[gPXE] Changes to Etherboot Project roles and responsibilities

[Michael Brown]

On Tuesday 30 March 2010 03:58:48 Marty Connor wrote:
> > You didn't develop the "concept of gPXE". You chose the name "gPXE".
> > That was, essentially, your sole technical contribution in the
> > evolution from Etherboot to gPXE. That was all that you did.
>
> Actually, I did a lot more than most people know, and you are using
> what is not known publicly to suggest that I have done much less than I
> have.
>
> In 2004, over a year before I initiated a fork of Etherboot called gPXE, I
> determined that the continued viability of Open Source network booting
> required us to support the PXE industry standard.
>
> <snip>

Yes.  You did indeed fund the development of PXE support within Etherboot 5.3.  
I tried to give you full credit for funding this work of mine at the time but 
(for reasons that I still don't understand) you refused and I respected that.  
If you want to add a funding credit to the current Etherboot tree, I won't 
object.  (Practically none of that code survived into gPXE, so it would be 
inappropriate to claim that you funded any of the development of gPXE.)

A year or so later, you decided to fork off Etherboot and chose the 
name "gPXE" for the new project.  Your technical direction was that gPXE 
should be a "pure" PXE stack, following the specification as exactly as 
possible.

While that is an interesting idea in its own right, it bears absolutely no 
relation to the product or the code that we now have.  What we have in gPXE 
is not the result of your creative vision or technical direction; it is the 
result of mine, and of my hard work over many years.

> [ For those interested, here is a paper that contains a concise history of
> the Etherboot Project:
>   
> http://edgyu.excess.org/ols/2008/H%20Peter%20Anvin%20-%20x86%20network%20bo
>oting%3a%20integrating%20gPXE%20and%20PXELINUX.pdf ]

And for those same interested people, here is an earlier draft that you wrote 
of that same paper:

http://www.fensystems.co.uk/~mcb30/gpxe_history_according_to_marty.pdf

Out of curiosity, I decided to tabulate the number of times that each 
developer's name is mentioned in your original draft:

5     for Markus Gutschke (who originally created Etherboot)
1     for Jamie Honan
1     for Martin Renters
5     for Ken Yap (who was Etherboot project leader for 9 years)
1     for Rob Savoye
3     for Jim McQuillan (of LTSP)
3     for Don Becker
9     for myself
4     for H. Peter Anvin
1     for Nikhil Rao
       ...and...
14   for Marty Connor!

I leave it as an exercise for the interested reader to compare and contrast 
your first draft with the version that was eventually published, following 
complaints that your first version of the Etherboot project history made it 
sound like "The Marty Connor One-Man Show".

As I said, I have for years sat quietly by and bitten my tongue at your ever 
more audacious attempts to claim credit for my (and others') work, but I now 
think it's important that people realise just how brazen you can be.

> > I have for years sat quietly by and bitten my tongue at your ever more
> > audacious attempts to claim credit for my work.  As I have confided to
> > some of the other people on this list, this has been one of the primary
> > reasons for my spending ever more time away from the project; indeed,
> > last summer I was seriously considering announcing that I was quitting
> > for good, because it was just too demoralising to carry on dealing with
> > your ego.
>
> It is my observation that you started communicating that my leadership
> was a problem

You seem to be refuting a point that I haven't actually made.  I didn't say 
that your "leadership was a problem"; I said that you repeatedly attempt to 
claim credit for my work.

> > As the person who actually developed gPXE, and who remains the majority
> > copyright holder, I would say that the responsibility to fork is yours,
> > if you wish to do so.
>
> No.  Whatever copyrights you hold, all the source code is licensed
> under the GPL. It is not your personal code; it is a FOSS resource.

We're going to have to agree to disagree on this.  Of course you own the 
physical machines and domain names, so there's little I can do to stop you 
betraying the absolute trust I placed in you when I accepted your setting 
yourself up in such an easily-abusable position of power.

I am holiday at present, but shall respond to the remainder of your message 
upon my return.

I am sorry that it has come to this, and I'm also sad that you have totally 
ignored the private message I sent to you a week ago to try to make peace.

Michael