Restarting this blog, seven years later
The old albertarno.net was a robot. This one is not.
The last thing published here went up on 23 February 2019. Then nothing, for seven years.
That gap needs less explaining than what came before it, so let me start with the awkward part. Between June 2012 and February 2019 this domain published 194 posts, and I did not write a single one of them.
What the old blog actually was
Around 2012 I set up an IFTTT recipe: anything I saved in Feedly got posted here automatically. The tags are still on the posts and they give the whole thing away — Feedly, IFTTT, Saved for Later, Recently Read. Some entries are even tagged with the name of the journalist who wrote the original, because the automation copied the byline across too.
The result was 194 excerpts from Engadget, TechCrunch, Bitelia, ALT1040, Boing Boing and Noticias madri+d, each one ending in “via” and a link. The median post is 134 words. Twelve of them are completely empty — the feed fired, the body never arrived. The longest are full reproductions of other people’s articles.
It was a bookmarking habit with a public URL attached. I think in 2012 that felt like curation. Read back in 2026 it is just someone else’s writing under my name, and republishing it would mean putting text that belongs to Engadget and TechCrunch and the rest back online as mine. So I have not.
I have not deleted the record either. The complete 2012–2019 archive lists all 194 posts with their dates, each linking to its capture in the Internet Archive. Everything that was here is still findable — it is simply attributed to whoever actually wrote it.
Why it stopped
It stopped because the work got heavier. onmedic had been building clinical software since 2000, and somewhere around 2018 the interesting problems stopped being the ones you can link to and started being the ones you have to sit with for months. Hospital data models. Consent. What you are allowed to infer from a measurement and what you are not. None of that fits in a feed item.
I also stopped believing that reposting links was worth anyone’s attention, without quite deciding what to do instead. So the recipe kept running until it didn’t, and I did not notice for a while.
What this is now
Over the last few years I have been building measurement software for the body: a set of sports and health analytics apps that read what your watch already recorded and compute the models a coach would actually use — critical power and W′ balance, critical swim speed, gait asymmetry, heart-rate variability. They run the maths on the phone rather than on my servers, which is a design constraint with consequences I find worth writing about.
That is what this site is for now. Notes from building it: what the models actually say, where consumer sensors stop being trustworthy, what happens when you take a technical decision for privacy reasons and then have to live with it. Things I have done and can show, rather than things I have read.
Fewer posts than 194, then. Written by hand, which is the only change that matters.