Re: [stp-dev] FW: [stp-newsgroup] IRC transcript from todays session

[Michael Norman <Michael.Norman@xxxxxxxxxxxxx>]

It may not be clear from my previous message what the precise issue is

Clearly code review is neither necessary nor sufficient to allow code 
committment within the rules set out in the charter.  It is not 
necessary, although it can be mandated by the PMC.  It is not 
sufficient, because it does not allow for vetoes by other committers 
within the scope(s) defined by the charter.

If we are to mandate code review we need to be very careful as to the 
scope over which we madate it. If we put in place any mechanism which 
allows committers from different companies to "buddy up" to exercise 
control over code committment  in a way that crosses (or defines subsets 
within) the component or project heirarchy, then those individuals (and 
their companies and the eclipse foundation) are potentially in breach of 
anti-trust law.  The thing that saves us within the project or component 
heirarchy is the detail of the committer election rules which, in 
principle, allow anyone to contribute and be recognised by their peers 
and become a committer thereby gaining veto rights within the scope of 
component.  There are also complicated rules for creating PMCs, Projects 
and Components which need to be followed.

Please also note every meeting (including discussions relating to code 
review) needs to invite everyone within its scope (project, component, 
etc.), to have an agenda posted in advance and to be minuted. 

In practice this stuff is best handled by committer vote at a weekly 
component or project meeting under the retroactive commit model.

Mike