My 2 cents, more as a CouchDB user than as a nominal committer. a) I really like Tim's suggestion.
b) I've built two IoT prototypes where the applications were pushed to the edge devices via regular couchdb replication of couchapps, and I loved it. In fact, this feature was reason #2 (after replication) for picking CouchDB. 3. Don't promise CouchApps are easy. This is very smart. Marketing it as an advanced feature makes sense. However, there's more than just ... Instead, *CouchApps are **a way for non-novice developers to use CouchDB > as both a database and an app server.* > CouchApps are also a mechanism for getting your app updates to the edge devices. CouchDB on Rpi-class devices is a thing, and not all edge devices are handheld clients with a browser. c) If I understand correctly, there would be still only one product: CouchDB 2.0 which stores and serves your data, replicates, clusters, shards it *and* it also serves couchapps. The separation between CouchDB and CouchAppServer would only be in the documentation/marketing, not in the software shipped. Correct? J
