10 things about end-user-programming

What is the future of malleable software?

Sol LeWitt, “Loopy Doopy, Blue/Red,” 2000

Hello! I’ve taken so much joy in the smell of freshly mowed grass this week. While enjoying the enchanting scent of said grass, I’ve also been wrapped up in the idea that we’re on the brink of our next wave in software. The feeling is based on a talk that Shishir from Coda did ages ago (As my friend K. says, “Shishir's distortion field is very funny because it's like zero anywhere. Unless you are a PM in Silicon Valley, and then it is over 9000.”)

Shishir, in a podcast he recorded over 4 years ago:

At the time [the early 2000s], what you were actually buying from these big enterprise software companies wasn’t a packaged piece of software. If you go look at how people deployed SAP or Oracle or so on, every single deployment was completely different. When SaaS came along, it meant that [you, the user are] adopting one model that works across different companies. Part of the value proposition was that it wasn’t significantly customized just to you. In fact, I think we’re tilting back now with the whole market moving back towards people customizing software

I think this is exactly right, and it’s the AI wave that will make this possible. We will have software that is the best of both worlds — 80% the same as whatever your peers may have, 20% customized to your needs. In the past, we’ve failed at this because it’s actually just…a really hard problem. Interfaces for this customization were too 1) code heavy 2) clunky and 3) poorly abstracted, in that they’re not really carefully revealing complexity — they’re either not revealing it at all, or throwing the user into the hard stuff at the get go! I think natural language oriented interfaces will REALLY help make this approachable and easy for users.

1. Before I dive into the rest of my links, I’d be remiss to not mention a big thing that happened this week: I launched a podcast with some of my favorite ladies! The story of Lego is so fascinating. Please give it a listen and a rating - it makes a huge, huge difference to get downloads and ratings in your first week especially. Lego is the original “programmable” toy — its well abstracted primitives are a perfect example of carefully revealed complexity, down to the simplicity of Duplo smoothly on-ramping into the slightly more complex Lego!

2. Gneiss, the streaming spreadsheet: Now that we’ve got that self-promotion out of the way….I find this paper and the demo tool to be very interesting. I think many of our collaboration tools are built with the idea that “data is created in the SaaS tool” but it feels increasingly less likely that the majority of data needed for a project will be generated in your tool. I like this idea of “streaming data into a spreadsheet” where you don’t have to write Javascript to make it happen. Curious to hear what you think!

3. Geoff Litt on the UX patterns for spreadsheet backends: One of the innovations in Gneiss (the tool described above) is that it requires you to have a “pull based” model of data population instead of a push based model (click a button and do something.) Geoff Litt wrote a bit about the tradeoffs of that UX pattern here. I like his synthesis at the very bottom, and his concise description of the flaws of each of the products. I think my meta point while considering these flaws is: do they matter to the end user? For example, Litt describes the flaws with Airtable’s enrichment API as: “API calls can't be triggered on scheduled intervals or in response to things that happen in the spreadsheet.c API calls aren't integrated into the formula language, so only narrowly prescribed interaction with specific integrated APIs is possible.” The former seems…like a feature, not a bug! I don’t want unpredictable actions happening in my spreadsheet, where API calls are being triggered without my explicit say-so. I’m not building the spreadsheet as an app, yet!

4. Andy Matuschak’s notes on Designing and Programming Malleable Software: Imagine a magical web, where there are no “boundaries between apps and create an end-user accessible “physics of interfaces” that dictate how different interfaces and documents can be assembled.” If you’ve ever had multiple tabs and applications open to accomplish a task, you know how ideal that world might be. I love reading Andy’s notes because he brings so much of himself into them — they’re not a summary! He also links all of his related thoughts in his notes, so that you can traverse the graph. I quickly rabbit-ed off to read this post on Figma as a surprisingly nice canvas for building docs, which makes me think that the nicest new interface for a docs tool is closer to a scrapbooking environment than the current Notion/Coda/Airtable highly structured interface.

5. File over apps: The founder of Obsidian talks about how the future should be files, not apps. It’s a very short read. An excerpt: “These days I write using an app I help make called Obsidian, but it’s a delusion to think it will last forever. The app will eventually become obsolete. It’s the plain text files I create that are designed to last. Who knows if anyone will want to read them besides me, but future me is enough of an audience to make it worthwhile.” To be honest…I disagree? I think that I already produce so much data — my digital contrails are smeared across the internet — that I’d actually like most of my data to be deleted, not for it to be portable. I don’t mind that my data is deleted when the app fails.

6. The enshittification of TikTok: These very jaded screeds on the internet are truly hilarious, but remind me of eating fast-food — you feel entertained but mildly ill after finishing it. The best part of this article is the “enshittification cycle” description:

   Here is how platforms die: first, they are good to their users; then they abuse their users to make things better for their business customers; finally, they abuse those business customers to claw back all the value for themselves. Then, they die.

   
   I call this enshittification, and it is a seemingly inevitable consequence arising from the combination of the ease of changing how a platform allocates value, combined with the nature of a "two sided market," where a platform sits between buyers and sellers, holding each hostage to the other, raking off an ever-larger share of the value that passes between them.

7. Executives on how the AOL & Time Warner Merger went so wrong: It’s so crazy to think about a deal like this — largest merger of all time! — talked about with such significance. It’s so absolutely irrelevant now. And the executives involved thought they were bringing about the future!

8. This thread on the shutdown of Merci Victoria Grace’s (early Slack executive) company is really interesting. The tweet below is true, and reminds me that the “tuning” work is where the value lies. I’ve always been interested in companies like hex.tech that give data scientists the opportunity to better meld the narrative and the data together.

   

9. A build your own CRM company — this is kind of the example of the type of tools that I’m talking about! I don’t know if this actually works, but it’s a fascinating idea.

10. Socks: The best thing I did for myself this week was to replace my entire sock drawer. You know that feeling of getting to the bottom of your sock drawer, where all that’s left is the crappy socks? I realized I didn’t need to put myself through that. I threw out all of my crappy socks (that have holes, that let me get blisters) and replaced them with amazing socks. Here’s my favorite white cotton tube sock, thin trouser socks, workout socks, and best of all, the thick wooly socks. I kept my Darn Toughs around for hiking, of course. Controversial opinion: down with Bombas! All of my Bombas had holes.