Meeting Minutes - Office Hours for CTI - 2026-08-28
Alexandre Oliva
oliva@gnu.org
Mon Sep 28 14:28:37 GMT 2026
On Sep 27, 2026, Siddhesh Poyarekar <siddhesh@gotplt.org> wrote:
> On 2026-09-27 19:53, Alexandre Oliva wrote:
>> In the current scenario, we have the GNU Toolchain community talking
>> directly to Sourceware volunteer service providers.
>> In the proposed scenario, we have GNU Toolchain community members
>> begging CTI (TAC, Board) members, who are not going to be appointed by
>> the community indefinitely (TAC) or at all (Board), and they will decide
>> which requests to pass on to the commercial service providers who will
>> be bound by a contractual agreement as to what service they're going to
>> provide.
> That's a gross misrepresentation of what the TAC is; it is *by
> definition* made up of members of the community:
> https://cti.coretoolchain.dev/gov/charter.html
> https://cti.coretoolchain.dev/tac/index.html
There's some certificate issue problem with that site, BTW. I managed
to get through, and I hope I got the info you meant me to get.
It's not reassuring at all, alas. Yeah, the charter says "contributor
to the technical projects". That's not the same as community member.
Users are part of the community as well. They're arguably the most
important part of the community, for projects that stand for users, such
as GNU.
It also clearly establishes that TAC members may suggest TAC membership
changes, but it's the CTI board that decides whether or not to make
them. So the board owns the TAC.
And the board is made up of enterprises who pay the most for membership,
plus one TAC representative.
One of the big risks I perceive in this arrangement is that TAC members
may eventually start perceiving themselves as *the* community, or as
*representative* voices of the community, without having been
*appointed* by the *community*. (This seems to be already at play)
Since they're appointed by the CTI board, they're at best representative
voices of the hosting sponsors who happen to have contributed to the
technical project. (assuming the board would ever refrain from
appointing someone who's never contributed to the project, or who isn't
an active contributor, which is not a given)
This is a sure recipe for them to drift away from the community, while
maintaining an illusion of alignment with the community as long as they
perceive themselves as representatives of the community.
Now, you seem to have raised your argument in an attempt to dispute my
assessment that the power structure is not meant to serve our community.
In doing so, you've confirmed my assessment. Because, you see, I noted
3 barriers set up between community requests and their implementation by
the proposed hosting provider: TAC, board, and professional contract.
You've objected to one of the barriers, as if that removed all 3. In
doing so, you yourself implicitly confirmed 2 of them, even if the TAC
were indeed perfectly aligned with the community.
> the TAC can be comprised of people
> outside the GNU toolchain community.
You attribute the quote above to me, and label it a lie. I acknowledge
it's a bit of an exaggeration, but it's unfortunately too true. The
only requirement for TAC membership is board appointment, and the only
requirement (to the board?) is that it's a contributor to the technical
project. But the charter can be amended by a 2/3 vote of the board, so
the board alone can make that requirement
go. But they don't really have to. They can interpret it quite
literally, and get someone appointed who has only contributed one patch,
or one bug report, or one mailing list post, or one IRC message 20 years
ago.
Now, maybe the current TAC members wouldn't nominate such a remote
contributor for approval. The board would have to be patient, slowly
changing the composition of the TAC to suit the interests of CTI buyers,
approving or rejecting requests for addition or removal of members, and
then eventually they will have captured the TAC as well.
And if the TAC remains perceiving itself as the community, they'll have
captured the project.
Congratulations! Well done! *slow clap*
--
Alexandre Oliva, happy hacker https://blog.lx.oliva.nom.br/
Free Software Activist FSFLA co-founder GNU Toolchain Engineer
Learn the truth about Richard Stallman at https://stallmansupport.org/
More information about the Libc-alpha
mailing list