would allow users to minimize the number of usernames and passwords they need to login to the various web-based applications they use. It’s a great idea, but as I did some research for a client recently I concluded that the conflicting incentives for OpenID stakeholders may prove to be a barrier to adoption.

The core challenge is that while everyone wants to control users’ information by becoming an OpenID identity provider, there’s less enthusiasm towards becoming a relying party that accepts OpenID credentials. This is not especially surprising – controlling users’ information is the means that corporations maintain lock-in, and derive competitive advantage that they use to drive revenues. Hence, the move by many large web portals to act as OpenID identity providers, but not relying parties, has been viewed as an underhanded means to exploit the interest in the OpenID standard.

What’s curious about the Zoho move is that the company has obviously made the decision to accept third-party authentication credentials in a bid to lower the barriers to adopting its products. Google’s Docs and Spreadsheet offerings are a major competitor to Zoho’s offerings, so it makes sense to try minimize the pain of switching from Google to Zoho products. However, the decision to include Yahoo! accounts in the mix confuses things somewhat. Given Yahoo’s current problems, why would Zoho want to include those users? And if you’re going to go to the trouble of supporting yet another authentication scheme to reach a wider audience, why not go for OpenID?

Overall, it seems very strange that Zoho would exert the effort to support GAuth (used by Google) and BBAuth (used by Yahoo!). Both Google and Yahoo! are now OpenID identity providers, so why go down the path that requires roughly twice the effort required to support OpenID? You could do less work, and reach more users!

I can only guess that either this work began prior to Google and Yahoo!’s OpenID support was announced. However, there is one other troubling possibility: that while OpenID solves the technical problem, using a Google or Yahoo! account to authenticate to a third-party is more easily understood by users.