Good question! So far, we haven't been overwhelmed with the app count. I think we can actually easily support far more. This is only because:
The main packaging cost is the first initial version. If you see the source code of any app package, we hardly have to make changes for a new release. It's mostly bumping the version. I am hoping once we have a CI/CD app on Cloudron, we can automate this part.
All our stable apps have tests. We don't test anything manually. The apps that don't have proper automated tests (like confluence), we have regretted publishing. But we keep them supported and test them manually since enough paying customers are using it. But that was a lesson for us to not add apps that don't have tests.
Many of the new apps that we are publishing now was packaged by our community - @msbt , @iamthefij , @Felix, @thetomester13 , @jimcavoli , @murgero (there's also 2-3 more but I am unable to find their id) etc. So, we are just building on top of their hard work (thanks a lot again guys!).
Finally, our goal is not to package all the apps possible. We just want the best 2-3 apps in each category. So, if there are 30 categories. This will only come to 90 apps or so.
All that said, if you see a drop in quality in support or general, please let us know and that will a hint for us to look into adding resources
