Hi Jonas,
dropping bug report as my answer has nothing to do with it
On 21/09/25 10:41, Jonas Smedegaard wrote:
> Package: librust-tower-lsp-dev
> Version: 0.20.0-1
> Severity: grave
> X-Debbugs-Cc:debian-devel@lists.debian.org,debian-r...@lists.debian.org
>
> Package librust-tower-lsp-dev is impossible to install: depends on
> missing package librust-async-codec-lite-0.0+default-dev.
>
> The cause seems to be the common situation in the Rust team of initially
> uploading to NEW, despite knowledge that some declared dependencies are
> not yet available in that Debian suite.
>
> Please do not knowingly destabilize Debian unstable, but release known
> broken packages to experimental, re-releasing to unstable only when
> dependencies are all there.
>
> debian-devel@l.d.o is Cc'ed, anddebian-rust@l.d.o as well since
> apparently the Rust team treats their Maintainer field email address as
> something else than for human conversations.
>
> - Jonas
As I've already explained in the other bug you cc'ed to d-devel
yesterday, with one exception which does not apply to neither this, nor
the other bug you cc'ed (see below), there's no such thing as
*knowingly* uploading broken packages to unstable in Rust team. If you
see this happening *knowingly*, you can safely assume the choice was
made by experienced DDs (because we would absolutely not permit it from
newcomers) in their personal capacity. From our part, we should probably
have better oversight of these situations, as unfortunately in the end
the Maintainer field lists the Debian Rust Team, so I have a hard time
arguing this is not "our bug", despite it not actually being our bug in
the sense that it does not reflect team policy or customs.
If I may, I would please ask you to refrain from characterizing this as
"The cause seems to be the Rust team doing so and so", after I've
already told you we don't allow this. As team, we work our ass off to
avoid such breakage, and it's insulting to hear we do this on purpose.
But I do know you mean well, so I say this with no hard feelings. I
imagine you're just looking at the Maintainer field, see above.
Back to the exception, what we do allow is uploading multiple dependent
packages to unstable/NEW at the same time. Then if a reverse dependency
gets accepted before its dependency, this leaves the former
uninstallable in unstable for some time. This is already being discussed
in the other thread, so I will not comment further here. I would however
like to note that neither this bug nor the other one you cc'ed to
d-devel yesterday fall under this exception. In both cases, the missing
dependencies were just *not* uploaded to NEW. I have no additional
information about what happened there. I also have no information to
suggest these packages were uploaded broken on purpose (*knowingly*) in
unstable. On the contrary, I have some non-conclusive evidence they were
*not* broken on purpose, but by mistake.
