This is great! I want sigstore-based or at least sigstore-style "keyless" short lived signatures to become more and more popular, so that support gets better everywhere. In particular, if having secure OIDC handling in a CI/CD pipeline becomes something that keeps people on GHA or GitLab then I'd hope third-party hosted CI/CD services begin to take that gap seriously and work on meeting those requirements.
I'm bought in on keyless signing, and I want there to be many more providers, so I like to see more adoption which builds more momentum.
>I'd hope third-party hosted CI/CD services begin to take that gap seriously and work on meeting those requirements.
How is a third party service, especially a self-hosted one, expected to meet those requirements? They seem to be lengthy, onerous, and also require a detailed vetting by the pypi team that is unlikely to scale beyond a few large hosting providers.
(What's more, this does relatively little to actually signal trust to someone downloading them. All it does is provide a signature that one of the big providers has built the package from a given source (but nothing about the trustworthiness of that source). This is a problem much better solved by reproducable builds and distributed verification of those builds)
Sorry, by "hosted" I meant "not self-hosted" but I see how that's super unclear.
I don't mean just PyPI either. The list of acceptable OIDC providers with Sigstore/Cosign is short too: GitHub, GitLab, Google, Microsoft, and I _think_ BuildKite and Octopus? (But maybe only if the runner is in GHA?) It's really that last category that I'm talking about here (BuildKite and Octopus). No reason that I have to be tied to GitHub or GitLab, so I'd love to see this proliferation of trusted OIDC issuers that PyPI, NPM, Sigstore etc trust.
That's all I mean. If one of them, like BuildKite, believes this is important, because users expect trusted publishing, then this could be the kind of motivation which helps.
Still, agreed about reproducible builds, I just think that didn't really catch on beyond OS packages and Go. (Just my impression)
So, if an easy-mode signature system can be created here, that could be a win for everyone.
I think OIDC tokens from workflows is basically table stakes at this point, since the alternative is a static secret, which is basically guaranteed to leak eventually.
All that's left is for PyPI to accept these tokens in their set of trusted publishers.
Please see the other comments in this thread (as well as on PyPI's issue tracker) for some important context here. It's not enough to throw JWTs over the wall, PyPI also needs to know that it can trust any given service to operationalize the ugly and risk-prone parts of managing PKI.
(I haven't looked at Tangled at all. If Tangled has a secret management story for its PKI that's stronger than a private key on disk and can demonstrate that a non-trivial number of PyPI users would benefit from federation, then it's a strong candidate.)
Sorry I didn't intend to insinuate that pypi should necessarily accept any old tokens just because they're there. Just that increasingly OIDC is more and more an option, which is generally a good thing.
PKI definitely is still hard, but if it can be done well by a smallish number of trusted parties to the benefit of their users, this can have huge ecosystem benefits. Let's hope we live in that timeline.
IMO, one way to retain the power user crowd is to do the hard parts that they might have skipped themselves. For a couple examples:
- after selecting your observability apps, every _other_ app could automatically configure centralized logging, its own dashboard of metrics, etc
- the host OS could ask the user simply for the backup destination (S3 url creds, attached local disk, SMB share, etc) and every app is again automatically configured with backups
- the host could have a control panel to roll back apps between versions (maybe it's FS snapshots or something)
- everyone can use containers, but if the project went through the extra effort to make each app its own VM instead, that could be worth something
- if there was a companion Android/iOS app for management and (importantly) alerting?
---
That doesn't really solve the monetization issue you described, but I do think you can still get powerusers on board if you handle even harder problems than they'd want to take on themselves. Or...harder problems than they'd risk taking on themselves if their family uses what's being hosted, for example. Less tolerance for downtime
This is neat! I've been building something very similar for personal use. I may lift a little bit of the app manifest part.
I'm a little surprised that there's just one backup app, instead of each app coming with its own backup sidecar. To backup Immich, for instance, requires coordinating between its four different programs. I guess I'll read more about how that backup app works too ha
It's weirder still because you often don't actually have to register with a party, even to vote in the primaries like other commenters have mentioned; it depends on which state and which party. [0]
For example, I just voted in the Minnesota primaries. Ours are open primaries, so you don't need to register with any party and the ballot has every party on it. Though, it spoils your ballot to vote in multiple parties' primaries. So you gotta choose, but you do that privately.
Funny timing! I just finished migrating my NAS containers all to be rootless last weekend.
"Migrating" because secrets management was a bit of a pain. I ended up with pretty close to the same design as this article though. (I put an extra layer of complexity in mine, we'll see if I regret it)
Along with the other folks in this thread, I don't think we can stop at rootless in 2026. But I'm optimistic that dropping in libkrun as the runtime will be an easy next step, given its integration with podman (--runtime=krun).
If the author tackles that next, I'd be interested in reading about it
I'm bought in on keyless signing, and I want there to be many more providers, so I like to see more adoption which builds more momentum.
reply