kineticac

We've been struggling with metrics for awhile now. We're hitting all the "enterprise" plans with our sessions + mau, and just can't afford it. Seems silly to build and maintain our own metrics, so glad we finally have Amplitude for all our needs.

The dashboard api is amazing. We only need a few simple numbers, and the API works fantastic.

Thinks like custom dashboards are cool, but we're able to just grab the numbers onto our own that we have melded with other services as well.

Just a bad guideline worded badly. I'd say this is far from the core value of what the service provides. I'd rather just hear:

"Quality products are guaranteed because of our extensive experience and knowledge of making these projects work with many dynamic circumstances." Ambiguous, sure, but that should be a good thing. Let's hear that they're constantly evolving and working around new challenges, and applying experience and learning from all the data points they have to tackle existing / mundane issues. The core should be just that.

Very interesting! As an avid photographer and advisor to a stock photo company, this is a great concept.

Where are you sourcing the photos from? And what do you believe your target audience would be?

Getting emails from CreativeMarket, for example, is not only a source of inspiration every day but also makes me think about redesigning the blog for the site, or maybe even swapping out icons or fonts.

I don't pay for the newsletter, but I can pay for the creative assets if I'm interested.

Have you thought of sending just teasers for free on a regular basis? Then if the user wants the photos they can pay the fee? I'd use it if it was like that.

Black text on orange was never great, so I'm with you on the white text in the header.

Overall, way too much orange going on. Y Meter is a good concept, but I think it detracts from the title of the post too much (being orange and all).

Are you going to do more versions?

Google Reader is perfect. Need implementation for Google+ pretty soon too...

this is really cool! after mentoring with the BUILD.org organizatin, I can only say good things about programs like these. I'd love to help but will be out of town that week! Keep up the great work.

Maybe I'm just really optimistic? =)

best answer yet ;)

Ooh thanks for posting this so fast! I actually didn't know the techcrunch article was out until I saw it here haha.

Can be done with javascript, comet and message queues in the backend, and only allows browser sharing. Joepestro and I did this a few years ago with pure js. worked even on mobile safari.

Ever have a new product that you have to release slowly, yet you are constantly asked to describe it? What if the current design and UI is changing so rapidly that screenshots and screen casts really won't be representative of the product? Illustrate your concept like you would on a whiteboard or piece of paper, and see if you can get your point across! Hope you like the video.

Just read the tc article. Looks really promising! A nice twist to Facebook event checkins + Foursquare checkins and blogging all in one. Right up there with what's really trending and working out. Excited to use it when it hits the California ;)

Finally someone is fixing public transport!!!! =)

congrats dotcloud! you guys deserve it! Glad we're hosted with such an awesome company.

Oh, an edge case: If facebook doesn't call your web hook to update your db on the newest permissions before a user revists your site, you may have some bad consequences.

Another edge case: say you have millions of users, only 1000 regularly visit the site. You would have webhooks for a million users updating your backend, rather than just confirming permissions with facebook whenever they arrive to your site. The hop to facebook is probably just as fast, if not faster, than checking in with your own backend. Let Facebook take the brute of the traffic ;)

Good question. There's a few things going on here. If a brand new user comes and you store his initial permissions and also setup the web hook to get updates on user permission changes (such as if she removes offline_access from facebook.com/settings/?tab=applications) then you can just query your backend (which would store this somewhere). Then you won't ever have to do this.

What if, for example, this is a pure client side application that doesn't store anything in a backend database? Everytime a user comes to your site, you can check to see if local storage has anything sure, but if they're on a new computer or have cleared their storage, you're in the dark.

I'll definitely be able to show you soon. We're finishing up a few final touches on our beta, but signup now and we'll send you a beta invite asap! We should be ready on Monday. http://feedtopic.com has a signup on the beta roadblock.

In the meantime, I can give you a quick description of what we have going on: each link that requires an extra permission not asked up front has a listener that will check what permissions are available. If it's missing a permission, we actually trigger a Facebox (from @defunkt's facebook lib). The Facebox defaults are really clean. Light overlay, slight shadow, nothing fancy. We put in different messages for different actions, one for example: "Yes you can like someone's post from FeedTopic! We need a new permission from Facebook to do this for you. Click the 'Add Permission' button to bring up Facebook's Permissions page". We have a button that looks SUPER clickable, like if you didn't click it you'd feel horrible because it has nice css gradients and looks like it's 3D. Underneath we have a message that says: "We won't publish anything without your direct permission to any part of Facebook, promise". Once you click the button, it will show you the FB permission prompt. yes, it's multiple steps, but it's clean and makes people feel really confident. BTW, liking something on feedtopic is probably the last thing people are going to the site for ;) It's just a small piece of info I can give away for now before the launch.

Feedback appreciated from Facebook API hackers!

great to bump into you at yc office hours today. let's chat more at some point, send me an email: art at fanvibe.com

Two quick updates I'm going to put in the blog post: Know that the server times are in UTC, so your cron schedule needs to be in UTC. Also you need the RAILS_ENV to be set to production in the scheduler as well as the worker.

Great work with Noteleaf! It saved me the other day when I had totally forgotten about a meeting I had. Keep up the good work!

Great product! You've basically built something every company tries to build when first launching a product.

Ever think about a badging system to go along with this? I mention this because this is similar to what we built to use on Fanvibe for awarding users virtual goods and real life deals on tickets etc.

thanks for the link!

Congrats guys. Dotcloud is awesome, the team is amazing! Even though we weren't hosted on Dotcloud, Solomon always was willing to help us out. That's the kind of service that comes from a company that you know will do well =)

haha that's true too. nice comparison dtran ;)

Forgot to mention, this is just for the NBA as All-Star weekend is coming up.

wow this is great!! congratz guys for the launch. I already spent a good amount of time on the site. Looks so perpetually useful, much more than things like Groupons which only work once in awhile.

This is a pretty amazing video. I like the tone it gives. It's respectful, yet informative. There's so many subtle things that makes it just perfect for the service. Huge props, you guys rock.

We've sort of just told everyone to do what they need. I worked through the holidays, but I'll be taking 3 days off next week. My co-founders took last week off and this week off.

Once you have VC's, are really raising money, have advisors, this is definitely the place to be. That doesn't mean you can't ever leave, or work in other places from time to time. If you absolutely hate it in the area, your product will suffer. If you also like to bounce around sometimes, and are very productive, you should do it. Bring your co-founders with you. Make sure they have the same mindset of working in different areas.

That being said, have a headquarter. Probably in Silicon Valley. Be a resident in the bay area. But by all means, don't subject yourself to never working elsewhere.