Ask HN: What are you working on? (August 2026)

(I'm kidding when I say this, and it's also concretely true.)

Been working on realtime DSP and ad-hoc A/V integration and sync.

Audio DACs are a solved problem. Pristine D/A conversion is a ubiquitous low-cost commodity. It's fun to buy a bunch of cheap "chi-fi" DACs – quality is excellent – and ad-hoc merge them into one big audio interface and do per-channel DSP. Toss in some TPA3255 amplifiers. Bodge together a crossover and room correction monstrosity in Reaper or Logic. Run system audio through that. It's easy to get super arcane. My home stereo is frightening. The interesting part is thinking about how to get a human interface on top of it. How to get it to assemble and stay assembled and present human-comprehensible knobs – or not.

Also had an extended family health crisis. About 9 years non-stop of extremely unlikely convergences of life-threatening conditions. (My close family has dissolved due to trauma and stress. (It's okay.)) And I haven't been working a programming or computer-science job since 2020… but I have very serious and significant work experience in things that are… how to even put it. It's not immediately clear how to add the past few years to my resumé?

Did an invisible PhD in mitochondrial neuroimmunobiology at the School of Hard Knocks? Stint of off-piste biomedical engineering? Amateur placental biotechnologist? Independent researcher?, or just… faceplanted into some PDFs on Google Scholar and accidentally read some out-of-my-lane stuff and it turns out it's actionable. Results from molecular biology are actionable.

It is said that there's a 15 year gap before something reaches clinical practice. Between having in one hand some concrete scientific results. On how to treat something in the body. And on the other hand… to clinical practice. A 15 year gap.

What's that 15 year gap made out of?, and can we cross it earlier?

We can.

Trying to figure out how to get this into words, and what to do with it.

And then when I've got the realtime-DSP-ed-to-the-hilt home-computer workstation going, I'm going to work on Idris 2 because we need dependent types.

https://checksitepulse.com

One thing I've focused on is producing reports that are actually actionable instead of overwhelming. I'm also putting a lot of effort into reducing false positives—for example with spell checking and URL analysis.

It's still under active development, but it's reached the point where I'm starting to get feedback from other developers and site owners.

I'd love to hear what kinds of website issues you wish existing audit tools did a better job finding.

paul@checksitepulse.com

I realized that all clues could be represented by Boolean formulas, allowing a solver to check whether my deductions and hypothetical scenarios were consistent without revealing the solution.

Of course the same solver can solve for individual variables, but that takes too much of the fun out of the game, so I left that feature out.

The game is meant to look like a retro RPG, but it's more of a puzzle about finding the optimal path. You must grab as many power ups as you can to level up your stats, while avoiding monsters to conserve your health.

It's inspired by a niche genre of puzzle/RPG called Magic Towers. If you already know Desktop Dungeon, Soulestination or Tactical Nexus, it's kinda the same idea.

We've got a Steam page up and the demo has been available for a few weeks. Try it out, it's pretty fun: https://store.steampowered.com/app/4142360/Paradox_Tower/