If you want a service that hacker news readers will use to find new friends, then listen to their advice about logins. If you want lay people to use it, Facebook login is clearly the way to go, which you'll learn once you start talking to people (e.g. soccer moms) who don't operate inside hacker/internet engineering echo chamber.

We are hoping to give people the best of both worlds, get started and opt-in to FB login.

Clearly, you'll want local login, and then integration to as many services as you can stand: fb, twitter, google, etc. There are some pretty robust solutions that are almost drop-ins that do this kind of thing for you.

Your feedback sample is HN; of _course_ these guys are going to dislike signing in with Facebook.

Fact is, for normal people (think: your mom, aunt betty, cousin Joe) Facebook login is easier and looks safer to them.

It circles back to the ancient proverb: Know your audience.

Not true. All the metrics I've ever seen for Facebook Connect usage are abysmal regardless of audience.

I worked on a project that had incredibly positive results using Facebook Connect. So I guess we're at one anecdotal data point on each side.

I don't know about the others, but the reason I didn't comment was precisely because I didn't want to log in using Facebook. Your giant trees were blocking my view of the forest.

Giving the choice between a third-party and a classic profile is the best way to go. It complicates things a bit but no one has a reason not to try your service out.

Lesson #1: Hacker News is not your audience!

Just forget about the nerds with a megaphone up their collective butts yelling about the evil Facebook and go for your actual market: everyone else.

I've considered GitHub and Bitbucket logins to be great options if you have a product that targets developers, but want to try and avoid (for a while, or forever) handling usernames/passwords/password resets/etc yourself. Both seem to have a general sense of "trust" that many do not have for Facebook.

I'm implementing Google & GitHub (w/ BitBucket to follow just after) OAuth2 flows on a small project I'm building now. Both services are extremely pervasive amongst the specific developers I'm targeting, so I consider users unwilling to sign-in with either of those services a minority (but one that I won't ignore, if they clamour for it!). I just had to draw the line in the sand somewhere if I'm going to get this out the door this summer (winter, here).

What does this say about HN -- that commenters here would rather troll and complain about FB login than to talk about the value behind a new social service?

Because nobody here is going to log into a service that requires Facebook, which makes it pretty damn hard to comment on it further.

I have no idea how criticisms over only allowing logins to a service through facebook could ever possibly be construed as trolling.

"Bait" someone, "troll"ing....to observe that demanding a Facebook login might not work for everyone?

Give me a break. Save the outrage for a case where it might be appropriate, however thus far the only troll here is you.

You might want to check out combosaurus.com . It's by the same people who run OkCupid and quite interesting. They are also quite good at predicting things I like based on the things I've already told it I like. What they don't have is a focus on actually meeting people.

Very nice writeup of the original thread and lessons to learn. Also I commend you for listening, learning and improving. I am not in your current markets but I'll be watching for when you expand to other areas.

Thanks for the kind words, hopefully we have helped improve our feedback channels even more and can expand to your area soon.

> "The email people used to sign up for Facebook was not one that they checked every day (which is especially important if that’s how you want to schedule events)."

That doesn't strike me as an issue that's unique to using Facebook Login. There are plenty of people that don't check their email on a daily basis.

Good point, what I meant was people use emails that they never check period.

I don't think you can really get a feel for the site without signing up for it. I wonder if you'd have gotten better feedback from HN if people could actually use it without signing up. This is probably a different issue entirely than relying solely on Facebook Login.

Not sure why you want to restrict your audience to the SF area. I can understand if you're shipping a product or delivering a physical service, but this is just people matching.

these are some valuable insights regarding something people default to without considering the tradeoffs.

I'd be interested to see a more in-depth analysis across products/companies and wider demographics regarding the disparity between an inferred personality (via likes) and actual preferences.

I don't know whether it's a mistake or not but I see a lot of new location/locality/city specific apps/startups coming out as only in city A,B or maybe works only in the USA. Though I'm sure majority of HNers are north Americans, most of these starups/services could actually open up to the world.

What happens is you see an app and from the outside it seems promising, you go inside and it's like "oh, great! USA only" or maybe "SF only" etc. One solution would be to just put a banner on home page "ABC only as of now" or whatever.