🧐 #Analysis № 1: Mini Apps Resources

As of now, there are few resources on Mini Apps. So I created this list of resources for future reference, and to help others starting out.

Link to post with resources here.

🤔 #Opinion № 1: Finding True Mini Apps

I am currently aware of only two places to browse for Mini Apps (that are not channels): Apps Center and FindMini.app.

I am not satisfied with these solutions, as I am interested in browsing for "true" Telegram Mini Apps. This means: applications that are useful because of their relevance to, or existence within Telegram. Not apps that just happen to open in Telegram, e.g., bots without a Mini App, (scam) games, crypto hype apps, and other sorts of "non-apps".

Apps Center is created by TON, and therefore is mainly populated with crypto apps and questionable games. FindMini.app comes with many bots (not apps), low-effort listings, and categories that do not fit the aforementioned criteria.

I would like to see a nicer true Mini Apps catalog. Let me know if you are aware of existing good solutions.

🤓 #Tip № 2: Cloudflare Tunnel > ngrok

ngrok is the most commonly suggested tool to establish a local network tunnel for Mini App development.

I recently came across Cloudflare's Tunnel, and I found it to be much better. It has a much more generous free tier, with fewer limitations. And it seamlessly integrates with your domain names. I overall find it cooler, and would recommend it over ngrok.

Honorable mention: Localtunnel, as a simpler alternative (Credits: @aabounegm).

Pavel @durov, the creator of Telegram, just posted that they will release a Mini App store this month! 🔥

And I just wrote about this issue yesterday! Is Durov secretly following this channel? 👀😁

🤓 #Tip № 3: TON to Stars

Fragment introduced a new feature to buy Telegram Stars with TON.

As of now, this method is ~25% cheaper, compared to other currencies (e.g., USD).

👨‍💻 #DevLog № 3: Referral Systems

Mini Apps utilize Telegram's social features to reach new users. Referral systems are used to incentivize users with rewards and track who referred who.

These systems are quite simple and elegant at their core. You can see how they function in the diagram.

Inspired by a question from @aystek ⚡️ Source code with docs.

👨‍💻 #DevLog № 4: Anonymous Messages 👀

There are many text bots for receiving messages anonymously, but a few Mini Apps.

A demo of an integration between Mini Apps and traditional web apps 🤝

Inspired by a Anonion ⚡️ #FreeDurov

Edit credits: @Sh1co for suggesting to add more colors 🥳

💡 #Idea № 2: Mini App Subscriptions via Channel Payment Links

Telegram introduced Star Subscriptions to join channels ⭐️ This can be used to power subscriptions for Mini Apps (and bots), and not just channels. Making this one of the simplest subscription implementations around 💪

How does it work?
1. 🔗 Generate a subscription channel invite (payment) link
2. ▶️ Add a button in the Mini App for the user to pay
3. 🤑 User pays and joins the channel
4. ❓ The Mini App's bot checks if the user joined the channel
5. ✔️ Payment confirmation and unlocking premium features

Dedicated to @levelsio for his continuous inspiration to launch products quickly via Stripe payment links ⚡️ Shoutout to @adalbright for the idea spark!

Here is to celebrate 5️⃣0️⃣ subscribers! 🥳

Other things I'm celebrating 🎉
- Completing my first Mini Apps client project 👨‍💻
- Telegram still growing 📈 and releasing awesome features ⚡️
- Durov still alive and free (for now) ✈️

Thank you for all your support! ☺️

Hint: your friends can join before tomorrow evening 😉

👨‍💻 #DevLog № 5: Waitlist Feature ⌛

A minimalist approach demo for a waitlist feature. The user can join and leave a waitlist.

Technical details 👨‍💻: with the current template setup, the feature took less than a 100 lines of code, in a single file. With one database record per user, that triggers a cloud function that send the message. ⚡️

💡 #Idea № 3: The First Fully AI-Automated Apps

I predict that the first fully AI-automated, widely-accessible, non-trivial software application would be something like a Telegram bot. But this statement requires some definitions.

Fully AI-automated: the person creating the app does not interact with any code nor requires any technical knowledge. For example, creating and deploying the Telegram bot solely by chatting with ChatGPT.

Widely-accessible: can be installed and used by an average, non-technical user.

Non-trivial: full CRUD features with a shared database. i.e., data is inserted and viewed by other users.

Here is a walkthrough of the thought process behind this conclusion:
- Apps typically have a frontend and a backend.
- Automating the development of frontend and backend is more complex than automating just the backend.
- What are the "simplest" apps that do not require "frontend development"?
- Arguably, terminal (command line) applications. But, terminal applications do not qualify as widely-accessible.
- What is a more accessible (easily sharable) version of a terminal app?
- Telegram bots.

Conclusion: Telegram bots (or something like them) are likely to be the first form of, widely-accessible, non-trivial, software applications, that would be fully created by AI-prompting.

Implications: any Telegram bot can be (deterministically) transformed into a regular (web) app. i.e., the second form of software apps to be fully AI-automated is likely to be Mini Apps (due to their less complex nature).

I would love to hear from you, if you've got feedback or related ideas. Shoutout to my friends for the discussions helping me formulate these concepts.

🤓 #Tip № 4: Free Storage 💯

For typical apps, a storage bucket is required to store users' files. The most popular solutions are AWS S3 and Google Cloud Storage.

But this is not the case for Telegram Mini Apps. Why? Because Telegram Bots have always stored files by sending them as a message to the user, without needing to store them elsewhere (in most cases). They relied on generous Telegram, as a file storage medium. ⚡️

Mini Apps can talk to bots to send files, and to retrieve them later when needed. All that's needed is to store the ID of the file.

To store files shared by multiple users, one can:
- add the Mini App's bot to a group, and send the file there
- forward the file message from one chat to another (to skip re-uploading)
- fetch the file (in memory) and send anew to another user or group

As a consequence: 100% free cloud storage 🤑

Feel free to donate your future storage cost savings in the form of stars on this post 😉

👨‍💻 #DevLog № 6: No-code Mini Apps Builder 🖥

Exploring what a no-code Mini Apps builder would look like 🔍 Just a mock-up for now.

Stealing Borrowing design from AppSheet 👨🏻‍🎨 with a bit of help from v0 ⚡️

100+ subscribers! 🥳🎉

I would like to use this opportunity to thank you all ☺️ and to ask you a question 👀

❓ Which mini apps do you use at least once a week (besides the Wallet app and games)?
And if you do not use mini apps, what do you think is missing? 🤔

Any thoughts you could share would be very helpful 😉 You can reach me @elbatanony

🤔 #Opinion № 2: The Future of Sales for Mini Apps 🛍

Traditionally, you create a service, setup payments (banks and Stripe), develop an in-house affiliate program or integrate a ready service (both costly options), convince and onboard affiliates, then implement the attribution/distribution of commissions. 😮‍💨

With the new Telegram Affiliate Programs, it only takes a few clicks. Transactions are already handled by Telegram (Stars). And users already use Telegram and trust the automated attribution and distribution of commissions. ⚡️

This paves a way for a reimagined, frictionless, and efficient sales process. When a business no longer hires a sales team or influencers. But when enterprising individuals independently look out for businesses and audiences and profit from the value generated from the connections they create. 🫰

Would have been cool if the affiliates service itself had an affiliate link that I could use in this post 😁

👨‍💻 #DevLog № 7: Soar Quest, Archival and Revival
A short story about an old project and its revival.

Soar Quest (v1) was an app development framework I worked on during 2022-2023. The idea at its core was simple: software should take fewer lines of code (to be built). And indeed, I was able to build functional "mini" apps in less than 100 lines of code. Initially, I was building this framework to make cross-platform applications with Flutter. Then in September 2023, Telegram held a contest for Mini Apps.

I was already very intrigued by Telegram as a platform, and in a matter of a week or so, I refitted the framework to build Telegram Mini Apps. This contest was especially fun because my friends were willing to try out my framework to build apps for the contest (after a considerable amount of persuasion 😁). We created some very cool things (with very few lines of code) in a short amount of time.
During this contest, I was making live updates to the framework for my friends to pull and use in their projects. Not the "stablest" of framework APIs, but it was so much fun.

With over a year after the start and more than 1900 mini Git commits, I drifted away from this project. There were a few reasons. One of which was the tech stack that I chose. I was building with Flutter, which not ideal for web apps like Telegram Mini Apps. And I also wanted to change the backend system. Another reason was because during this journey, I discovered more what I would like and would not like to build. For example, I figured that I don't like using mobile apps as much as desktop-screen apps. I halted the framework development.

A few months ago, I was working on a project. While developing it, I remembered the motivation behind Soar Quest and I felt the code redundancies again. I had a plan for the upcoming year (but that's for a different story time), and a framework to build mini apps fit in that vision. So I restarted this project with fresh tech and with Telegram at the core.

Today, I archived the old project. But, I am not done with this quest (or it is not done with me). #SoarQuest