[gPXE-devel] [PULL] ipxe-sync

[Michael Brown]

On Thursday 08 Jul 2010 00:09:17 Marty Connor wrote:
> Quoting numbers like 60% is poor form.

You are welcome to keep it at 0%.  I have no particular interest in having my 
code backported into the gPXE tree.

> You clearly have committed lots
> of useful code, but there is a lot more to FOSS than that, and lots of
> people help us in ways that are not so easy to quantify, but are equally
> worthwhile.

Actually, it's not too hard to get a rough and ready quantification; some time 
ago I introduced the practice of including additional non-coding credits in 
the git log, such as Reported-by, Tested-by, Reviewed-by, etc.

This allows one to observe, for example, that although Stefan is the primary 
author of only one commit in the iPXE tree (during the time in question), he 
has contributed towards many more.  Similarly, although I rewrote from scratch 
Timothy Stack's patch to handle multiple TLS handshake records, he is credited 
with having originally fixed the problem.

You yourself are credited in the iPXE tree with having reviewed the jme 
driver, reported a bug in the jme driver, and signed off on two other commits, 
despite the fact that you have written no actual code.

(It's nowhere near a perfect system of measurement, particularly if the 
committer forgets to add the relevant credits, but it does give a very good 
first approximation, and aims to ensure that the equally-worthwhile non-
authorship contributions such as yours are not forgotten.)

> > On a more serious if annoyingly pedantic note, please note that
> > romprefix.S is using the EISA vendor ID assigned to Fen Systems, and so
> > you must take care not to change the functionality of romprefix.S in any
> > way that would break compatibility with iPXE.  (Or, alternatively, obtain
> > and use a different vendor ID so that there is no potential conflict.)
> 
> If you can be a bit more specific about what number(s) you are referring
> to we will explore what is required to replace them. I don't immediately
> see the number you refer to, but perhaps someone else will:

The EISA vendor ID in the handle used for PMM allocations (see 
PMM_HANDLE_BASE).  See the PMM spec for the relevant rules.

> I am also aware that there are a number of magic numbers in gPXE that
> were embedded to help detect cases where people are using gPXE without
> adhering to requirements the GPL, some of which are registered to your
> company, Fen Systems.

Yes, there are.  All of these are embedded within code over which I hold the 
copyright and so am the only person with legal standing to address any 
potential GPL violations.  If you encounter a suspected GPL violation, feel 
free to let me know and I can take appropriate action.

Michael