[sr-dev] [RFC] Patches ready for merge

[Alex Hermann]

On Tuesday 08 March 2011, Daniel-Constantin Mierla wrote:
> indeed it was not something complex, it is why I thought it worth doing.
> However,my idea was slightly different to use a parameter for
> allow_trusted(), so you can have many calls of the function with
> different behaviour.

I looked at that option and I'm all for it (there are already enough functions 
dependent on 'global' parameters), but there are already 2 allow_trusted 
functions for 0 or 2 parameters. Adding another (1st or 3rd) optional 
parameter to those would make a complete mess of it, requiring 4 
implementations and forwarding the option through a series of functions.
Making them non-optional would not be backwards compatible, which was the 
whole purpose of making it configurable.
-- 
Greetings,

Alex Hermann