I’m a big fan of user library software. I bought a license to Delicious Library as soon as it came out, and then undertook a process of hacking a number of CueCats to assist in scanning my book, CD, DVD, and game library.

Once I began that quest, though, it struck me as odd that my library was tied to one computer; I couldn’t access/update it online, or share it with others. This led me to LibraryThing, less UI-candy but a powerful online library tool nonetheless. Further, LibraryThing helped and encouraged folks to use CueCats to input their collections, and seemed focused on open standards.

Enter: Shelfari.

I got an invite to Shelfari back in July from my friend Lowell. There seemed to be a level of design professionalism around the product, and it was backed by Amazon. Cool. Also, it looked like they might be trying to move towards a more visual approach, ala Delicious Library. I created a Shelfari account, added some books, and then largely forgot about the product.

Enter: Facebook’s F8 Developer Platform.

With waves of 3rd-party development of social profile apps, a number of book cataloguing apps appeared. Shelfari’s Facebook app appeared at one point in my News Feed, and I decided to add it to my profile page. After adding it, I went to Shelfari and added a couple other books I owned to make my profile module look a bit more impressive.

Then, I made a mistake. I opted to let Shelfari look through my Gmail address book to see who I was already connected with might also be using the service. (I have a general policy when it comes to inviting others to a service I belong to: if they are already a member, I have no problem asking if they’d like to link up; if they’re not, I may invite one or two folks who I think would be very interested. Otherwise, I either mention the service in person, or forget about inviting folks to it.)

I entered my Gmail login credentials, and pressed the Continue button.

The resulting screen showed a half dozen some users (from my address book of 800-some) as members of Shelfari. All the addresses were checked. Just to make sure, I scanned the page first… I didn’t want to spam everyone in the list. There were two sections of addresses: “Your Friends already on Shelfari” and “Your Friends not yet on Shelfari.” Each had a “Send Invites” button, and a toggle to select/unselect the addresses within that section.

Pretty clear.

I hit the “Send Invites” button beneath the “Your Friends already on Shelfari.”

Seconds later, I got the first of several mail delivery errors in my mailbox. It appeared Shelfari had emailed some of the addresses from the section whose button I didn’t press. I was alarmed, but not too worried.

Then came the acceptance emails. Emails from my mother-in-law, my dentist, past clients, eBay sellers, and folks on mailing lists whose addresses I didn’t even have. By the very fact that I had received mail from an address, Gmail had stashed it in my address book; that I found useful. What I found harmful was that, now, all those addresses (especially all those addresses of folks I didn’t ever even email… they had emailed me) had been pinged by Shelfari by me.

Should I send an email to everyone in my address book to apologize? That seemed even lamer.

So I channeled those feelings into an email to Shelfari support: “It’s bad enough to auto-check them all in hopes of spamming the whole list, but to then tie those auto-checked addresses to a disconnected button?” The response I received back (rather quickly) was infuriating. “That certainly is disconcerting,” it started. But then it took an odd turn. “We have actually evaluated numerous designs for this process and have chosen one that we felt was extremely clear explaining the process and what is happening.”

So while acknowledging what happened was not in a user’s best interest, they defended the design of the system. This struck me as something usability testing (even paper prototyping or casual man-in-the-field testing) would catch easily. Qualitative feedback captures gems like this in a way that raw, aggregate data can’t.

I replied again, but never heard further. I grumbled to myself, and to several co-workers.

And then, finally, this past week I came across O’Reilly Radar’s ‘Shelfari and the New Social Contract’. I wasn’t alone. Dozens of other Shelfari users were as angry as I, LibraryThing’s CEO had done an investigation (and expose) on the exact issue, and now O’Reilly was drawing attention.

Shelfari’s response seemed plausible on first read. And then I thought about it. Their new fixes, while solving my particular problem, still will result in unwanted, unsolicited emails being sent. (They choose to pre-select all email fields still; a convention long ago dropped by most companies wanting qualified members, not those too lazy to hit “unsubscribe” from the ensuing emails-to-come.) And, further, the pattern exposed (of their intern who hid under the guise of a happy fan while interacting with co-workers in comment threads) suggests that they don’t get it.

Long story longer, I’m still a Shelfari account holder. Part of me feels bad bailing on a service to which I inadvertently attuned so many people. Another part of me wants to believe this was just a couple of bad things coupled together, and Shelfari will learn from the mistakes.

But honestly? I’m doubtful. And it makes me want to use LibraryThing more. (I appreciate their allowance for site publishers to include their own Amazon Affiliates link in their widgets.) And the upcoming Delicious Library 2.

So, in the meantime, while Shelfari’s invite process isn’t on the up-and-up, I figured I’d publicly join in with the kicking.