DISQUS

In addition to everything that has already been mentioned, they could also potentially mine all those contacts they scrape from Gmail. Y'know, after people hand over their Gmail passwords to go scrape them from Google's servers. And of courses all the other places they import contacts from.

I think the most likely to actually show up there though would be people that were invited but haven't registered.

well, the new semi-private "rooms" part of Friendfeed differentiates it in a new way, no?

Oh wait. This part might be more idiotic:

"Within two hours, eighty percent of the school's male population had voted, and Harvard's computer system overloaded, effectively shutting down as the University's bandwidth was overwhelmed."

yeah, right.

"a school-wide search for the creators of FaceSmash led to campus security literally storming into Eduardo and Mark's dorm room."

yeah, right.
(anyway, wasn't it facemash?)


...and then it just gets worse and I can't even comment

"Eduardo, using the algorithms he'd created to track and rate weather patterns across the nation, helped construct a program to order the hacked pictures"

This has to be one of the most idiotic things I've read this year.

My guess is that facebook's take on twitter-like usage will come from the redesigned feed (after all, they're merging it with the wall).

Making that feed more real-time will make it more twitter-like and friendfeed-like. Public conversations naturally fit a feed that shows you history.

Actually, it may be a bit of both - they can just make status updates go into their feeds.

The main element I would wonder about is how to track back-and-forth exchanges. Friendfeed has replies nested under feed items. Facebook has "wall to wall" currently. Twitter has nothing except the @ convention, which people navigate manually. Wonder if Facebook has an improvement in mind in this respect?

It looks like exactly what Facebook's blog says: Google just used their public api and ignored their terms of service.

Nevermind, strike that. It just turned on for me in the last three minutes.

I'm going to have to contradict that. I'm on the Denver, CO network, and facebook chat is not turned on as of now.

I only knew about that after a friend told me about stories she was seeing that I'd been removing for my mini-feed, and I totally agree.

Wasn't the reason they gave for not making the newsfeed public that people wouldn't be able to pull items about them off of the feed?

Are you serious? Google App Engine is Huge. And it doesn't really compete with facebook at all.

App Engine and S3 are options for hosting a distributed application on a 'cloud', taking away the details of running a bunch of servers yourself.

App Engine gives an incredibly easy way to write and deploy an application that can scale to millions of users in no more time than it takes them to propagate it across their network (minutes? hours?) No design headaches, no complex master/slave database failover setups with memcached servers and the rest of it. No headaches managing the fundamental disconnect between an object-oriented data model and a bunch of relational tables. And all in plain, open-source python. Write it and push it out to the world in 10 minutes.

I won't be at all surprised if open source packages come out soon that mimic App Engine's API, since that's almost certainly a good model to design to for scalability. Other services may crop up to compete with App Engine, so that the app you write could be dropped into any or all of them.

App Engine basically lets you handle http requests and use a data store. If you need more control over the architecture than what Google's easy offering gives you (say, heavy backend data crunching) then you should look at EC2. They give you root on all of the machines you care to buy time on, and you can easily shrink or grow your cluster. It's not as easy though, since it's up to you to administer the machines. It's different enough from App Engine that in most situations there is probably an obvious answer as to which is better for the job.

For an easier time on S3, there are also of course a whole bunch of startups that build a service that runs on S3 and then offer their users something easier. Just to pick one example, heroku.com lets your write a ruby on rails app right inside your browser and then deploy it in S3. Pretty sweet.

I don't understand why App Engine would be considered a threat to the facebook platform. You don't host an app inside facebook platform, you host it outside and communicate using standard web protocols. If anything, App Engine provides a new, easy way to write and deploy a facebook application that won't have any trouble scaling to facebook's entire user base, should it catch on. That's good for both platforms.

Yes, you can tie an App Engine app to Google accounts, but that's optional. You could use OpenID instead of Google accounts for an App Engine app. You just need to include the code to do it (hopefully a library will make that dead simple).

Was there. Good times were had by all.

Between your video and this interview, it looks like you've been a good source of info lately. Nice going. I imagine your hitcounts have gone up...

As far as developer feedback, I am a developer and I'm working on an application that I believe is very much in line with what Zuckerberg wants to see people building (I have yet to release it).

What I would ask of facebook's platform team is to please allow my application to store user data, after showing the user a very obvious button and telling them exactly what they are opting into. I don't want to trick anybody, and I don't want to store anybody's data if they don't want me to. What I want is to let them set things up to seamlessly work with a single click, without having to go somewhere and register separately for yet another web app.

Quite simply, I want to provide a good user experience when registering to use my application.

Know what would be cool? Some kind of crowd mood monitoring system that crawls the twitter feeds, does some text analysis, and controls some simple display up on stage. Could be as simple as a mood light that glows red for bad and green for good, and gradations in between, so you can watch what the mood of the crowd is doing.

Or a tag cloud, or the full text of posts...

This isn't what I expected, after reading all the negative buzz. Maybe that's due to lowered expectations?

Here's my sense of it as I watch it now for the first time:

Starting right in, some awkwardness. The story about him sweating him hit a bad note, the stuff about his age gets old, and the French announcement was a flub.

But, the rest of the first half wasn't bad. Seems like she's interrupting mostly because she wants to cover a lot of ground. He's actually talking quite a bit throughout most of the first half, not very shy at all. I've been following facebook's progress for a while now and I'm getting a lot out of it. He's giving a great sense of what he's trying to do with the site, and I feel like most of the people commenting are just not reading into it the things that I am reading into it.

Things seem to take a turn about halfway through when she interrupts to ask about music. She's condescending more now, and that is what's really losing her points. After the remark about 45 minutes, it slips some more.

"Presale on amazon", Hackathon "disgusting" - now she's sinking fast.

The 15 billion dollar talk comes next. He's not interested, the crowd isn't particularly interested, and here is the hair twirling, impossible to ignore. This crosses some kind of ditz line.

After all this 'efficient communication' and 'user control' talk, I am wondering why we haven't touched on the obvious question of when the site will put the public/work/family/school separation of profile data front and center. That would make facebook suitable as a business network and a friends network at the same time, and I'm betting it's the first thing most users would want to ask him.

Not that the crowd is hating her for the sins above, the little things start piling up, unforgiven. I imagine the twitters serve as a huge amplifier. People are not thinking, "does everyone else think this is going bad?". They're seeing proof that everyone else thinks it's going bad, and it's making them all a little bolder by the minute.

Now at 38:12, another nonquestion elicits a simple "ok". This might be fine for a lot of subjects, but by now we all know that Zuckerberg won't take that and run with it - 'you have to ask questions'. Her awkward laugh as she crosses her legs is like a parody of herself. Things build some more and the crowd is definitely laughing at her.

Telling him to read her book, digs a little deeper...

41:46 Giggling for no apparent reason? Gets an audible reaction...

Around 43:00, with the "yeah"s and "mm-hmms", she is obviously just waiting for her turn to talk... (happens several times throughout)

Now with the major softballs, "is [beign CEO] hard for you?" etc... noboby cares, although it actually gets a good answer in my opinon

Now the questions are getting better again, and the answers are good too. I think people at this point have lost the patience to listen to her talk and it doesn't matter much what she says.

By the way, it's funny how often Mark says, "empathy"


"One thing a lot of people don't know about Mark..." ok now we are headed back downhill...

The trouble with this story about the books is that she just wants to tell the story, so he sits back and listens. And there's the hair twirling again. She realizes that it's his interview so she tries to get him to pick up the story by saying, "right?" And as she should know by now, you shouldn't do that with this subject.

I heard an episode of "This American Life" titled "Fiasco". In it, they describe the tipping point moment, where a situation turns from an ordinary disaster in to a Fiasco.

The obvious candidate for tipping point comes right here (49:00), when he tells her to ask questions and the crowd cheers for half a minute.

But the real tipping point comes after right after 51:35 ("you made that up!") when she yells to the back of the room to try and defend the claim that Mark burned his books. As she would say, WTF? Now the crowd is openly heckling her, and it's all over.

Now now, you may not like a particular product or service, and that's fine, but isn't it a little petty to call its user base "undeserved"?
:)

Wow, if that's the best they can do then this is a joke. That looks like a basic use case outline for every web form driven search routine ever written. And the rest is worse for them of course. Stupid jocks.

> I’ve had developers who’s livelihood rely on the success of
> their application contacting me to express how these new
> changes will negatively impact their application.

Serves them right for spamming the rest of us and (probably) ruining lots of opportunities for others. If their applications were good to begin with, they shouldn't have any trouble.

Good. If they don't do this, I think they'll find themselves in decline as users stop seeing the "utility" they're offering.

Yes... and no. This is what dataportability.org is about. Yes, someday you'll likely be able to do this things but, if data portability gets the support it deserves, you'll be able to use some other service if you prefer. And you'll be able to make the services play nice with one another so that things "just work".

Where is myspace on this list? It's still the biggest one, and it's by far the biggest one in the US. Strange that they wouldn't include it in this chart...

I feel oddly satisfied in regards to my earlier comment on this thread. Now that facebook and google have joined DataPortability.org, I feel like they listened to me, just five days after I wrote that.

Of course it wasn't a good integration.

A "Facebook" debate would work well if done similarly to the youtube debate. Have people submit questions before and during the debate. Have a ranking system working in real time that lets users bump the best questions to the top of the heap. The realtime nature of the system could allow followup questions to be quickly thrown back at a candidate later in the same debate.

Apply instant community judgements of candidate responses. If a guy avoids a question and gives an irrelevant response, in the judgement of the community, have a big neon sign come up somewhere that says, "Avoiding the question!"

And since it's facebook, it'd be interesting to use the social graph somehow. Sure, you could have it show a user what his friends are saying, but I'm not sure how a social effect would get to the candidates...

Ah, well yeah, a script shouldn't be allowed to just crawl the site like that.

But I am sticking with what I said about facebook applications and users' data. I'm hoping people will back me up on this and get facebook to make this change.

It's the difference between a walled garden and an open web. It would essentially solve the multiple login problem and establish facebook, and other similar platforms, as providers of an online identity that other sites can tie into.

I think a facebook application should be allowed to import any and all of a user's contact information into its own database for permanent storage...

PROVIDED THAT THE USER THE INFORMATION BELONGS TO EXPLICITLY ALLOWS THE APPLICATION TO MAKE A COPY OF HIS/HER DATA

The information is the property of the users. Facebook is simply keeping that information with its users' implicit trust that it won't abuse the privilege. Facebook should not release this information to anyone without the users' consent.

Similarly, if users want an application such as Plaxo to have their personal information, it is not Facebook's place to stand in the way and force the user to reenter it.

I think the options presented to the user when s/he adds an application (and later in the application preferences) should give users the opportunity to make clear what they prefer to share with applications and whether they consent to having the applications store that information long-term. The default settings should be chosen to protect the user, but the user should have the final say.

Interesting...

Going simply on the naming conventions used, I like the idea of a standard "Social Network Markup Language" rather than "Facebook Markup Language".

You appear to be suggesting that Facebook is going to switch to this more neutral-sounding name when you write, "Facebook is going to be the primary competitior to OpenSocial but rather than using the often critized FBML, they will use SNML and SNAPI to take OpenSocial head on."

What are you basing that prediction on?

I agree it would be nice, since the alternative is a world where application developers have to spit out one set of FB* tags for facebook and a set of SN* tags for any site that adops SNAPI. I like that name by the way; it's snappy (sorry, couldn't resist).