OpenSSF’s Alpha-Omega Project – Good, but not enough.
On February 1, the Linux Foundation’s OpenSSF announced the launch of its Alpha-Omega project, an initiative aimed at ...
On February 1, the Linux Foundation’s OpenSSF announced the launch of its Alpha-Omega project, an initiative aimed at securing what are considered the most “core ecosystem” Open Source projects as well as another 10,000 widely-deployed open source projects. Microsoft and Google have kicked in a collective $5 million to help fund this effort.
There are three important points to make about this initiative which need to be raised. One is technical and two are non-technical. Let’s start with the technical.
On the surface, the OpenSSF’s initiative has a really lofty and noble goal – to secure the open source software supply chain. However, its reliance on sigstore’s technology and particularly cosign and Rekor is less than optimal. The flaw in the architecture of this system is two-fold. First, cosign is based on certificates and keys – ancient technologies. What happens if you sign billions of artifacts with a certificate and then that certificate is now compromised? Will you go back and revoke the trust of all your assets? Furthermore, what if the actual trust root becomes compromised somehow? Serious questions to be asked and answered.
OpenSSF also plans to use Rekor, which is meant to be an immutable, tamper-resistant ledger and function as a transparency log of metadata about artifacts in a supply chain. Rekor utilizes Google’s Trillian for it’s underlying append-only data structure and requires MySQL or MariaDB and Redis underneath. Ultimately, this leaves room for vulnerability. There are too many moving pieces at play here, and the more pieces means more pieces you need to trust, or actually shouldn’t need to trust (you should be employing Zero-Trust). Once the underlying layers are not unequivocally immutable, there is room for the introduction of doubt and risk. To us, that’s far from optimal. The Rekor site also states: “IMPORTANT: This instance is currently operated on a best-effort basis. We will take the log down and reset it with zero notice. We will improve the stability and publish SLOs over time.” Yikes!
Now let’s look at how Codenotary solves these issues by utilizing our open source immutable database, immudb. It’s been very popular lately. Since immudb is itself an immutable database and doesn’t rely on any external piece of infrastructure, there is no room for doubt, no room for risk. What goes into immudb is stored in immudb, in a tamper-proof and verifiable way. What’s more, the state of the database can always be verified by the client as well, so you never need to blindly trust what immudb is telling you. It is one single integrated solution that provides all the requirements and functionality. This is a more compelling model, a more assured model, a more confident model, and it’s the reason we’ve built Codenotary Cloud on immudb, which can scale to billions of artifacts and notarizations/authentications.
Aside from these technical issues, there are two other big issues. While $5 million sounds like a lot of money (and it is), it seems like a paltry sum to carry out this mission. The depth of penetration of open source into the global software supply chain is daunting. When you start to take a look at your potential Alpha projects, say, the Linux Kernel, etc. the effort there alone can potentially utilize almost all of that. How do you get to all your Omega projects then? Which leads us to the second non-technical point.
Did you hear about the Fortune 500 corporation that emailed an open source maintainer to “Demand Answers” about his software package, which they’ve never paid for, and which they’ve now realized is being used in their software? We did. And we didn’t think it was funny. This story showcases what is perhaps the most obvious truth about how the software supply chain can be secured – by paying maintainers. Any project or initiative, such as the one launched by OpenSSF, will never be complete and never fully actualized unless money is being earmarked to pay those maintainers of Omega software. Otherwise, the outcome will just be to identify more holes in a maintainers code which will cause them to work more hours, not less, for the same non-existent compensation. Open source has always been about community and it’s time the community re-think the value proposition when it comes to paying maintainers.
So on the whole, yes, the OpenSSF’s announcement is a positive step. People are starting to take securing the open source software supply chain seriously. Now, we need the right approach in combination with the right technologies to make the dream a reality. It’s clear, we still have a long way to go.