[gPXE] Changes to Etherboot Project roles and responsibilities

[Marty Connor]

Michael Brown wrote on 3/19/10 10:11 PM:
  ...
> I will compromise on respecting your process where it affects anyone
> else's contributions, despite my reservations, but will continue to
> push my own changes directly.

So basically you're saying that, after abandoning the project for
close to a year with no warning, and after ignoring emails and phone
calls, you have now decided to swoop back into this project and
think it's appropriate to set policy and overturn procedures that
were put in place largely to deal with your own unavailability and
unreliability over the past couple of years.

No.  No more compromising.  No more negotiations.

You have repeatedly abused the privilege of being lead developer for
this project, and thus you have lost that privilege, along with a
healthy measure of my respect in the process.

Therefore, in summary:

- You are no longer lead developer for the Etherboot Project and gPXE.
  You abandoned that position over nine months ago.

- You may no longer commit to the main gPXE repository.
  That role is part of a review process that you've declared you will
  not follow.

- You will not be assigned a student for our GSoC 2010 engagement.
  You have repeatedly neglected students you agreed to mentor.

I am taking these actions in public, and will defend them in public if
I must, because you are using the fact that I have kept private your
abuse and neglect of this project to spread misinformation.  I want
there to be a public record of what I am doing and why.

In the past two years you have become increasingly unreliable as a
contributor to this project. You have disappeared and become
unreachable for ever longer intervals, ignored your mentoring
responsibilities, and caused enormous amounts of additional work for
myself and other project members.

During this period I have repeatedly (when I could reach you)
communicated about the problems your disappearances and unwillingness
to work cooperatively have caused.  Your responses have been to
suggest that the project should adapt to your preferred way of doing
things. You speak about working collaboratively, but your actions
loudly and repeatedly demonstrate otherwise.

Because of your programming skill and past code contributions, and
because I wanted to believe you would change, I have done my best to
cover for you publicly, in the hope that you might improve your behavior.

Then, last year, during Google Summer of Code 2009, for the second
summer in a row, you abandoned your mentoring duties almost as soon as
the program started.  This meant that for the second summer in a row I
and other mentors had to assume extra responsibilities to cover for you.
We were able to manage, but it was grossly unfair of you to commit to
being a mentor and then to disappear.

I'm not willing to do this again, and can no longer in good conscience
ask others in the project who are reliable, consistent, and team
oriented to make up for your shortcomings.

Your desire to commit whatever code you want, without review or
approval, essentially paralyzes the project because nobody else can
develop code without worrying about how it will collide with what you
might be developing. That process, along with your frequent neglect or
dismissal of other peoples' patches, slowed the project significantly.
That is a completely non-collaborative process that dis-empowers and
demoralizes everyone else.

During your most recent multi-month disappearance, due to necessity, I
began the process of building a development team that could move the
project forward without your involvement.  Together we have been able
to support and extend gPXE, and to work cooperatively, and in a
positive fashion.

It's been a refreshing change to discuss changes to gPXE in a
collaborative fashion, with an eye on building a solid set of features,
rather than our old model where you would throw in whatever code you
had recently been paid to develop, in a dump-and-run fashion.

We no longer are hostage to your technical skill.  We are a
project that has enough technical ability to maintain and extend our
code without your help.  We have increased our "bus factor"
significantly.

You're a great programmer, and I've always said so, but this project
instead needs talented people who are also communicative and
collaborative team members, which you are not.

And then there is this from you:

> I respect your willingness to coordinate the contribution of patches
> to gPXE, and am glad that you have now found a role that allows for
> you to contribute in ways directly connected to the actual code,
> besides the numerous ancillary tasks you have kindly taken upon
> yourself in years past (such as Linuxworld organisation, GSoC, OSUOSL
> hosting, etc).

This paragraph is so full of willful deception that I hardly know
where to begin.

When you publicly attempt to suggest that my contributions to this
project for 10+ years have been purely bureaucratic and
"ancillary", it is definitely time for you to find another project
that will appreciate you more.

I have done everything in my power to promote you, fund you, and paid
expenses out of my own pocket in order to position you to get lucrative
contracts, while all the time raising the profile of network booting and
gPXE, which also benefited you.

This was in addition to developing the concept of gPXE in the first
place, developing and continuously updating ROM-o-Matic, and
contributing code to both Etherboot and gPXE.

Since you have repeatedly demonstrated that you are not interested in
working in a supportive, collaborative manner that respects all
members of the project, please fork or find another project to work on.

This is one of the most painful decisions I've ever had to make, but
my primary concern remains the continued health and well-being of the
entire project, and I firmly believe in this course of action.

/ Marty /