Who Shares Code With Artists - http://blairneal.com/blog/who-shares-code-with-artists/

star_counter.py

import requests
import json
import operator

companies = [ "adafruit" , "akqa" , "B-Reel" , "thebarbariangroup" , "cinder" , "BBDO" , "breakfastny" , "CLOUDS-Interactive-Documentary" , "deeplocal" , "digitas" , "morethanlogic" , "dreamworksanimation" , "fakelove" , "field" , "google" , "googlecreativelab" , "dataarts" , "MadSciLabs" , "HeliosInteractive" , "hellicarandlewis" , "ideo" , "ideo-digital-shop" , "labatrockwell" , "legworkstudio" , "local-projects" , "msopentech" , "MidnightCommercial" , "mpcdigital" , "otherlab" , "PixarAnimationStudios" , "Psyop" , "razorfish" , "redpaperheart" , "sapient-global" , "sapientnitro" , "sparkfun" , "stopp" , "tbwa" , "therumpusroom" , "warprecords" , "Vidvox" , "warprecords" , "wieden-kennedy" , "YCAMInterlab" , "Your-Majesty" ]

company_stars = {}

for company in companies:
	request_string = 'https://api.github.com/users/' + company + '/repos'
	r = requests.get( request_string )
	company_stars[ company ] = -1 #negative star count indicates invalid request
	if r.ok:
		repolist = json.loads(r.text or r.content)
		start_count = 0
		for repo in repolist:
			start_count += repo['stargazers_count']
		company_stars[ company ] = start_count

sorted_companies = sorted( company_stars.iteritems(), key=operator.itemgetter(1), reverse=True )

for name_count in sorted_companies:
	print( name_count[0] + ", " + str( name_count[1] ) )

@laserpilot I actually prefer the table without any numbers (as it is on your blog now), but this is interesting nonetheless. We're closer to a "this is valuable & useful code" metric.

Yeah definitely interesting, and much more representative of what people are actually finding useful from these places. I also need to find another way to include code from places that share but don't post explicitly to github - or individuals who develop at companies and post to their personal githubs sometimes (I'm guilty of this). Github is unfortunately not always the best place for all of this data collection, but I mostly just wanted to contribute to the conversation in some way.