How not to add new code to the tree

(1) File a bug.
(2) Write up a demo and evangelize on why this is a good idea.
(3) Spend two weeks turning your demo into a working piece of code.
(4) Write up a spec because someone asks for one.
(5) Spend another week fixing your working code to conform to the spec, and start tweaking the spec.
(6) Write documentation on the whole thing.
(7) After doing steps 1-6, only then do you ask if the module owner would actually like to have your code checked into the place you’ve been writing your patch for. UNTIL THEN, assume you can do whatever the hell you want.
Most importantly, never, EVER read http://www.mozilla.org/hacking/new-features.html . They don’t know what they’re talking about.
On the other hand, if you want to add new code to the tree, it might be wise to think about more than the code you’re writing… <alex:grin type=”rueful”/>

3 thoughts on “How not to add new code to the tree”

  1. There’s nothing wrong with new-features.html, at least from the C++ coder’s pov. I was being humorous about my own mistake of not reading it.
    Personally, I think a lot of the hacking documents at mozilla.org are oriented for the C++ developer. Many of them don’t seem to apply the same way when it comes to XUL code.

Comments are closed.