Hi there, Gerard here. I work for Core as a firmware engineer, happy to answer questions as well.
I personally understand Rebble fears, for example when we forked and kept development under Core Github. However, I think we tried to be as transparent as possible and explained the reasons behind. While Liam (ex-Pebble) did an excellent job integrating NimBLE, it is also true that we also offered to do the work. However he had more availability by then to do so. At the same time, we fixed quite a few bugs after integration, or implemented many missing non-trivial features to make it functional. If you also check Github statistics, you will see that as of today ~93% of commits are from Core employees or paid contractors.
All development is happening in the open, and released under Apache-2.0 license. This is an exception in the industry, specially for core product components. It is also common for companies to fork when developing new products because you need to move fast (check our commit rate!). Think about Linux, can you use upstream Kernel on most new ARM SoCs? No. Core took a risk here because Rebble could have kept adding new features, adding overhead for us with upmerges. Reality is that Rebble repository has been dead since we forked. Nobody except Core, and Liam were contributing by then.
Another fear I've heard is about PebbleOS being sold to another company. Well, the company doing that would be pretty dumb as they could clone it for 0$. And thanks to Apache-2.0, they could even add new proprietary features! Not only that, but if Core winds up, the IP will stay open forever!
I think the best, fair long-term solution is to join a well established OSS organization. Rebble lacks many formalities that are common in many OSS projects: board elections, open and regular meetings, public accounts, voting rules, etc. This makes it a dysfunctional community to me. It is up to Rebble to fix these problems or join forces in a new OSS org. Core can't do much more than that. It is also not bad that the two parts have different views, e.g. Core may think a local voice-to-text model is better but Rebble may disagree because that could imply a revenue loss. That's unavoidable, in the end, people could choose at that point.
I know it's not your focus, but what's your take on the Core app frontend being closed source? I know libpebble3 is open and has the important bits, but it still feels bad to be unable to build an APK or grab that from F-Droid.
I had initially assumed it was because of some kind of dependency redistribution issue, but I think I read somewhere it was to stymie clones being developed and using the app. But that's part of an open ecosystem, no? That anyone can integrate into it?
Not firmware, but is there any chance Core would release the app as open source too? It's weird to have the library open, yet the app itself closed source, especially with how bare bones it is, it could be a nice gesture of good faith, show it's not about being "closed" for example
I personally understand Rebble fears, for example when we forked and kept development under Core Github. However, I think we tried to be as transparent as possible and explained the reasons behind. While Liam (ex-Pebble) did an excellent job integrating NimBLE, it is also true that we also offered to do the work. However he had more availability by then to do so. At the same time, we fixed quite a few bugs after integration, or implemented many missing non-trivial features to make it functional. If you also check Github statistics, you will see that as of today ~93% of commits are from Core employees or paid contractors.
All development is happening in the open, and released under Apache-2.0 license. This is an exception in the industry, specially for core product components. It is also common for companies to fork when developing new products because you need to move fast (check our commit rate!). Think about Linux, can you use upstream Kernel on most new ARM SoCs? No. Core took a risk here because Rebble could have kept adding new features, adding overhead for us with upmerges. Reality is that Rebble repository has been dead since we forked. Nobody except Core, and Liam were contributing by then.
Another fear I've heard is about PebbleOS being sold to another company. Well, the company doing that would be pretty dumb as they could clone it for 0$. And thanks to Apache-2.0, they could even add new proprietary features! Not only that, but if Core winds up, the IP will stay open forever!
I think the best, fair long-term solution is to join a well established OSS organization. Rebble lacks many formalities that are common in many OSS projects: board elections, open and regular meetings, public accounts, voting rules, etc. This makes it a dysfunctional community to me. It is up to Rebble to fix these problems or join forces in a new OSS org. Core can't do much more than that. It is also not bad that the two parts have different views, e.g. Core may think a local voice-to-text model is better but Rebble may disagree because that could imply a revenue loss. That's unavoidable, in the end, people could choose at that point.