TNG-stable

[Jeremy Allison]

On Tue, Sep 26, 2000 at 03:08:18PM +1000, Luke Kenneth Casson Leighton wrote:
> > code it. No other options are acceptible. And remember it has to be done
> > *professionally*. No memory leaks, no buffer overruns, full
> > I18N support (no ascii only code please).
> 
> i was waiting for a comment like this.
> 
> this is one of the reasons why i will not work with samba any more.
> 
> the standards are excessively high to be able to do any kind of
> incremental development.
> 
> open source projects are all about incremental development.
> 
> start off small, with something that just about does the job.  continue to
> do improvements, and continue to accept improvements.
> 
> the expectations of the primary samba developers have gone well beyond the
> bounds where it is possible for anyone to help except those people and
> their contributions that they consider to be worthy.
> 
> i spent three, maybe four years encouraging various people to contribute.
> that includes comments on APIs, specifications, documentation, bug
> reports, FAQs, and code. i can recall the following who have made various
> coding contributions.
> 
> steffan lauer.
> 
> elrond.
> 
> sander striker.
> 
> luke howard.
> 
> timothy cole.
> 
> danny breiss.
> 
> 
> all of these peoples' efforts, through insatiably high standards, have
> been rejected.  i did not realise that i represent these people, and i am
> sorry that i let you all down.
> 

I don't think these peoples efforts have been rejected.
The code in TNG is slowly being moved into HEAD, and then
into release.

I don't think we have excessively high standards. Remeber
we are shipping a program that is used on *millions* of
servers worldwide.

We are a trusted source of server software used by millions
of users, we have to have standards for production code.

That's why we were never able to say TNG was a production
branch, as it hadn't gone through the QA or testing that
the release branches do.

TNG was meant to be the "test it in a non production environment"
area - many useful advances have been made here, but not *all*
of the experimental ideas have been adopted. I think that is the
core of your complaint with the other members of the Team.

Remeber, the core of Open Source is peer review, many of the
ideas I and others have get passed by Andrew, JF and others who shout
"crap" loudly and they never see the light of day in a production
release. Often we open the process to a direct vote if it is a
contentious decision. You don't seem to like this process applied
to your code. 

But you have to be able to play well with others to get robust
code out there.

We've had this argument so many times, with me, with Andrew,
with JF, with Gerald...... but the common factor is you always
being on the other side. Please think about what this means
w.r.t. peer review of your code.

We don't have insatiably high standards - we just have standards
full stop. And we can't drop them for anyone - not even you, sorry.

Regards.

		Jeremy.

--
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------