I’m currently refetching items. That takes a lot of time.
What have I changed:
1. The new db Model (hopefully) supports better scaling performance
2. The userid will now be the Google id and not just a random number.
3. I can now do some new language filtering tricks. It’s lickely that the language filter gets worse at first until I have time to build some algorithms.
4. Double counting items from a user should happen no more. From now on I really know exactly what you guys are sharing 😉
5. To be honest: When the site leaked I parsing tags was not integrated. That’s why no such feature has been available. I’m now parsing them and will soon release something like a tag cloud.
6. When I solved scaling (could take some time) I will start fetching shared items as far back as possible.
Well, so far so good. The parser really has a hard time re-fetching all those items, so please be patient with me…