deal with rating system differences between tosdr.org and phoenix

The site has the following rating system:

points are rated good / neutral / bad / blocker and on top of that, a weight from 0 to 100
services are rated based on their points, as per https://github.com/tosdr/tosdr-build/blob/master/scripts/checkclasses.js#L29-L39:
- add the weights of all good points
- substract the weights of all bad points
- substract the weights of all blocker points
- substract an extra 100 for each blocker point
- if the result is <-500, e.g. 10 bad points with weight 50 each, or 3 blocker points with weight 67 each, the service gets class E
- if the service did not get class E, but there is at least one blocker, the service still gets class D
- if the result is between -500 and -200, but there are no blockers, the service gets class C
- if the result is above -200, but there is at least one bad point, the point gets class B
- if there are no bad points and no blockers, the service gets class A

Phoenix has a slightly, but significantly, different rating system, see https://github.com/tosdr/phoenix/wiki#rating:

I think points now also get good / neutral / bad / blocker, even though this is not documented on the wiki
As implemented in https://github.com/tosdr/phoenix/blob/master/app/models/service.rb#L40-L49 services get rated by assigning 0 to blockers, 2 to bad points, 5 to neutral points and 8 to good points, and then taking the average of that; 0-2 -> F, 3-4 -> D, 5-6 -> C, 7-8 -> B, 9-10 -> A

This means a services with only good points gets a B in Phoenix (getting an A is impossible? see https://github.com/tosdr/phoenix/issues/375)

tosdr / tosdr.org