There's a lot more than that meets the eye happening beneath the surface. When you have a thousand photos up there, you'll get a better feel of our algorithmic clustering.
There's at least some work by the algo to recognise different moments of the same day, I still need to upload more pictures to try to understand it correctly.
I understand but that's like saying you won't be using Google because they didn't explain the algo. Or you won't try the iPhone because they don't explain the algo that makes it scroll so gently. And so on.
It's free, you can try it if you really want. You can even use someone else's pictures. Asking for the algo seems too much.
As someone who runs another photo sharing site that organizes for you by date, I can say yes.
First, it's important to know that uploading a thousand photos is way easier on such a site, since you do no organizing work. Try uploading a thousand photos to Flickr without culling, organizing into albums, tagging and captioning, and see what your result looks like. It doesn't have to be like that.
Second, they have to be your photos for you to appreciate chronological organization. When I look at my own, I can say, "Wow, I just click 2005, then December and I see my Puerto Rico pictures. And I didn't have to organize them myself." You don't get the same effect looking at someone else's.
You need that many to understand the basics of the algo behind it? If that's how you debug your code than you might need to learn some better ways to do it...
Sorting by date is a breakthrough for normal people. I've been developing and advocating it for years. The big photo sharing sites make you sort manually into albums because that's what professionals and photography enthusiasts want. Photo sharing for the rest of us is just getting started.
I'm eager to see more about the algorithmic clustering, but the site's overloaded now.
What's the real breakthrough here?