I've created Rowt as an open source solution for deep linking and analytics infrastructure- built on NestJS. With Firebase Dynamic Links being phased out and Branch.io's pricing structure being "call sales", I wanted to offer a community-driven alternative.

You can host yourself or there's a super-affordable managed service, including a free tier with 50 links/7 day analytics to try out how it works.

Core features:

- Postgres and SQLite support

- TypeORM migrations (no SQL needed)

- Single/multi-tenant ready

- Built-in auth and JWT handling

- Cleanup chronjobs for expired data

- Analytics dashboards via Console SDK

- Multi-project management

Setup takes just 5 minutes from zero to running locally. Configure via config file and env vars.

TypeScript SDK works with Expo and web apps. React Native improvements coming soon (accepting help with improving the native intent listeners from a real native dev)

Links:

Docs: https://docs.rowt.app

GitHub: https://github.com/Rowt-Deeplinks

NPM packages:

   https://npmjs.com/create-rowt-server

   https://npmjs.com/rowt-sdk

   https://npmjs.com/rowt-console-sdk

Managed service: https://console.rowt.app

Any feedback or questions are welcome, thank you!

Thank you! I do see the use for these, though a few things are a little off target for what the app hopes to solve. Component preview and mock props is actually something I have planned for the future with a subtype of snippet for templates, so when you intend to have custom inputs there is a way to test with different data. First though I'd have to build a different system for live previews and a dynamic environment that can adapt to which component it might be hosting.

Storing in the cloud to grab them with something like a cli tool is super cool. I have a feature on the mvp+ list to download a list as a library or to fork to github as a lib, just converting it into a folder. But I'm trying not to let scope creep distract me from just making it as easy as possible to grab a quick tool without downloading a whole library, not trying to compete with npm or any other package/version manager.

At present the only AI features I can think could benefit the app are search tools and framework detection, but these can also be improved traditionally without the need for linear algebra or paying for use of any ai api I think.

Basic version control could be cool too! like storing previous versions if a tool was useful to you in different forms

I do overall though feel the goal of this is a little different from the Bit platform, but some lessons could be learned about what people want based on the product's popularity :)

Hey! thank you! I actually have a great use for my site itself, you can find the language detection utility on this list: https://snippp.io/list/18

but tldr I'm using hljs to detect and return a string with the language title, and comparing that key against my array of categories.

I will be putting some more focus on the mobile experience after seeing how many people will be using/checking this out on mobile! It's tough as a primarily desktop app I assume since most will be working, but I won't let that stop me from making the mobile experience as good as it can be. Admittedly I did have mobile on the sidelines :)))

But thank you so much!

Thank you so much! This is all incredibly helpful, I felt like I needed to respond when it had my full attention haha

1: The countdown is definitely something I can adjust so we have some way of letting you know without thinking anything is loading.

2:Search is super bogus right now, what I'd actually love to do is run the search on the database itself but wanted to prioritize speed. Even so, I had not included the code itself as searchable since I thought it may bloat results a bit too much vs only searching for the high level descriptions of what you need to do. I am playing around today with how I can weight and sort so the search is more useful to everyone but may need more data to make it great :)

3: Absolutely agreed, went over the site and pulled the 100s up to 2-300, ty :)

4: hahaha I am struggling to figure out the riddle of the first sentence, but I do agree this was an embellishment. In an app where the goal is efficiency and clarity there is no need for a bell or whistle like that. I do want to give character somehow, without getting in the way. Some way to make it feel snappy without completely stripping it down.

5: Love clickable tags, the tags being in the show more is just a matter of visual clutter. My thinking: Most users will likely not need tags > description, so they are easy to find if you need them but not distracting. Open to other points of view on this

6: oooooo I will do more testing on browserstack

7: I don't have an image atm, but I can put that together :) though aside from consistency and scalability, I don't know how much of a benefit it has as an end user, but I admit I am semi-ignorant to the k8s world

8: This is foreign to me and seems like a decent project to take on, but I can't knock the benefits.

9: I'm working on putting this together for the faq and to display on signup, but more or less never sharing with any third party, I only collect google profile information for sake of creating a user profile on the app, with email to future-proof the accounts system if I decide to expand to something other than google accounts. I'm using a trusted cloud storage provider for data. The app currently only retains some details in local and session storage to smooth out the app's operations, but don't save any cookies on users, that's all on the client side. I do currently use vercel/google analytics so it may be reporting things like location, device, browser, and page visits.

10: Lmao I'll take that compliment, It was for me and I will not be accepting revisions <3

Thank you for all the information! I can use a lot of this to improve experience, and hopefully make this useful to someone :)