After checking some resources I believe that the following looks like the ideal migration plan for existing apps and no-coders.
I’ll first take care of the backend and keep using Bubble as a front-end.
Xano seems to be a good option for a no-coder like me even because it has a great plugin (https://www.youtube.com/watch?v=1xs48PNhmB0&t=1s). What more it is GDPR compliant (Bubble is not). Alternatives I didn’t check are Supabase (also has a plugin) and Backendless (no plugin). After checking this video, I believe Xano is better as it comes with better workflow management.
- I will start creating twin pages for my app where i) users are created on Xano, ii) all data is written both on Xano and Bubble but read just from Xano. It will be a long phase as my app is very big and I will need time to learn Xano.
- After testing everything I will: i) bring old users from Bubble to Xano (I still have to figure out how to do that), ii) bring old data into Xano, iii) deactivate the old Bubble pages and keep the twins.
- Afterwards, little by little, I will deactivate WFs that write data on Bubble.
- I will try to have as little logic as possible left into Bubble. I don’t yet know if I’ll be able to rely on Xano only or if I will need an additional tool like n8n.
- The intermediate goal is to get to a super light front-end in Bubble and a back-end in Xano (and maybe some WFs in n8n).
Only after I get to point #5, I will decide which front end I should use instead of Bubble. For the moment WeWeb looks like a possible option as it’s easier than Wappler. Other options are FlutterFlow and Draftbit but I guess they would require to re-submit a build to the stores and I’d rather avoid that pain and keep using a wrapped web-page.
I don’t want to go again with a solution that has backend+frontend because i) I’m not going to put again all my eggs in one basket, ii) I believe that the migration would be more complex iii) separating the back-end from the front-end and managing everything with APIs looks like a good move for future enhancements.
Some full stack solutions mentioned here are Drapcode, Noodl, Appgyver, Appmaster, Bildr, Dittofi. I didn’t check them, maybe one could be a good option if used as front-end only.
During the migration period I will consume al lot of WUs but it doesn’t matter. The final goal is to complete the migration before 18 months.
Please let me know if the above plan makes sense for you. Thx!
UPDATE 1. I started studying Xano and I found out 2 material drawbacks:
- You can’t schedule an API at a future date. You instead have to create a cron (what they call a “background task”) that will check the db every X minutes and take an action if a condition is true (see this post in their community).
- What’s worse is that you can set very few background tasks in their affordable plans! (0 in the free plan, 5 in the 59$ plan , 15$/mo in the 356$/mo plan).
This looks like a dealbreaker. I will start studying Supabase.
UPDATE 2. Supabase is not a no-code tool when it comes to functions. It’s way too complex for me. I started implementing Xano. As for he background task I will use make.com
UPDATE 3. I found out that Xano is not a real time database. In my app I need data to be updated without the user doing a reload. I’m now checking www.back4app.com. It seems to be quite easy for a non developer like me but, as far as I understood it has limited WFs. The logic will need to stay elsewhere. As for the moment I think the final configuration will be back4app backend + make.com for the logic + Noodl for the front end. Noodl has an amazing editor and, to me, it looks much easier than Flutter Flow. Noodl also has his own backend if you don’t mind putting all the eggs in one basket. I know that many Bubblers are going with WeWeb for the front end. I haven’t checked it properly but it seems weak in case you want your front end to have both a web app and a mobile app.