Re: [Fsfe-uk] RFC: Free software project grants

[Alex Hudson]

(My 10p - I wasn't going to get involved in this debate, but what the
hey ;)

On Fri, 2003-10-17 at 15:05, PFJ wrote:
> While I am used to filling in grant applications for research projects,
> I can't see many having the time (or ability) to fill in a grant form
> for the AFFS or FSFE, jump through the hoops required and generally lose
> valuable time from software development.

There has to be some process of dividing this money out. We cannot
approach people; people have to approach us. 

However, I do agree that it needs to be braindead simple. I don't see
any requirement for anything more than:

     1. who i am, what the project is and what my relation to the
        project is
     2. what I want to do
     3. why this is the greatest thing since Perl 5, and why I require
        funding for it.

If you can't be bothered to fill out a (web) form with those three
things - say, 400 words or something - then you obviously don't want the
grant :)

When people make applications, I think they ought to be publically
viewable. I've not really seen a convincing argument that they should be
private. 

I think there ought to be a publically viewable process, but not
necessarily a public decision. What I mean by that is that I'm not sure
that there ought to be some kind of poll or vote or anything like that -
I don't think this should turn into a popularity contest on that basis,
since it biases towards projects which have buzzwords / street cred /
great script writers. 

There ought first to be a set of guidelines to what we accept proposals
for, and proposals should be marked against them. That should remove the
'chaff' proposals, and the choice in the end should be towards what
projects would make the most difference (really, in terms of
bang-per-buck). 

I don't think we should be throwing money at people for just developing
software; it ought to be goal directed. For example, with Scribus -
'improving Scribus' is a project which I don't think should be
acceptable. 'Writing Quark support into Scribus', on the other hand,
gets my vote. The difference here is that there is an end result: we can
point to something and say 'we did that'. If you're not able to do that
at the end of the day, then it's something I don't think should
necessarily be funded.

Other examples I could think of: 'Writing a Free driver for piece of
hardware X'. Wouldn't necessarily even need to be a software project:
'Designing hardware to interface Psion wavefinder DAB radio with PC to
use GNU Radio on it' (this is made-up and probably impossible ;) could
be cool. 'Writing a Free manual for software X'. 

The two main criteria I see to judge on, then, is 'How worthwhile a
project is this?' (how far does it advance free software?), and 'How
crucial is the funding to the project?' (if you guys are going to
continue to develop Scribus anyway, why should we fund you?). 

AFFS' responsibility doesn't end with the giving out of money; it only
begins. I would envisage we have a page on our site which basically
tracks progress: people should be able to see what is going on, and how
far to total success it is. The projects need to have deliverables, and
AFFS needs to check that they are indeed delivered. 

I think the only thing that we are going to end up arguing about is
going to be that most sensitive of things: how projects get picked.
That's essentially the crux of the issue. And, at the end of the day, no
process is going to be even near perfect. Lots of good projects will get
turned down, no matter how we do it. Arguing that process X is better
that process Y is likely a tenuous argument; but I think it is
definitely key that the whole process be entirely transparent:

      * up-front guidelines need to be available
      * guidelines should be what projects are judged on
      * projects should chosen on the basis of how highly they fulfil
        the guidelines

The most important decision, therefore, and probably the one which needs
most input, is what guidelines are used as the basis for judging a
project. The projects are going to be judged somehow, and whoever is
judging them should be judge them all on the same basis. The key
question is: what does it mean to be a good application?

Cheers,

Alex.