More Facebook Integration for SpokenWord.org

I’ve added a new feature to SpokenWord.org for Facebook users. When you submit a program to our database or you add a program to one of your SpokenWord.org collections, you’ll be given the chance to post it to your Feed (Wall) on Facebook. Note that this only works if you’ve previously logged into our site using your Facebook ID and your’e currently logged into Facebook.

CHI Conversations Launches

CHI Conversations is a new channel from The Conversations Network, the home if IT Conversations. “CHI” refers to Computer-Human Interfaces and this new channel initially features programs produced by BayCHI, the San Francisco Bay Area chapter of ACM’s SIGCHI. BayCHI has been recording the speakers at their monthly Silicon Valley meetings for many years, but those programs (many with extraordinary presentations by some of the great technologists of our times) have gone unheard by the general public until now.

Those recordings would have been lost forever if it weren’t for the efforts of BayCHI’s Steven Williams, backed by the support of BayCHI’s membership and Board of Directors. Steve has single-handedly pulled together all of the bits and pieces it takes to create a new channel on The Conversations Network. It has taken nearly a year of Steve’s efforts to get to this launch, and we’re indebted to him for his perseverance. Steve is now serving as Executive Producer of CHI Conversations.

Along with the members of TeamITC including new volunteers from BayCHI, Steve plans to publish their new monthly programs as well as work their way through the archives of past-year’s recordings at the rate of two or three programs each week.

Collection Limits for SpokenWord.org

Because SpokenWord.org collections can subscribe to feeds and even follow other collections, they can grow to a size that is unmanageable. We’ve therefore added three ways in which you can keep your collections under control.

  1. Limit the number of programs.
  2. Limit the age of programs.
  3. Limit the size of a collections RSS feed.

On your collection’s page, click the Info link under “Edit This Collection”.

1. “Remove oldest programs when there are more than [count] or [age].” The default value for [count] is 1,000, the maximum number of programs any collection can contain. If you want to keep your collection smaller, select another value: 10, 25, 100 or 250. As you add new programs, earlier-added programs will be removed in order to maintain the maximum size you specify.

2. Likewise [age] tells us how long to keep programs from the date you collect them. The default is never to delete them (by age), but you can change this to automatically remove programs that have been in your collection for more than one week, one month or one year.

3. “Most-recent programs to include in RSS feed: [count].” By default, we’ll include up to 100 programs from your collection in its RSS feed. But you can use this option to change that value to 10, 25, 50, 100, 250 or “all”.

Note: Although you can set all of these values now, only #3 (RSS limits) is operational. We won’t turn on #1 and #2 until at least Wednesday morning (July 15) at 9am Pacific time to allow you time to modify your collections that may be affected by the change.

Facebook Connect for SpokenWord.org

Yesterday I rolled out Facebook Connect for SpokenWord.org, and if you have a Facebook account I urge you to stop by, give it a try, and let us know if it works for you. The integration is about two-thirds done, but you probably won’t notice the missing one-third. It has been an interesting process so far. I previously implemented OpenID, and I expected something similar, but that’s not the case. The concepts of the two systems are similar, but the realities are quite different. For example:

  • Facebook’s documentation is awful. Rather than one or two coherent documents there are dozens of wiki pages written, as far as I can tell, by the developers themselves, not good tech writers. Each page is written in a different style and documents (usually incompletely) one small piece of the big picture. To actually integrate Facebook into an existing identity system, there are many — more than becessary — moving parts.
  • Although a FB user explicitly authorizes your application, FB refuses to supply his or her email address through the API. Instead, there’s a very Baroque system by which you send FB hashed versions of the email addresses of all your existing registered members in advance so that Facebook can then let you know that one of them matches a FB user at the time that user authorizes your application. But if a new (to you) FB user logs into your site, you don’t have that existing data. (OpenId’s API gives you an email address if the user approves.)
  • The Facebook Terms of Service are oppressive. They must have been written by Facebook’s Business Prevention Division. For example, you are not allowed to store (in a database) any personal data you receive from Facebook Connect. When a user authorizes our app, FB sends us the user’s first and last names. We’re allowed to display those while the user is connected, but not thereafter. (We get around this by asking the user to give us this data independently.) I noticed that TechCrunch uses Facebook Connect for comments, so I was curious what would happen if I left a comment on their blog and then de-authorized the TechCrunch app. Sure enough, my comments disappeared from their site, and when I re-enabled the app, the comments re-appeared. Weird.
  • The email thing is particularly nasty, for while we’re not sending FB our users’ emaill addresses unencrypted (which would violate our own Privacy Policy), we are sending an MD5 hash of those addresses. This means FB can compare the hashes we send them to the 100+ million email addresses they already have, allowing them to determine that someone is a registered members on our site even before that person authorizes the use of his/her FB identity to access our site.
  • FB requires that if a user is logged in via Facebook, you display that user’s Facebook photo on every page they view. No reason is given for this requirement, and very few Facebook Connect sites do so. (Digg is an exception.) Note that this (and other ToS issues) requires that you load FB’s supporting JavaScript on every page.
  • Oh, did I mention how bad their documentation is?

All of that said — and there are many more issues — we’ve had many requests for this integration as a way to make it easier to register for and login to SpokenWord.org. I hope you find it valuable.