FRESH

Hacker News

Show HN: VAM Seek – 2D video navigation grid, 15KB, zero server load

40 points by haasiy

by Scaevolus

1 subcomments

Client-side frame extraction is far too slow to be usable for large volumes of data.
You want to precompute the contact sheets and serve them to users. You can encode them with VP9, mux to IVF format, and use the WebCodec API to decode them in the browser (2000B-3000B per 240x135 frame, so ~3MB/hour for a thumbnail every 4 seconds). Alternatively, you can make the contact sheets with JPEG, but there are dimension restrictions, reflow is slightly fiddly, and it doesn't exploit intra-frame compression.
I made a simple Python/Flask utility for lossless cutting that uses this to present a giant contact sheet to quickly select portions of a video to extract.

by fc417fc802

1 subcomments

> All frame extraction happens client-side via canvas – no server processing, no pre-generated thumbnails.
Doesn't that mean the client has to grab a bunch of extra data when it first opens the page, at least if the user calls up the seek feature? Since you effectively have to grab various frames from all throughout the video to generate the initial batch. It seems like it would make more sense to have server side thumbnails here as long as they're reasonably sparse and low quality.
Although I admit that one line client side integration is quite compelling.

by Anthony76

2 subcomments

There’s a recurring pattern where tools fail not because they’re inefficient at scale, but because the first 30 seconds feel heavy or intimidating. Zero-config, zero-backend approaches often win simply by lowering that initial cognitive cost.
We’ve seen similar dynamics while working on long-term adoption and visibility for developer-facing Web3 tools at AixBoost.com — the solutions that feel instantly usable tend to compound trust faster than technically “better” setups with more friction.
Curious if you’ve thought about a hybrid mode later on, or if keeping the mental model simple is the core philosophy long-term.

by haasiy

0 subcomment

I’ve read all your feedback, and I appreciate the different perspectives.
To be honest, I struggled a lot with how to build this. I have deep respect for professional craftsmanship, yet I chose a path that involved a deep collaboration with AI.
I wrote down my internal conflict and the journey of how VAM-Seek came to be in this personal log. I’d be honored if you could read it and see what I was feeling during the process: https://haasiy.main.jp/note/blog/llm-coding-journey.html
It’s just a record of one developer trying to find a way forward.

by dotancohen

1 subcomments

This looks absolutely terrific if it is performative. How long does this library take to generate the thumbnails and the seek bar for e.g. a 60 minute video, on 8-year-old desktop hardware? Or on older mobile devices? For reference, my current desktop is from 2012.

by littlestymaar

1 subcomments

The idea is very compelling, it solves a real use-case. I will definitely take inspiration from that.
However, the execution is meh. The UX is terrible (on mobile at least) and the code and documentation are an overly verbose mess. The entire project ought to fit in the size of the AI generated readme. Using AI for exploration and prototyping is fine, but you can't ship that slop mate, you need to do the polishing yourself.