[FFmpeg-devel] Sovereign Tech Fund

Stefano Sabatini stefasab at gmail.com
Thu Feb 1 01:15:03 EET 2024


On date Wednesday 2024-01-31 18:10:57 +0200, Rémi Denis-Courmont wrote:
> 	Hi,
[...]
> Sarcasm aside, I take that to mean that SPI has been involved with those 
> discussions for months in a private and closed process. Michael asserted that 
> an open inclusive process is better than the usual closed approach whence the 
> funding goes through a company.
>
> It looks to me that those SPI discussions were just as opaque and closed, and 
> all the talk of openess is just pretense. It does not help that Michael, and 
> now you too, misrepresent any challenge to SPI proposed *process* as an 
> attempt to reject the idea of STF sponsorship, under the convenient pretext 
> that there is not enough time.
> 
> This is further aggravated by the context that Michael brought forward the 
> idea of funding developers through SPI 3 months ago (in actual Earth units). 
> From your statement, I have to infer that Thilo, Michael and SPI already knew 
> of the STF plan and concealed that key piece of contextual information back 
> then.
>

José already provided and excellent summary from his side. On my side
I can say I was involved in the discussion, and that this was mostly
about the feasibility and the groundwork of approaching STF and later
SPI.

So in my opinion there was no need to involve the community at that
early stage, especially given that until the past week there was still
no evidence that STF was providing a grant (and BTW, still this needs
to be substantiated with a SOW and then it will have to be reviewed
and approved on the STF side).

Also SPI was involved at a later stage, after the investigation about
using the donors fund for active development (which also involves the
handling of SOW from SPI to the individual contributor).

The main result of that investigation was discussed in the open and
can be found here:
https://ffmpeg.org/pipermail/ffmpeg-devel/2023-October/315702.html

If I undestand it correctly, it was never committed because there was
a disagreement about where to put it (ffmpeg or ffmpeg-web) and about
the general intent, then at some point the discussion died off before
a conclusion was really reached.

Note that the focus in that case was to make good use of the donations
fund (keeping it in the account is not a good use for it).

> In hindsight, it feels hypocritical to me that they were arguing for the SPI 
> path, and against the corporate path, on the basis of openess already then, to 
> be honest.
>
> I can only agree with Anton that this looks like an attempt to strongarm the 
> community. This is ostensibly being to ignore all the objections that were 
> already brought in October and are being brought again now, with the 
> complicity of SPI. I can't say that this looks well on SPI, but that's just my 
> personal opinion.

SPI was involved at a later stage to act as fiscal sponsor for
STF. Just to reiterate, SPI involvement was requested, and was not
actively seeked by SPI itself.

I cannot read any attempt to strongarm the community, nor I see why
this should challenge the corporate path (which has a different focus
and has its own merits).

> With all that said, I don't think anybody will attempt to prevent this from 
> happening (if they even can?). But that will take place without the consent of 
> the GA, without any legitimacy on the claims of openess and inclusiveness, and 
> obviously without any form of preclearance from the technical appropriateness 
> of the resulting code contributions.

It's unfortunate there is a tight deadline - one option would be to
try to delay the deadline and ask General Assembly for a vote before
the application is sent - we might probably want both things to avoid
the feeling that this is done against the "community" and create a
tense environment, but any of this might probably result in voiding
the opportunity.

Also, it should be assumed that this proposal was done in good faith,
in view of the sustainability discussions done in the past months.


More information about the ffmpeg-devel mailing list