Updates on object0 and VocaSync
The last few weeks have felt like the “real” part of building—less about announcing a thing into the world, and more about quietly earning the right to rely on it. Both object0 and VocaSync started as solutions to problems I was already living with every day, and progress lately has been about reducing the remaining friction: smoothing the edges, strengthening the foundations, and making each project feel less like a promising prototype and more like something you can integrate into your life without thinking twice.
object0 has gone through the kind of shift that looks cosmetic from the outside, but changes everything underneath. The app has migrated from Electrobun to Tauri, and the important part is that nothing was lost in translation—the original functionality is still there, only now it sits on a leaner, more stable base that feels like it’s built for long-term iteration. That migration wasn’t about chasing novelty, it was about getting closer to the shape of a tool you can ship confidently: fast startup, tighter binaries, fewer moving parts, and a platform story that doesn’t feel fragile.
The bigger leap, though, is that object0 now has true background, local-to-remote bidirectional syncing in place, with three-way diff resolution when reality gets messy. That’s the point where “storage manager” stops being a nice interface and starts becoming a system you can trust to keep state coherent across time, devices, and human error. I wanted syncing that behaves like an adult: not just “copy A to B,” but “notice what changed, reconcile it, and make it legible when conflicts happen.” Alongside that, the job queue has been improved in a way you can feel—less waiting, fewer awkward stalls, and a clearer sense that the app always knows what it’s doing next.
And for the first time, distribution feels like it matches the intent. There are standalone binaries and installers available across Windows, Linux, and macOS (both Apple Silicon and Intel), plus an AUR package for Arch users who want to install via paru/yay without ceremony. This matters because tools like object0 either become daily companions or they die behind “I’ll set it up later,” and the distance between those two outcomes is often nothing more than whether installing it feels simple and native. The UI has also been refined in the same spirit: more responsive interactivity, more animation where it helps comprehension, and fewer moments where the app makes you do extra thinking just to confirm you clicked the right thing.
On the VocaSync side, the progress has been about filling the biggest missing piece in the pipeline: what happens when you don’t have a transcript at all. VocaSync now supports transcription of audio files using WhisperX ASR, which means you can generate a strong initial transcript, edit it into shape, and then feed it straight into alignment when you want timestamps that actually feel publishable. That workflow—draft transcript → human correction → precise alignment—closes the loop in a way that makes the platform dramatically more useful for real-world content, where “perfect input” is usually a fantasy.
That focus on flow has also pushed VocaSync toward a more deliberate idea of “workflows” as first-class product design. The goal is to make the platform feel less like a set of separate tools and more like a pipeline you can move through without cognitive overhead. You can now one-click an alignment job for any synthesis jobs, which sounds small until you realise that it removes the mental bookkeeping that always sneaks into voice tooling: keeping track of artefacts, remembering the next step, checking you didn’t miss a format, and re-uploading what you already generated five minutes ago. More of these workflow improvements are in motion, because the long-term vision here isn’t just power—it’s power that feels obvious to use.
One of the most concrete expressions of that vision is the VocaSync Astro plugin, now published as an npm package for anyone who wants the Valeon-style audio + word highlighting experience on their own blog. The plugin handles the parsing, synthesis, and alignment automatically, and it also takes care of generating publishable keys for public access so your content can be served cleanly without you building a bespoke storage and signing setup. With VocaSync, hosting of the generated artefacts is included, which means users get the entire pipeline—generation, alignment, storage, and playback—without having to assemble it from ten fragile scripts. The included player brings the “living text” experience into place: word highlighting with a toggle to turn it on or off, and a click-to-seek feature that lets readers jump to any word in the audio (with an option to disable it if you want a more traditional listening flow).
What I’m watching, across both projects, is the same pattern: once the foundation exists, the real work becomes making the tool disappear. Not by becoming invisible, but by becoming dependable—so predictable that you stop thinking about it and just build, publish, sync, and ship. There’s still a lot planned for VocaSync, including new integrations and plugins like a React library and a WordPress plugin, and object0 still has plenty of surface area to expand into. But this update feels like a threshold moment: less “look what I made,” and more “this is becoming the thing it promised to be.”
Whilst not related, I also took everything I’ve learnt over the past few weeks and revamped the UX for GigFin. In my humble opinion, it’s starting to feel more like a complete product, and I’ve plumbed in a few more widgets with richer stats.
