[FFmpeg-user] The Guild

David Niklas deference at null.net
Mon Feb 5 02:25:00 EET 2024

Mark was good enough to privately email me and we were discussing this.

He asked to take it to ffmpeg-user (below), so here you go. The only
thing you missed (IMHO) was a bit of misunderstanding between us, which
isn't included, as quotes or otherwise, herein.

On Sun, 4 Feb 2024 15:28:36 -0500
Mark Filipak <markfilipak.imdb at gmail.com> wrote:
> On 04/02/2024 11.34, David Niklas wrote:
> > Mark, currently, the system that's set up works so that those who have
> > give money to the devs who then do work that benefits everyone. So the
> > rich help the poor.
> Sounds good, but it's a false analogy, David. The real relationship
> isn't rich v poor, it's contractors v users. The contractors are
> companies -- some are very large -- that pay the FFmpeg developers via
> consulting contracts. They provide the support. Users don't provide
> support. If anything, users are dead weight. Users are interested in
> getting their video project working after which they disappear. Between
> those two there's contributors: Users who have technical knowledge and
> the motivation to make FFmpeg better. I believe that includes you and
> me.

And here I was looking forwards to the day when I could support FLOSS
devs after getting a job... sadness.
How do you know the money only comes from companies?

And calling regular users a "dead weight" is harsh. We each contribute to
society, from which ffmpeg devs benefit, in one way or another. And all
ffmpeg devs are, by your definition, "dead weight" before they contribute
to ffmpeg -- which is probably the majority of their lives.

> > Your proposal included an exclusive cloud based book. So now we have
> > those who have (bourgeois) and those who have not (proletariat).
> > Surely you see what happens next, right?
> I don't really know how to react to bourgeoisie v proletariat. Methinks
> it's another false analogy. Users will not participate beyond posting
> to ffmpeg-users. The book is for the contributors. It provides a means
> for contributors to create real documentation. It would also provide a
> way for the more serious users to 'plug in'.

So the monetary contributors will create their own documentation? Or are
you now expanding your idea to encourage more contributions, in the form
of documentation and/or code?

If you want more users so as to get more contributions, paywalls will not
accomplish that.

Again though, one of FLOSS's greatest strengths is it's accessibility.
Having bug free and "feature" rich code that "Just works" (TM) is not
something that FLOSS SW typically does. We should play to that strength,
not try and create a YT-er fanclub style system (It sounds like that's
what you're proposing).

> Such a book is research, not code, so it can't violate the FOSS license.

Just to quell your concerns, as you mentioned this on the ML and here,
I do *not* think you're trying to violate, directly, or indirectly, any
license, FOSS or otherwise.

> > The change you're proposing is a political one. It gives power to
> > those who have money to spare, and, depending on the organizing
> > principles, could result the devs that depend on the money for their
> > income being harmed by the financial officer of the guild.
> How? The developers would own the guild and would control the finances.

Barring the possibility that each contributor donates directly into the
financial account of the chosen developer recipient, there's going to be
someone in between the contributors and the devs transferring the funds to
each person.
If there is only one person in the role, or if the role works by voting,
there's still room for abuse and so considerations should be taken to
ensure that this doesn't result poorly for any parties involved.

There are more examples than just Mr. Stallmen and the unnamed Mozilla
CEO. There's the XEmacs debacle, and reiserfs, which didn't involve any
money at all. I've read through both accounts throughly. This is a matter
of accountability and leaving politics behind instead of using it against
perfectly good code, respectively.

> > And given that normal devs dislike politics in general, or so I've
> > read many times over from various devs, the fact that 2 different
> > people in 2 different foundations suffered from political strife
> > within the foundations should suggest to you that a degree of caution
> > is in order in proceeding with your goal of better supporting the
> > ffmpeg project.
> > > And that was the purpose of my original ML email.
> What do you think now, David? There used to be shareware. That's how I
> got most of my software. But FOSS killed shareware. That was not in
> anyone's best interest. What can be done to replace shareware?

Shareware was always a dead end. Once the person who wrote it decided
against supporting it, or was unable to, it would deteriorate due to
differing ABIs, APIs, bit count 32/64, security vulnerabilities,
tacked-on ads, or spyware, or malware, etc..

FLOSS didn't kill shareware at all. Anyone can write shareware and there's
still sites dedicated to distributing shareware alive today.

> This should all be on ffmpeg-user. What we are having is discussion.
> Discussion is constructive. Disagreement is not.

At your request, I'm sending this to ffmpeg-user.

> Do you agree that there needs to be a way for contributors to gain
> power?

Monetary contributors shouldn't gain power. Power shouldn't be a goal at
all in FLOSS SW, baring some extreme circumstance -- like the only guy
with write access to the repository cannot be reached and there's another
guy who wants to take on the project.

> Do you agree that there needs to be better documentation?


> Do you agree that a way should be found that adds shareware aspects to
> --Mark.

What particular aspects do you value about shareware that are not present

But if you really want to do something like this, Mark, how about a bit
of a different system than you had in mind? How about a "pot luck" system?

Anyone who wanted to could contribute as much or as little as they wanted
into a "pot". The pot would be emptied every month evenly into the
accounts of any dev whose contribution (code/docs) to the project was
accepted, provided that they asked for some of the pot when they made
their contribution. Any excess pennies leftover would be left for next

It would be pot luck how much the devs get, and pot luck what got worked
on. That way, devs would be free to fix that really really tough bug, or
could contribute a single line of code/documentation. The devs would,
therefore, feel free to pursue whatever thing tickled their fancy.

If someone wanted a particular bug fixed or some other particular matter
attended to, they could use the preexisting "hire a dev" system.

In order to avoid abuse, the contributors emails would be made available
to the recipients (unless they request otherwise). That way there would be
some accountability (between devs and contributors, the info could easily
get out), where a dev who was mistreated would be able to suggest that
contributions cease for a while until the matter is sorted out. The
results of each pot and who asked for some of it would be publicly
announced on a dedicated ML monthly.

You could even have two pots. One for documentation contributions and one
for code contributions.


More information about the ffmpeg-user mailing list