Al Hopper wrote:

> If the project is good enough to integrate - then - step aside and let 
> it integrate.  If there is a "philosophical argument" - then follow your 
> own procedures and file a bug report against ksh93 *after* it
> integrates.

That is most assuredly *not* our procedure.  You don't get to fix stuff 
after integration, you get to fix it before.  That's why code reviews 
take place before integration and not after.

I could understand your gripe if the bar Roland was being asked to jump 
was higher than that set for anyone else, but it isn't.  I know it's 
difficult for people outside to believe, but we get exactly the same 
treatment internally, and I'm saying that both as a 'victim' ;-) of 
meem's reviews in the past and as an ex-S10 cteam member.

This isn't arbitrary - the gate machine is going to build Roland's stuff 
the night it is integrated, and the following machine it will upgraded 
to the previous night's build and will be *running* his bits.  Lots of 
other people will be pulling over his bits and building and running them 
too.

> You're being obstructionist and blocking ksh93 integration. In my book - 
> you're either part of the problem or part of the solution. Where do you 
> want to be in terms of getting ksh93 integrated (rethorical question - 
> don't answer).

He's being a thorough code reviewer, no more, no less.  Honestly.

-- 
Alan Burlison
--

Reply via email to