This project contains a database model (one table, one proc and a view which is optional) and the .NET code to generate an EXE

You will need to create the table and proc in some SQL Server database somewhere. SQL Sever Express will work if need be. The view assumes 2012+ as it has a lead/lag call in there. This is simply an interface for quickly finding change and can be overcome with a less elegant query.

If you use the .exe in the Release folder, edit the ReputationMinder.exe.config to point to the SQL Server instance. You will also need to edit the list of user ids and URLs.

Finally, you will need to schedule invocation of the executable on some regular basis. I run every 10 minutes, you can dial that up or down as you see fit. Just be cool

Note that the userids reset per site so your ID on one site will be different, even on meta.

Future changes, if anyone finds this useful, would be to
  1. factor the source site into the reputation table
  2. make the table not be a heap
  3. cloud, because everything is cloudified

Yes, crawling the site is legit http://meta.stackoverflow.com/questions/84352/can-i-crawl-stackoverflow-articles

Last edited Dec 18, 2013 at 4:01 AM by bfellows, version 1