Running Lego thread

Wanted to start a running thread to talk about Lego. In some ways the most accessible and familiar example to talk about protocols.

Recent conversations on bluesky and farcaster have convinced me I was wrong to be skeptical of the value/power/expressivity of Lego. In particular, I’m impressed by the ability of builders to develop radical curved forms. I mistakenly thought the “grain” of a model couldn’t change easily, let alone smoothly, because of the way the studs on standard bricks face in one direction. So the protocol is more expressive than you would guess, especially at larger scales where you can exploit effects like joint play that are too small to be useful at small scales.

This build for example, uses the circular frame part of Harry Potter glasses from a hog warts set along with a 1dof hinge to create a curved spine around which this dragon is built. https://www.youtube.com/watch?v=8bM0Lax0a0w

And here is another video with curved builds that exploit play/wiggle room in the tight tolerances of non-hinged regular bricks to create surprisingly robust curved forms.

I no longer think the system is a closed walled garden platform pretending to be a protocol. It’s a true protocol. It has a bias towards skeuomorphic artistic representation over functional engineering (even with Technic) but that’s fine. It’s just more of an art/sculpture protocol than an engineering one.

1 Like

I’ve been thinking about Lego from the perspective of “all models are false; all frameworks and protocols are provisional”:

  • Protocols can appear more limited and constraining initially than they are in practice.
  • Human creativity and curiosity can overcome perceived limitations in protocols by identifying and accessing overlooked degrees of freedom, wiggle room, and other latent potential. What looks fixed from one perspective may prove an “enabling constraint” when approached from a fresh perspective or different scale.
  • The possibility space of many protocols only fully emerges over time as new generations of designers, builders, and users test boundaries, shift perspectives and uncover new affordances. Lego creations once considered implausible are now possible by reimagining structural capabilities.

Looking ahead to the prospect of AI-mediated communities, the gradual timeline for exploring possibility spaces in protocols is likely to be accelerated. Instead of spanning generations, I anticipate AI and automation will be able to rapidly test boundaries, shift scales, and generate ideas for human consideration within a matter of years or even months. The implications are fascinating given that protocols are perpetually provisional. Their possibility spaces will be revealed not through the slow march of time, but through stretching of their seams by both human ingenuity and automation.

1 Like

I’m always wary of exploiting wiggle room because you never know when they might tighten it up and kill your whole thing. It’s like closing loopholes in laws or tweaking APIs.

And of course there’s an xkcd for that. xkcd: Workflow

I think the people who are able to harness the value and expressivity of Lego are a small fraction of a percentage. From talking to kids/parents recently about their experience of Lego, it does seem like they’ve moved away from interchangeable blocks that start with the value of expressivity to specific kits (harry potter, bionics) that are more build once and display/play with forever. Leads to more total legos being sold, but lowers the overall builds.

I think there’s an analogy there worth noting which is that the popularization of a protocol can lead to managed, specialized kits that limit the experience of a substantial portion of the population (like blockchain API wrappers that manage decentralized operations by abstracting away complexity and creating another layer of centralization).

It’s interesting to see LEGO discussed here independently of the terminology and platforms I got used to thinking about it in terms of during my brief foray into online LEGO communities in the early 2000s. Being introduced to the sideways/alternative building options of “SNOT” (studs not on top) and “SNIR” (studs not in row) as specific new methodologies felt revolutionary, and the fact that they were encapsulated under such definitive terms maybe made them seem more like inventions or breakthroughs.

It’s also funny in general how so many abbreviations get thrown around (below is a list including some I’m not yet familiar with); a lot of the anacronyms are of course juvenile-sounding, maybe reflecting the nature of LEGO as a toy, but the repeated reliance on abbreviations also feels slightly military-esque in a way.

Other websites which felt like they used to be the center of the online LEGO community world: https://www.lugnet.com/ (almost like a proto-Reddit, but only for LEGO), and BrickLink - Buy and sell LEGO Parts, Sets and Minifigures for buying/selling individual pieces.

I also remember there being a specific branch of the LEGO community catering to people who had a more general interest in model railroading which focused on making model railroads in LEGO, and they kept a little bit more to themselves where you wouldn’t necessarily hear about everything they were creating unless you sought it out. (Edit: here’s one of those communities I remember from that time: Welcome to PSLTC.org)

Another collaborative activity I just remembered was the Moonbase project which IIRC Lugnet organized. Different people would build modules in a stilt-elevated interconnecting moonbase, with standardized docking heights and dimensions between modules, and then sometimes combine them together during IRL meetup events. Examples I’ve found of people following the standard: My Moonbase Module – BrickBuildr
LEGO at Hafhead.com: Misc. Moonbase Modules gallery
t-reichling.de
Reddit - Dive into anything (more recent)

Combined moonbases following the standard (links here are automatically showing up as images):
Imgur
Imgur
https://brikwars.com/wiki/index.php?title=File:P1010360.JPG#filehistory

This is great info! @dorian mentioned SNOT on bluesky but couldn’t recall the expansion.

Do you know of anywhere that has stats on kits and parts like what percent of parts used are reused from older kits vs designed new for that kit, and how many kits use a given part? I’m interested in statistical research into the protocol level openness/interoperability etc.

Current state of my stone soup meta-kit

In general (at least from when I last kept up), new parts are introduced when LEGO introduces a “theme” of several sets with a similar aesthetic or subject matter at a time. The vast majority of parts appear in multiple different sets, though there can be a few more rare and unique parts. I’d assume that in general once LEGO makes a mold for a part they’d want to re-use it on parts for multiple sets. Rare and unique printed patterns (or if the company is being lazy, stickers) on standard parts are more common than unique part forms.

For seeing which pieces came in which sets I used to use Peeron™ LEGO© Set Inventories, which looks like it hasn’t been updated since 2012. BrickLink Appearance In Inventories also has what might be a more recently-updated directory where parts can be looked up (to see what sets have them) if you retrieve their numbers from another page (like the set listings) on the site.

I don’t remember any site offering good statistics about sets that I saw in my past experiences. https://brickset.com/ looks like it might have a few set-specific curatorial guides, and Rebrickable | Rebrickable - Build with LEGO looks like a new site focused on sets which I haven’t heard of before.

In general the parts are very interoperable (though there’s plenty of really obvious and intuitive limits like pole-based pieces w/o stud holes, such handheld objects, being able to connect with say open studs or hooks but not solid studs), but there have been edge cases of “illegal” builds identified where certain combinations can put excess stress on the pieces, provide inadequately secure connections, or in some cases even be relatively irreversible combinations, and IIRC the company doesn’t authorize doing them in official sets. Here’s a PDF that’s from back in 2006, but it’s nicely documented: http://bramlambrecht.com/tmp/jamieberard-brickstress-bf06.pdf (Also had an HN discussion: “Illegal” Lego Builds (2006) [pdf] | Hacker News) SNIR in general may be kind of on the “illegal” side of things from the company’s point of view, though I haven’t looked into it. I also IIRC remember reading about how LEGO’s manufacturing quality is/was very high, and that might factor into them setting standards for legal moves to not impact their high expectations of long-term durability.

Also another building technique I forgot about: in addition to SNOT/SNIR, there was also a “studless” building trend where builders tried to cover up every stud on their models with plates and/or sideways construction. I thought it was particularly elegant when done well, and there were some neat builds particularly among the community of people building space-related creations. Paging through the archives of https://news.lugnet.com/space/moc will probably reveal some neat approaches.

I’d guess there might be some kinds of existing online guides and resources on Reddit or somewhere specifically for adults returning to LEGO, or adult beginners. The term I used to see used for people getting into LEGO again as adults was that they were returning from their “dark age” where they stopped playing with it between childhood/the start of adulthood, so that could be a possible term to search for.

(Edit: another interoperability issue I just remembered: the early 00s controversy when LEGO made the color of grey they used slightly different - slightly bluer - from before, so new/old bricks no longer exactly matched in color. There was a big uproar about that from adult customers, which the company made statements specifically in response to. The term fans used to criticize the new color at the time was calling it “bley.”)

Studs Not On Top