Site "Picst (Picst)" that ranks Twitpic, Flickr, instagram etc. photos in order of Twitter popularity

Twitpic·Flickr·Instagram·Yfrog·PlixiFor photos and images that are published with services that you can easily share photos with, such as analyzing Twitter tweets and ranking this "Picst (Pict)"is.

Currently only Japanese tweets are supported, and you can view photos with Twitter's recent comments on individual photo entry page.

Details of what kind of images are available is as follows. It makes me feel happy just by looking at it.
Picst | Popular ranking of twitpic, yfrog, instagram etc. pictures
http://pic.st/


For example, at the moment the following image is in the top.

Picst | Station got 2 ch


Picst | Uh, I sat on a panda seat, I saw it, but whatever you think, my neck!


Picst | Chuo Dori "Forbidden catalog" "My sister" "God only" "Bakuman" trailer is stopping #akiba


Picst | blown www outside the store www [Rowan Yokohama West West Exit]


Picst | Because things that I thought as garbage are graduated, it is hard for Miyabi to cleanse up ... # mau2010


Picst | When opening the image folder, a hidden image came out and try to raise (sweat)


2011/01/23 23:18 postscript
Since the service seems to have increased, we added a note to that effect, and a comment on the technical background of this web service etc was published on the author's blog.

Twitpic and other photo sharing site ranking site "Picst (Pikst)" released! - Yuisubebe diary

Well simple technical explanation. How many tweets to service entries are shared is the method of using Twitter Stream API and Search API which are good at you. The problem is getting information such as the thumbnail URL and title of the photo, but this uses only the embedly API. It is exactly suitable for this service because more than 30 photo sharing sites can get information in eEmbed format following a common interface.

In addition, I also put a little effort into operation, this time I borrow VPS firmly with linode. Regarding the back end, we are using DBIx :: Skinny for O / R Mapper in the Web Application Framework for reason (stable excuse?) That Rumi, API which I have created recently is stable. When I made a mock, it was Catalyst + DBIx :: Class, but it seems quite resource saved.

So, if you find the topic of the current topic, "Picst!" Please see if it is good. Enjoy!

Sometimes it will be fun to watch it.