After I decided to quit my job at the end of 2020 to become an indy maker. I made many things just for sake of the innovation, ideas, and love. Let’s see what I did even It’s seems failed terribly but It’s something.
Getting started with AI again
by picking up the old project of mine like stock price prediction using a recurrent neural network. I updated the code to the new version of the Tensorflow framework. I was trying to utilize GPU for training but It seems like it is slower than On CPUs. I tinker the model a bit but it is not helpful. I’m not looking for the result because I know stock prediction is the best to use the RNN model. I just try to train it faster using GPU but it’s not successful. At least I got a new way to easily install Tensorflow GPU in Ubuntu, you can read it here.
HNOOM — A stupid short video app
TikTok become popular and it’s reached a younger audience. It would cost heavily to run such a video platform. I want to design an app architecture that can support a short video platform without spending any penny so I design and develop the prototype app called HNOOM. The truth about the name is it stands for Hope it Not Out Of Memory because I was using the react native video player framework that must change the video player view quickly by swiping up and down. It did not handle the memory usage very well and it cause the app out of memory. HNOOM is using the IPFS to host the video by using the HLS technique and using the user’s phone to encode the video to small files uploaded to IPFS, and using Cloudflare IPFS gateway to act as a free CDN and using free firebase app as the video listing. I’m quite proud of this app architecture. It cost nothing to start building a video platform. I just created an innovative app architecture that is not like it before. It's probably really dumb so no one ever thinks about it. Anyway, I learned the app architecture design, coding a react native with the camera and video view. However, HNOOM is not the smoothest video scrolling experience by using react-native. It needs to be an actual native app for future work. Short video platforms become mainstream IG, YouTube, even LINE have it, called VOOM. Vish it not out of memory like mine. You can read about HNOOM here.
Zwipe — Swipe Tinder on your bike
It was lockdown again. I was doing two things to pass the time, swipe Tinder and indoor cycling. My bored mind was thinking why we don’t combine both into a single activity(no pun intended). I made Zwipe, a Chrome extension for indoor cycling to swipe Tinder. Zwipe name after Zwift an indoor cycling game and Swipe from Tinder, you get it. It’s using and a client app installed on a PC to access the ANT+ USB speed sensor and the Chrome extension to fetch the speed data to determine whether you swipe left or right on the Tinder web app. The usage is simple to ride the bike at your normal speed to pass a profile and ride 5% faster to like and 10% faster to super like. Faster average speed you ride, faster-swiping Tinder. The idea is really good for keeping me engaging in the exercise. I wish it could be turned into a business but building a new dating platform is hard. The success is based on how many people are on the platform, building it for a niche user group is a big no in my opinion. I’m still single the result can tell. You can read about Zwipe here.
Vision — A camera pendant.
The lockdown end and I’m looking for an outdoor activity to do. The one that’s not cycling is photography, especially street photography. One day I found beautiful scenery the phone is not fast enough to capture the moment. I wish I can have a small instant camera ready to take a picture with me all the time. So I design Vision a small pendant camera. It’s heavily inspired by RayBan Story from Facebook But instead of sunglasses it will be a small aluminum stick hanging around the chest as a pendant with a camera ready to take a photo by pressing a button on the side. I was designed it to be fashionable and elegant like an Apple product. I was trying to design in on CAD but CAD design is not my talent so I hired an industrial design from Fiverr. This is my first time hiring someone to do my ideas. I was turning out fine but not the best since I was not figured out many things. I am not a hardware developer neither. In my calculation, Vision will suffer from the small size because the battery wasn’t big enough to standby camera all day, and charging a wearable device every day is not what I’m looking for, what am I, an Apple Watch?. In terms of photography results, it’s not great nether, because it’s cannot see what in the camera frame the photos it took feel like a spy trying to stalk people for a photo. I have simulated the result using the phone on the shirt pocket to see what it will look like maybe it will change a bit when it becomes a pendant but not by much. I have no idea how to develop hardware that’s no for this project but the passion for photography will live on. I decided to carry a DSLR every time I go outside. If I got something to shoot it is fast enough and my DSLR battery lasts forever when it is lightly used. If I found someone interesting, I take the photo, ask for permission, be a real photographer. That’s how I solve my stated problem.
Random Bullshit — A Lightroom plugin for random photo adjustment
I was looking for a new way to improve my photography. I turn to what I do best, coding. I develop a lightroom plugin for random photo adjustment, just to get the fresh idea, style out of my photograph. It just works..hahaha
Read about it here.
WONDERSHOT.CO — A photography community and NFT marketplace.
I’m starting to be serious about photography and I’m producing some interesting works constantly. I was looking to sell my work. The Getty image was the first option but it’s my choice. I like 500px and like to show my work but summit for sale is a drag, many photos were rejected just because of lacking commercial ability. Of course, I was meant to sell it as an art, not a stock photo. I turned to the NFT marketplace. At the time not many people sell photos as NFT. A marketplace like Opensea is saturated by other kinds of art. The biggest turndown for NFT is it need to mint it first and the transaction fee cost on Ethereum is bananas. You need to spend ~$100 to mint an NFT whether it will sell or not, using the side chain is free but not convenient for buyers. I have some experience with Web3, IPFS, Solidity smart contracts. I should build a photography NFT marketplace for myself. I design the NFT marketplace UX from the ground up. Not relying on the user experience of web3 like others NFT marketplace. I name it wondershot.co. It is in the super early stage but you can try the website. It’s just a test on Rinkeby for now. I’m adding features to making more like a community, not just a soulless marketplace.
…
That’s it, my 2021 year in review as an indy maker. I was trying to search for myself to find out who I am. It seems all over the place but I’m trying. I wish I could turn the ideas into the business. Please let me know if you were interested in my ideas, what I could do better.
