“The Lost Apps of the 80s” never went away
A response to “The Lost Apps of the 80s”, originally published on Secure Scuttlebutt at %rweoPItwNqwpL6yM7NNph6nljjTUAWXdUkvosSJJ/o8=.sha256
Sometimes Dave Winer says smart things, though there are often strange gaps (maybe he is like Kevin Kelly, who says smart things because he has surrounded himself with smart people but generally can’t reconstruct the reasoning behind those things).
In this case, the difference between the 80s and today with regard to writing tools is sort of clear: in the 80s, most available tools for home computers either were small-computing products or (as in the case of Lotus) began as small-computing products and grew due to popularity. Small computing products are quirky because they are created by a single person or a handful of people to solve their own problems in a way that corresponds to their own workflow; as they grow, they tend to become highly customizable application platforms with reasonable defaults because expansion occurs through supporting new users with new workflows.
The browser text field, like most webtech, comes out of someone saying “this theoretical feature is something we ought to have in order to compete; let us consider abstractly what we, as developers who touch only a tiny minority of use cases, would expect the minimal feature set could be in order to satisfy all users”. In other words, a big tech mentality from the start. You get a committee together to imagine what a very distinct set of unrepresented users might want, and then the developers get a checklist of intended behaviors, some of which contradict each other. Then it gets standardized or de-facto standardized, and you get almost identically-behaving implementations in all the major browsers.
Tiny and quirky writing tools still exist, still are in use, and still are created. Vi and emacs are arguably quirky writing tools, and are widely used. I wrote a specialized text editor that had support for color coding word black/whitelists, recommending the next word via markov models that could be easily swapped out (or have their probabilities reversed), rhyming dictionary and thesaurus lookup, and S+7 replacement. Robin Sloane wrote a plugin for Atom that tried to complete his sentences with char-RNN. Botnik Studios uses a specialized “keyboard” that’s really a markov-based autocomplete oriented text editor. And, of course, we have a wide variety of proprietary ‘boutique’ editors like Scrivener and Hemingway. On top of that, there remain specialized editors for proper hypertext (like Twine, which though it produces web-compatible output, really has a lot more in common with 80s hypertext-fiction editing systems like Hypercard and StorySpace).
These systems don’t get a lot of exposure because there are lowest-common-denominator systems that eat up most of the bandwidth for familiarity for most users. Word is an ascended-quirky-writing-tool that, because it became institutionally mandated, had its quirks sclerotized into standardization; it and its bug-compatible clones (Google Docs, libreoffice) are used by everyone. Completely minimal editors like notepad are everywhere too. It’s been a long time since anybody thought it was a good idea to ship a computer with a program as quirky as LotusNotes by default.
Basically, if a field has incumbents, then so long as you are deciding what to look at based on popularity, you will not see quirky fringe experiments and specialized tools. When a field is young, quirky fringe experiments and specialized tools are all that exists, but they can never grow to dominate without losing their quirk and specialization.