There are a lot of upset developers, and rightly so. The Versa 3 and Sense trackers have JUST been released. This means that 99% of their consumer base is still wearing legacy Ionic and Versa devices. It makes no sense to drop 4.2 as a development platform or make 4.2 overly difficult for developers to support.
I understand that the OS needs to move forward with 5.0. Happily, most of the developer code is high-level JavaScript, SVG and CSS. I moved one of my clocks successfully and ONLY the visual files (icons, images, styles, etc.) in the resources folder needed to be changed. The CORE clock face code and companion app were unchanged. So, even if the two OS platforms are truly incompatible under-the-hood, the high level code that the majority of the developers create is perfectly compatible.
So then, all that really would need to change is the tooling. Personally, I think it would be as simple as versioning the "resources" folder into "resources/4" and "resources/5" and allowing the tooling to dynamically select between these locations during the build process.
Plus, developers would need a way to test whether a feature was available on a particular device from within the JavaScript code before using it. For new features like temperature, the 4.x branch would simply return "false" for everything. The benefit would be a common API between 4.x, 5.x and any future releases.
Hopefully, Fitbit will consider supporting the current 5.x API branch, as well as one MAJOR release backward -- which would be 4.x. Whenever OS 6.0 is released, support for 4.x should end. Hopefully by then, most of the Fitbit consumers will have updated their physical hardware to Versa 3 and Sense devices, and Fitbit will be enjoying the success of selling all those new trackers.
Further, if developers think that maintaining separate 4.x and 5.x versions of each of their clocks and apps is difficult, just imagine how fragmented the Fitbit Gallery will become if developers are forced to upload multiple versions of the same app. I believe that is where the REAL HIDDEN danger lies, and Fitbit will definitely need to address it.
Developers should keep in mind that Fitbit is still a relatively young platform. Mistakes are going to be made and resolved. However, the platform that Fitbit created is EXTREMELY cool. You've got first-time programmers being as productive as folks like me, somebody who has been coding professionally for 30+ years on a variety of platforms. That's no small trick to SUCCESSFULLY pull-off, and I don't think that as a community, we should lose sight of that.
Fitbit's primary focus has been on bringing the new OS 5.0 products to market -- and RIGHTLY so. Let's face it, Fitbit doesn't make any money on products that they've already sold, and supporting them is nothing but overhead. However, new apps and clocks need to be written for 4.2 as well as 5.0 because 4.2 is where most of their end consumers actually reside.
Let's give them time and see what they do. 🙂
