Parts 1 - 2 covered the why and how of building the initial parts of v0.2.0. In this installement, I'll cover how I pulled off the seemingly impossible feat of putting this thing on the web and what's next for us as a company.
Before I dive into that I wanna say a few words about AI. AI is a force multiplier but you have to overcome the initial inertia yourself. Expecting it to do all the work is a lazy mindset that will lead to average results. Having very good fundamentals and deploying it tactfully ,on the other hand, leads to endless possibilities.
So how did I end up putting this thing on the web with desktop-level performance? Well the exact details are a trade secret (copycats hardest hit). What I can say is that I relied heavily on the Emscripten tool chain. Simliar to how FFmpeg is the gold standard for video internals, Emscripten in the gold standard for WASM compilation. It's not perfect but for 95% of devs, it'll get you where you need to go.
So at this point my stack was SDL, DearImgui, FFmpeg and Emscripten and I had a working desktop demo. Trying to compile the entire code to WASM became an enormous challenge and for a long time, nothing I did actually worked. I damn near used every feature in that tool chain and funny enough built a lot of half correct solutions. Recall that at the end of Part 1, I talked about how what I thought were small gaps in my knowledge were still actually large chasms. I didn't actually realize this until I started dissecting the half-correct solutions and figuring out why they were in fact half-correct solutions. This is also when I really cranked my AI use (s/o to xAI) into overdrive and eventually was able to build it the correct way. If you want more details than that, get a job on our dev team in a few years and I'll be happy to go into more detail.
Now that I had a working web version, I realized something else. I had figured out how to put AAA level video games inside a browser. Still not sure what I wanna do with that knowledge but it felt very full circle that I had used a lot of video game tuts to learn about GUIs and ended making (what seems to be) a pivotal discovery in the space. Serendipity and all that.
At this point it's April 2026 and I'm besides myself with excitement. What started out as a simple weekend project had turned into a tool that I could build a business and community around. One thing I forgot to mention in the last installment was that after building the desktop version, I had opened up a waitlist and gotten a lot of interest which was very encouraging. I knew people would want this but getting external confirmation via people trusting me with their emails was really cool. To everyone that's joined the waitlist, thank you, I won't let you down.
As off the time of writing this, it's June 2026 and I'm now comfortable enough with all aspects of my knowledge base to build the full thing. Moreover, it has evolved from simply being a video editor to becoming a full video production suite. For too long, us editors have had to cobble multiple tools together for projects and I realized that building "One Tool to Rule Them All" was the way to go.
Thanks for reading!
Sign up for our
waitlist so that you find out about new developments in the project before anyone else.