really, the issue is that i am not part of a collective
a good software collective is established with two primary goals :—
1. creating good software
2. generating a public record of documents which demonstrate the process of making good software
considering this, one might be forgiven for thinking that there must exist an overabundance of tools for generating and proliferating documentation on the creation of software. after all, this is a problem that every good software collective must solve. however,
@Lady (You have an eye on what Worm Blossem is doing, right?)
(Both in terms of "software collectives, and documentation", and in terms of "sharing documents".)
@gaditb no, never heard of them
@Lady Oh!
Okay gimme a sec to grab some links.
Do you prefer FOSDEM talk (aesthetically highly based on Legend of Zelda, primarily about their design approach/goals and how it informed their spec-writing),
protocol specs,
protocol conceptual documentation,
example application (running but still hacky and using only a scarce few of the features in their design-space),
or collective devlog/shared blog of how things are more generally?
or fediverse handle of one of the people in it, I guess.
(... grabbing...)
@Lady
Main site / devlog+ : https://worm-blossom.org
Projects page: https://worm-blossom.org/projects/
"Willow": Data Model/Protocol that is their primary output/locus-point of effort: https://willowprotocol.org
(Has data model explainer, protocol specs, links to their other spec-work this builds on)
"Sneakerweb": Application built on it, intended to build an experience approaching zines, uses directories of HTML. Uses their Willow (Willow '25 specifically) for serialization and for authenticated updates to prior content.
https://sneakerweb.org
FOSDEM talk: "Willow - Protocols for an uncertain future"
https://fosdem.org/2026/schedule/event/CVGZAV-willow/
@Lady Specifically relevant to their documentation (copying descriptions straight from their Projects page)::
WebTT: Lightweight tooltips for the web, without any server-side logic.
https://worm-blossom.org/webtt
Macromania: A macro-based system for generating plaintext files, including this very website.
https://codeberg.org/macromania/macromania
(Maybe there's more of their process/tooling available/documented but I don't know?)
@Lady (... I'll be honest frankly I based on their aesthetics/approach/ideology I think I assumed you must have known about them before I did,
or possibly been indirectly involved.)
@Lady (I think probably you will either go for one or the other -- or switch off both, I don't know how you process attention -- of their devlog+ [ https://worm-blossom.org ], immersing linearly in the way they approach themselves & their process & what-happens-alongside-thst-informs-it (if you want to go fully this route they have their archive forward-chronological from the beginning: https://worm-blossom.org/archive/full/ )
and the Willow data model & spec documentation, exploring network-ly from the Start Here overview of the data model: https://willowprotocol.org/specs/data-model/index.html#data_model )
(I don't know how you feel about video-of-conference-talk-as-way-of-intro, but maybe that ~20min talk. That was my first intro and set a solid base.)
(I think the thing it took me longest to grasp in the concepts was the time/authority/data-path 3D axes they kept drawing in their visualizations. That it was those 3 axes (.. I'm pretty sure), and why it was meaningful they were independent.)
@Lady (You don't need to look at or respond but if I did hear your thoughts on them I'd be interested in your thoughts.)
@gaditb ah, well, you might need to be more specific on genre of thought. technological? rhetorical? social?
@Lady ... whichever you had your first thoughts on? I like hearing your thoughts.
(Possible specifics are "how well/interestingly do they do the documenting-themselves stuff that prompted me to mention them in this particulsr thread". but if you think other thoughts are more interesting,)
@gaditb Regarding communications: I think their website is cute. It is something they have put a lot of care and heart into and I think it shows.
It also very obviously is something that takes a lot of time. And I feel it is important to point out unless you are grant‐funded or have tenure (a very very small minority), time is not a luxury that most working programmers have. Most people writing open‐source code for their jobs are programmers at understaffed public (or at least public‐interest) institutions with portfolios that vastly outsize their workforce, and people who write open‐source code in their free time are typically scrounging for a few hours outside of work that they have to balance with their families, their welbeing, and their other interests. This blog exists in a different world from those programmers. It is not a blog written by a labourer with too little time and too many responsibilities trying to convey the barest essentials of their work and practice to another labourer with too little time and too many responsibilities. This is not a criticism, exactly, but it is necessary to acknowledge the ressentiment that this evokes. Culturally, worm-blossom exists in a space that is something like the Ivory Tower of programming.
@gaditb Analysis along the cultural-technical front reinforces these feelings of privilege. The first implementation of Willow was written in TypeScript; the current implementation is written in Rust. Small teams choose programming languages based on the comfort levels of their members; this TypeScript‐to‐Rust pipeline suggests a certain background and history which, if not directly a participant in the startup culture of the 2010’s, was at least attending the same conferences as them and sharing code, methodologies, and worldviews. It is rude, of course, to overindex on these choices when these are two of the most popular programming languages in use today, but these things do send signals about who someone is and who they associate with. TypeScript was a core part of Microsoft’s strategy to capture a share of the developer market (a share which they are now trying to cash in on with Copilot). Rust is now a primary target for AI investment and vibecoding. It is, if nothing else, kind of embarrassing to constantly find yourself making the same technological choices and sharing the same spaces as these kinds of people.
@gaditb On the purely technical standpoint, I was annoyed but not surprised at Sneakerweb’s enshrining of `index.html´ as the directory root to the exclusion of alternatives like `index.xhtml´ or `index.xml´. It is of course the opinion of WHATWG that the XML syntax of HTML is a legacy format which should not be used, and it is a common opinion that HTML is friendlier to novices that XHTML, so this is an understandable piece of ignorance. However, you have to ask yourself whether the goal of Sneakerweb is just to accommodate indie handcrafted websites with bright stickers and cute pieces of string, or also more formal and durable repositories of value like, say, ebooks. Even a cursory exploration into EPUB would reveal that it is just a Zip file of XHTML, and from this logically one can assume that most ebook publishing workflows are probably XML‐based, so failing to account for XML in your software gives the distinct impression that you do not fully understand the problem space you are trying to solve.
You can only get so far trying to invent programs for publishing without talking to any publishers, programs for library science without talking to any librarians, et cetera. The impression I get from worm-blossom is that they exist primarily in a bubble of technologists with no real stakeholders, and they come up with solutions which sound interesting and trendy to technologists but lack the problem statements, requirements gathering, and necessary compromises that are inherent to solutions for stakeholders. A pithy way of putting this would be to say they are architecting but not engineering. It is frankly probably impossible to do any real engineering with a team of two or three people, so this is not exactly a criticism, but it does call the longevity and applicability of their solutions into doubt.
@gaditb This all sounds very critical so I want to close by saying that I am not trying to harsh their vibe and actually want them to keep doing what they are doing. At the same time, I don’t think they do or can serve as a useful model for working programmers who are trying to build community and share knowledge with other working programmers in the limited time they have on or off the job. Nor do I think it necessarily makes a very good demonstration of the process of conversing with stakeholders and refining issues based on their needs, which is the bulk of the task of software engineering. It is mostly theory and manifesto, with the manifesto sometimes taking the form of code.
