Craig Newmark, Founder of Craigslist
FlickR
Video
Search for more photos:
from
from
from Microsoft
Video
Search for more photos:
from
from
from Microsoft
Labels:
america,
commerce,
Craigslist,
top100
About the Top 1000 Media Life Project
- We're testing the photokit platform - a cloud-based system with no dedicated servers and databases - true mash-up using JSON AJAX. With dependency on dozens of servers for every page, is it robust? Will it stand the stress of heavy traffic?
- Google Analytics and Feedburner track results. Privacy notices are the same as those services.
- We abide by normal use rules of photos on the web, using mostly thumbnails and maintaining attribution links to the source.
- We did quickerfy - reducing deployment from months to one week.
- We're comparing the performance of search engines - how quickly robots respond and how relevant their answers. Anyone can search using their name or any term. Already we've noted significant differences among the search results - rarely presenting the same answer on any query.
- Every page is dynamic, reflecting the current state of the web. We rarely use static photos.
- We've also defeated the client-side cache, but not the photo cache, for faster results. Click the buttons to refresh the photos.
- We don't control the selection of photos. Robots control the results. Please complain to the search engines for wrong photos. Thanks for understanding.
- We're asking the community to join us, outside of the comfort of Facebook. Will Google's FriendConnect work? Do we need to install Facebook Connect?
- We're creating a friendly competition to identify the top 1000 media personalities. Will the community help? Will the long-tail lost in the global crowd add our badge and compete to join the elite?
- We purposefully targeted a larger number, 1000. The web is global and there is more room for deserving personalities than just the small elite of the top 20.
- If needed, we can expand to 10k and segment into sections.
- The rules for placement are vague. We can be bribed ;-) Currently, the names are the obvious top 20 plus friends, particularly boomer-aged publishers. Others will be recommended by trusted community members.
- If needed, we can automate with some robotic method to fairly judge the results - although we don't really wish to rank people. It's a friendly competition.
- We don't expect advertising to generate significant revenues. We're wondering if there is advertising value.
- We're open to suggestions. Have any?
Enjoy!
Me
The search below finds 'me' photos - not my photos, but those that are tagged with 'me'. The results to vary day-by-day and moment-by-moment.
Labels:
about
Subscribe to:
Posts (Atom)