Day 4/100
My favourite project is WeMake Inner Circle. It's a very innovative project to me because it bypass the standard limitations of the Google ecosystem.
It manages identity, storage, and data decay in real-time.
Firstly, the frontend; which was the most hardest to create but very intringing.
Most web apps hide content behind a login screen, but this project frontend goes further with a Deny-by-Default.
For Index.html UI,
I used window.onload to check for valid invite code in the URL or a user token in local storage.
If neither is found, JavaScript removes the login form from the Document Object Model (DOM), ensuring that even if someone inspects the element, the interactive parts of the Inner Circle disappears.
For the isolation of my code, I used a completely different HTML file via the doGet(e) function, making administrative logic invisible to the public.
To keep the chat real time despite Google Sheets' physical limitation and performance degradation, I used a Scalable Sharding Engine (SSE) to implements database pagination using a hidden system metadata sheet as the brainbox to store the active shard name, in the form of ChatDB_1.
So every time a message is sent, the backend checks the row count and, upon hitting 40,000 rows, it automatically creates a new sheet named ChatDB_[N 1], sets headers, and updates the metadata pointer, ensuring the fetchNewMessages function only scans a light sheet and keeps polling latency under 2 seconds.
Also, Google Apps Script lacks traditional file stream uploads, so i had to create a custom media pipeline which uses client-side encoding to convert media like HTML5 Audio or images into base64-encoded strings, it's then sent as text payloads for the backend to decode back into raw binary blobs using Utilities.base64Decode().
The blobs are then saved to a Google Drive folder, and the resulting Drive ID is returned to the chat and rendered via the
google.com/uc?id= bypass, displaying media directly in message bubbles. It takes approx. 5 to 8secs.
The ephemeral nature of the whole network is maintained by a self-destruct sequence running on a time-driven trigger that utilizes backward indexing to loop through database shards from the bottom up, ensuring index stability while deleting rows.
The multi-cloud incineration then proceed to identify messages older than 24 hours and uses the content ID to find and trash linked files in Google Drive, wiping the digital footprint from both the database and storage simultaneously.
That's all,
Also, I used Vanilla JS and Tailwind CSS to keep my system readable and portable.
#GoogleAppsScript #Serverless #SoftwareArchitecture #ZeroTrust #DatabaseSharding #BackendEngineering #CloudComputing #WebDev #SystemDesign
#GhostNetwork #CyberSecurity #PrivacyTech #Ephemeral #InnerCircle #Encryption #DataDecay #SecureCommunication #Bootstrapping #SoloDev #TechInnovation #BuildInPublic