> From: Raju Bitter [mailto:rajubit...@googlemail.com] > Sent: 11 January 2012 18:59
>> I will repeat my cautionary statement that we are working in a >> constrained environment and will likely need to cheat on occasion. >Right, we only need to know when we should cheat. :-) I got the impression from earlier threads that many (most?) folk were in favour of reviewing all commits. If this is the case, then rule on cheating is simple: cheat only when you think you can justify it to others in a review. If you can't justify it, don't do it. For example, if we have a standard that requires methods be, for example, less than 10 lines long, if I committed a change with a new method of 20 lines, I'd have to argue why it makes sense to keep it that size, rather than split it down into two or more smaller methods. (this is just an example, not a standards proposal :) David.
