A suggestion for the 'rules'; experimental/testing branch

[Daniel Dickinson]

Hi all,

A personal aside; I seem to be doing better so hopefully I can be a
positive factor in this project.

Anyway,  I read the rules and have one suggestion:

There is no mechanism for removing problematic committers.

Problematic can mean one person who makes the project miserable for the
rest, or it can mean someone who commits too many untested patches.

I understand the urge to approve a seemingly innocuous change, because
properly testing everything takes time, of which we are all in short
supply, but I think it would be good if committers were held to the same
standards as patch submitters (which is something I saw going awry at
times in openwrt).

Having an official policy in this regard helps committers avoid the
temptation to take shortcuts (I know this is a failing I have to work
hard to guard against myself) and assume a seemingly minor change, or
quick test, is sufficient.

Certainly having my own tree that I don't feel as much urgency to try
and get my stuff 'out there' has helped me, and I think the text
encouraging the practise is good.

One other thing that I think would be helpful is to have an experimental
or testing branch to which one could submit RFC patches, if one needs
help getting patches tested.

Regards,

Daniel