After three years of work by more than thirty volunteers across twenty companies, I’m proud to announce the Sandpiper Framework 1.0.0-FINAL release. For the technical details, head on over to github to read the full release notes, or just download the documentation in PDF format.
In 2019, Jason Riegel gave a presentation at the ACPN conference, highlighting the mounting pain of full-file distribution in the automotive aftermarket. I’d gotten to know Luke Smith, chairman of the ACPN at the time, while working on my own smaller presentation about starting a data department from scratch. I approached him with an idea that started bouncing around in my head after Jason’s speech: couldn’t we treat product collections more like sets of smaller stateless elements? If their unique IDs referenced the exact state of their content, then any change would result in a new ID, and comparing the sets would automatically provide a method to resolve them against each other.
Luke did something I didn’t expect, but now recognize as one of his unique character traits: he immediately started grabbing people and turning the idea into an effort. First he brought in Doug Winsby, then Jonathan Larsen, Tom Vierling, and Dave Hastings. Before I knew it we had a group of people all excited to make something new happen, and started having weekly calls.
Over the past three years, participants have come and gone as their time allowed, but every one has given a part of themselves and their experience, freely and passionately. The list is too long to place here (and I know I’d forget someone), but I want to say thank you to everyone who has joined us on a technical call, participated in one of our business surveys, reached out to a colleague, or just told us that you support what we’re doing. There were a few days that I would doubt our abilities and consider giving up, and invariably someone would email me excited about our ideas, and I would be ready to jump back in again.
So what’s in the future?
We’re taking the next month off of our technical calls to catch our breath and turn our attention to implementation. Originally we wanted to develop a reference client/server program alongside the standard, and for the first half we did. Critically, this allowed us to vet and solidify the foundation of the standard on top of which the rest is built. But after about two years we ran out of application development horsepower as job roles and membership changed, so we had to scale back to providing the schemas and data structures independently.
As a silver lining, this has given us the opportunity to work more closely with solution providers, who in the automotive market and its very traditional attitudes would have a tough time contributing to a functional, potentially competitive open source offering. We’re attempting to create a collaborative environment between proprietary implementations, where the standard is improved by everyone’s efforts at the same time as their own internal work is improved.
Thank you all once again for wanting to build something together!
One last note, if you want to stay in touch, feel free to sign up for our newsletter. We only send out major announcements — once a quarter at most — so it won’t fill up your mailbox.