This is a post from Robin Sloan’s lab blog & notebook. You can visit the blog’s homepage, or learn more about me.

Attention router

January 28, 2023
A shard of a broken vessel, fabulously colorful, with streaks of blue, green, and yellow. You almost can't believe it's 2000 years old; shouldn't it be GRAY?
Fragment of a vessel, circa 1 CE, Roman

I wrote a new story that was just pub­lished by Brand New Box, a dig­ital product studio in Lawrence, Kansas. It con­cerns a giant mod­ular syn­the­sizer, and the web edi­tion fea­tures a vir­tual syn­the­sizer you can actu­ally play, so I feel embold­ened to share it with this crowd.

Here’s the story: In the Stacks (Maisie’s Tune)


I’m very taken with a little wave of apps that allow you to pub­lish and main­tain a feed as easily as you’d tweet.

Beluga is an iOS app that pub­lishes your feed to an S3-compatible storage provider. It also pro­vides a simple feed reader. All together, it really does look and feel like Twitter … except it’s all feeds, feeds, FEEDS!

Postcard takes a more web-centric approach, bundling together a simple home page, feed, AND email newsletter, with no app required. It is an offering of the appealingly-named Contraption Company.

Microfeed is a light­weight CMS “for podcasts, blogs, photos, videos, doc­u­ments, and curated URLs”. Heterogeneity, yes! You set it up by deploying your own copy to the free tier of Cloud­flare Workers.

These apps all have a won­derful light­ness to them; impres­sive work all around.

The ques­tion remains: what about discovery?


Val Town is a new plat­form for tiny cloud functions, and if it’s some­thing dif­ferent from the apps above, it still seems to “rhyme”, somehow.

All of these offer­ings want to sug­gest that it should be triv­ially easy to pin a feed, or a function, to the internet: as easy as sticking a magnet to the fridge.

There are some provoca­tive ideas lurking in Val Town. After cre­ating your tiny cloud function, you can make it public, and other people can call it from their tiny cloud functions, using a clever addressing scheme.

This makes me think of the ambient “composability” of the Ethereum Vir­tual Machine — only simpler, and faster, and cheaper. (It wouldn’t be such a ter­rible fate for Web3, in the end, to be “raided for parts”, all its gen­uinely humane ideas detached from the heavy chassis of the blockchain.)

I think Val Town’s console.email is great, too. Can we just add that everywhere?

An exer­cise for the reader: what might a net­work of feeds built entirely on Val Town look like? That is to say, they wouldn’t quite be “feeds” as we know them, static con­tent updated at intervals, but rather flexible, free-run­ning code. How might these feed-functions call one another, relate to one another? What new inter­ac­tions might be possible — richer than RSS, weirder than social media?

Those are murky ques­tions, pos­sibly bad ideas — so, obviously, I am having fun thinking about them.


From my notes:

Again, this sense of, “if they’re gonna make it, I’m gonna use it.”

Pin­ning bits of code to the universe. As easy as posters on utility poles.

This is the com­puting envi­ron­ment we are swim­ming around in — this is water. “Pure” plat­forms and do-over pro­to­cols are just not tenable.

Live in the world!


Max Krieger’s Voiceliner is a lovely, inven­tive app for cap­turing voice notes on the go, in a struc­tured format; its approach feels, to me, truly new. If you’re an invet­erate note-to-self-er, or aspire to be, you ought to check it out.


I always leap when a dis­patch from Jackie Luo appears. She is a sen­si­tive chron­i­cler of her milieu, diaristic and polit­ical in the deep sense — I wanted to write “in the French sense”, I don’t know why — just end­lessly readable. Her recent New Year newsletter artic­u­lates a widely-shared feeling of stuckness:

there’s a run­ning joke (is joke the word?) on twitter that we’re all still stuck in 2020, or that we’re about to begin year eight of 2016. in my own life, at least, that has felt true. 2016 is the last year i can recall feeling deeply opti­mistic about what the new year would bring, for me and for the world at large. since then, the fragile hopes i bore for each new year have been flat­tened again and again into the form­less same­ness of a world where time means nothing and yet somehow every­thing man­ages to keep get­ting worse. the future began to feel less like an unbounded space of poten­tiality and more like a pre­cious resource that kept diminishing, untouched. eight years is a long time. where did it all go? how did i get here? it’s hard, living in such per­sis­tently unprece­dented times, to know what is the nat­ural process of aging and what’s the spe­cific pecu­liarity of aging in this time.

Although this feeling isn’t only, or even primarily, about the internet, I think it con­nects to the widely-shared sense of “ … is this it?” that I have dis­cussed in this newsletter. And I think it’s impor­tant to under­stand the search for new avenues in that richer context: of pol­i­tics and economics, work and meaning, hope and dread … all of it.

Jackie’s “stuckness” makes me think of Zyg­munt Bauman, the philoso­pher who saw it all with such clarity, way back in 2000:

We feel rather than know (and many of us refuse to acknowledge) that power (that is, the ability to do things) has been sep­a­rated from pol­i­tics (that is, the ability to decide which things need to be done and given priority) [ … ]

Read Jackie; read Zyg­munt. This feeling of stuckness, the inability to get traction, under­pins every other link in this newsletter.


What were the early explorers and engi­neers of the web dis­cussing and debating, 29 years ago? Review the proceedings of the First Inter­na­tional Con­fer­ence on the World-Wide Web and find out!

I love this kind of doc­u­men­ta­tion “in media res”. As you read, a sense of legit­i­mate dra­matic irony takes hold: these people didn’t know what was coming.

I also enjoyed browsing the code for version 0.5 of NCSA’s web server, circa 1993. It’s very simple!


Web Browser Engineering is a web book that uses the browser brilliantly. Its sub­ject is: how to make one.

The bal­ance between light­ness and inter­ac­tivity here is wildly impres­sive. The book’s sturdy clarity sup­ports mag­ical flourishes: inline code exam­ples that really run, along­side visu­al­iza­tions revealing what they’re doing. And the side­notes are beautiful.

The web of 1993 this ain’t!

A col­lab­o­ra­tion between Pavel Panchekha and Chris Harrelson, this is about as close to a per­fect web book as I’ve seen. (The other con­tender is Rodrigo Copetti’s Architecture of Consoles.) If you spend any of your time thinking about browsers and what they do, you’ll get some­thing out of this material — even if you only make it through the introduction.


I like this: widget.json, “a file format designed to push con­tent from the web to your home screen.” Its cre­ators write:

We had access to an internet that was playful, weird, unexpected, and inspiring. We love the vibrancy and tex­ture that comes from people having access to simple tools to express them­selves and create the things that they want. Wid­gets are an approach­able and per­va­sive canvas to have fun with. We hope people will create things we can’t imagine.


Stephen Wolfram’s inves­ti­ga­tion into the life and thinking of Ada Lovelace makes for a pow­erful read. I found the piece via Brandon Rhodes; his recommendation, just a single paragraph, is per­fectly con­cise and stirring:

[ … ] A deep dive into the printed and hand­written evi­dence of the col­lab­o­ra­tion between Bab­bage and Lovelace, copi­ously illus­trated with both. Con­clu­sively demon­strates that, of the two, Ada was the real programmer: the first member of our species to expe­ri­ence the joy of hacking a gen­eral pur­pose com­puting engine to do things that even its inventor didn’t realize it could do. Bab­bage comes across as a some­what imprac­tical startup founder and Ada as the hacker who really under­stands the technology.

Another shard of a broken vessel, just as colorful, this one with a fat stripe of rusty red or brown, and a glimpse some pattern, now long-lost.
Fragment of a vessel, circa 1 CE, Roman

Faster, now:


That’s it for January. If you your­self have pub­lished anything, in any format, that roughly fits the theme of this newsletter, please don’t hes­i­tate to send me a link. I’d love to include it in a future edi­tion and route a bit of atten­tion your way.

Regarding my query, in December’s newsletter, about localStorage and cross-browser synchronization, this post was illu­mi­nating. Sev­eral cor­re­spon­dents sent links to PouchDB, which is new to me, and looks great.

To the blog home page