Find n nearest locations in flutter: local database (Sqlite) in combination with Cloud Firestore vs Firebase Realtime Database

Question

I am working on the following topic right now:

• about 8000 stores/objects(uniqueID, name, adress, latitude, longitude, geohash, amount of productX, amount of productY.)
  • in the future there will be new stores and also some of the stores will be deleted
  • I am planning to add a version attribute to each data: for example I start with version1. If I receive an update or a new store then it will have version2
  • it is not needed to receive realtime changes. This data will be changed only a couple of times each month
• Flutter app with a feature: find n stores at this location

---

• because the data is not changed frequently I was planning to put an updated local sqlite database to each build

• I will use Realtime Database or Firestore to receive updated data within my app to avoid to publish a new version only because database was changed

• Firestore: I will add only a new attribute to my document and I will request all data that is newer than my local database version and I will also have a collection that has all deleted items

• Realtime database: here I would have something like this

       -updates
                version1
                         newData: List of unique IDs
                         deletedData: List of unique IDs
  

---

I read a lot about geo queries on firestore and I also found a working flutter plugin to do this filtering on serverside, but because my data is not changed that frequently I am very confused about the best approach.

Questions:

1. Is my approach a good idea or should I forget about the local sqlite DB and use only Firebase/Firestore?
2. If I will use the local database and will only observe updates from the backend, which of the services should I use in terms of pricing etc.? (Advantage Firestore: If I use Firestore I can easily switch to it by using the flutter plugin and forget about my local database)

Answer 1

If you have only a couple of thousand objects, and the data hardly every changes, I'd typically consider querying the data locally on the device. You can likely very quickly iterate over the 8,000 items to check their lat/lon against the user location.

You could then use something like Firebase Storage or even Firebase Hosting to distribute updated data sets to the users. Of course you could use Firestore or Realtime Database for that too, but if you're not using their querying or realtime capabilities, there are cheaper options to distribute a static file.