gemini://zaibatsu.circumlunar.space/~solderpunk/gemlog/announcing-the-smol-earth-compendium.gmi

One is that people (mostly critical outsiders) keep claiming that Geminispace is just a bland and one-dimensional place where nerdy tech people go to write posts to one another about nerdy tech things that nobody else cares about. Anybody reading this likely knows that this is not actually true (if you somehow still need convincing, I am slowly accumulating a page dedicated to linking to as broad a variety of long-form non-technical content as I can, link below!), and I shouldn't really let this get to me. But it does, and I want to help turn the non-tech content on smolnet up to 11 to help put this myth to death once and for all. I want to explicitly encourage and incentivise the publication of non-tech content on Gopher and Gemini and make it very easy for people who only want to read that content to find it.

Another is that I want to stir up enthusiasm and discussion and experimentation in the area of new and different and better ways to organise the smolnet and facilitate finding interesting content. In both Gemini- and Gopherspace, the dominant paradigm here is the "waterfall aggregator", in the form of Bongusta for Gopher and things like CAPCOM, Antenna and Cosmos for Gemini. Let me be clear that I enjoy and use all these services from time to time, and even run one of them, and nothing in this post should be construed as a criticism of these fine projects. I would like to see them augmented, not replaced. A second paradigm is the categorised directory, exemplified by the Gopher Lawn or a variety of similar projects for Gemini.

Both these paradigms are typically characterised by a lack of any particular thematic focus. The feed aggregators aggregate large swathes of their respective spaces (although there is no shortage of quality content that's off their radar, and random exploration is oh so rewarding!) regardless of what they're about (and this helps add to the perception that there's only tech talk going on, because the non-zero-but-minority non-tech content is scattered amongst the majority tech content). The directories do naturally organise things into categories, but I haven't seen any yet which specialise in a relatively small range of related categories, they all seem to be efforts to organise entire spaces. Many of the Gemini directory projects were started with enthusiasm in the early days of the protocol but abandoned during the explosive growth in 2020 when it became hard to keep up (though some are still going). Nobody should blame anybody for this. Some of them were single person projects, and for a single person to try to categorise an entire space is a colossal task, which only gets harder as the space grows. It doesn't really make sense to even attempt this. But it does make sense for multiple separate teams of people to work together to categorise just a subset of a growing space, the subset concerned with stuff they care about. Divide and conquer!

Finally, over the six and a bit years I've been active on the smolnet, I have spent a continuously increasing amount of my time being outdoors or wishing I was, and thinking, reading and writing about environmental issues, and I don't expect that to change any time soon. In the past six months or so I've also become quite interested in the way that information was organised and distributed in the 1960s US counterculture movement, particular the small, low-budget book and magazine publishing enterprises that sprung up around it. I'm talking about things like the Whole Earth Catalog and its spin offs, or the Domebooks and other products of Lloyd Kahn's Shelter Publications (early this year I snagged a 1973 edition of "Shelter" on eBay and it's a wonderful thing). So as my motivation to do something concrete to address points one and two above has increased, I've naturally been thinking about leading by example by starting a project which ties into this outdoorsy-greeny-hippy lo-fi alternative media bundle. If you're thoroughly disinterested in that bundle, as no doubt some of you are, I do hope you'll read on anyway and keep a loose eye on this project, because I hope that if it is successful it can be a role model, both conceptually and in terms of technical foundation, for other similar projects on entirely different topics, maybe something you are actually interested in.

> The Smol Earth Compendium is a community-curated collection of links to Gemini and Gopher content related to the natural world and the great outdoors, especially content which fosters a deeper sense of connection to nature, encourages long term holistic thinking about the natural world, or contributes to an increased awareness that all human activity takes place within the natural world, is inescapably dependent upon that world, and has unavoidable impacts on that world.

I hope the Compendium will grow to be a one-stop-shop for smolnet greenies to find interesting content both new and old. If you've been toying with the idea of writing on subjects within its scope but you've been unsure whether or not anybody wants to read that kind of thing, or you've worried that even if they do they'll struggle to find it, I hope the launch of the Compendium will put an end to that and encourage you to start! At the moment it is just a declaration of intent and a rough technical sketch, not backed up by code. But I desperately want it to be more, and I'll start writing code to make it a reality with speed and enthusiasm proportional to the amount of interest and encouragement I get back from this announcement, and if I can find some volunteers to help me get started.

In deciding how a project like this ought to work, I had a few criteria in mind. One was that I didn't want to add any extra workload to the people who actually write content for Geminispace or Gopherspace. This immediately ruled out the idea of asking everybody with a phlog or gemlog that includes posts about more than a single topic (which is 99% of them) to maintain a separate feed or index file for each topic, so that I could just grab some existing aggregator software and point it at relevant URLs and call it a day. That would be a huge chore for authors, a big turn-off, and most people would never play the game. It could work relatively well on the web, where it is basically a given that you generate your blog using some kind of CMS toolchain that can easily generate all those separate feeds for you if you tag your own posts, and even though some people do use tools like that for their Gopherhole or Gemini capsule, plenty don't. I sure don't. I maintain mine by hand, always will, and love that these technologies are simple enough that I can do that and it's not even hard. It's a big part of the appeal. I don't want to pressure people to automate what's easy and fun to do by hand.

Another criterion was that I didn't want to create a burdensome moderation workload for myself or anybody else. I could just write some CGI scripts and ask authors to submit the URLs of their individual posts within the scope of the Compendium, but that would be vulnerable to spam submissions of off-topic content, and without moderators in the loop it would only be a matter of time before some clown submitted something offensive which automatically turned up in the Compendium and made us look bad. No good. Plus, this is still burdensome to authors. If this idea takes off and people setup similar projects for different subjects, then the expectation is there for authors to keep track of all those projects and to manually submit their individual post links to the relevant place(s) every time they make a post. That's no fun at all, people will just opt out and the whole thing won't get off the ground. The workload should rest entirely upon the shoulders of the folks who want to see the organisation and categorisation happen, not the folks writing the content (of course, there can and probably will be overlap). In my ideal smolnet world, writers would not feel the need to take any more than the bare minimum of active steps to ensure that their content is found. They should write what they want to write because they feel the burning need to write about it, and they should do it understanding that in the fullness of time somebody, somewhere, somewhen will find it, and it'll resonate with them and maybe they'll share it. We should work harder to incentivise exploration and sharing of discoveries than we should to incentivise self-promotion. Not everybody will be onboard with this vision, and that's fine, but it's the one that I want to personally work toward.

The solution I've arrived at is as follows. I'd like to assemble a smallish team - perhaps up to a dozen people at most - who like this idea a lot and are willing to put some time and energy into bringing it to life. The only requirement aside from being into nature, the outdoors and the environment, is that you have some presence in either Gopherspace or Geminispace or both. That's because when you volunteer to be a curator for the Smol Earth Compendium, you don't have to setup an account on the smol.earth server - that sounds too much like work on my part! Instead, all you have to do is maintain a single file in a very simple format (basically a list of URLs for individual posts and associated tags), hosted on the smol protocol of your choice on the server of your choice, updating it by whatever means you prefer every time you find a new bit of content that you believe should be included in the compendium. You email me once to let me know the URL of your file and then I'll use a cron job to fetch a copy of it once every 24 hours. I'll write code to combine the information from everybody's curator file into some kind of consensus directory and spit it out in gemtext and gophermap format - the result will look roughly like the venerable del.icio.us "social bookmarking" service. If you don't want to be involved anymore, just delete your file and that's it, you're out, simple as that. If you consistently submit low quality, off topic content, I'll stop fetching your file, simple as that. No accounts, no usernames, no profiles, no client certificates, none of that. Static files fetched via cron job and boiled down into more static files.

The really exciting part is that in addition to outputting the consensus directory in a human-usable Gemini and Gopher format, I'll also publish it in exactly the same format as the original curator files from which it was built. If other projects in the future do the same thing, those compendia can themselves be combined into larger collections using the same software. It scales down as well as up. We could have obsessive individual curators building fine-grained directories of very niche topics singlehandedly, and other people could aggregate multiple of those directories on related topics up into something a little broader. Maybe those tiny niche tagging projects will only add new items once every six months, but so what? It's not hard to just leave a file sitting there most of the time and add a few new lines every year. If you forget that it exists entirely, that's no big deal. People can still fetch it, combine it with others, glob stuff together and remix it and build large, broad, structured, frequently changing things out of a pile of multiple smaller, simpler, slower moving projects. Doesn't that sound neat? Really simple and really lightweight, entirely distributed, but with so much potential. There are interesting questions and details to figure out around this basic idea, and I hope the Smol Earth Compendium project will provide some useful data on how a simple and naive version of it fares in the real world.

If you think you're up to maintaining a curation file (there's some more information on exactly what they'll look like at the Compendium itself), give it a try! Create a file with at least three URLs in it that you think are worth including, and tag 'em up as you see fit. If you already have stuff bookmarked on relevant subjects, that's an excellent sign that you could be right for the job! Email me with the URL of your file. If enough people do this and I think the results look promising, I'll get hacking, and curating myself. I *want* people to volunteer for this, and too many is better than too few, but please don't do it on a whim, and please see sticking your hand up as a pledge to put some continuing effort into this project, which I hope to keep running for many years. We'll need at least a few people who check in on aggregators like Bongusta or Antenna semi-regularly, say once a week at least, who can keep an eye out for new in-scope content as it appears, but that won't be enough. We'll need some people who genuinely enjoy spending time exploring pubnix user directories to help dig up old and un-promoted content from accounts abandoned years ago - sometimes that stuff is pure gold. It would be great if there's somebody who already has the habit of routinely checking DiscoGem or some other "random capsule of the day" service (is there one of those for Gopher?) who is also a greenie or outdoorsperson who can recognise when they stumble upon something within the Compendium's scope. A few of the Gemini search engines also publish lists of the most recent newly discovered pages or hosts, and it would be great to have somebody watching those too. I'm making this sound like more work than it really is! No individual curator needs to work very hard at this. Nobody needs to regularly check more than one or two of the things listed above, and "regularly" sure doesn't have to mean anything remotely close to daily. It's fine to take time off. If you get bored after a few months you can retire and somebody will take your place. This will be a slow and steady cumulative project and there are no deadlines and no sense of urgency. We'll never find every single thing we'd like to, and that's fine. I'm just trying to gather folks who think they can plod along at this task, even at a slow pace, for months or years without getting bored of it. You should already be an enthusiastic reader of Gopherspace and/or Geminispace, natch, and you should already be hungry for more green content, that way it won't seem much like work at all.

I'm really excited about this! If this goes well, I will consider expanding the Compendium's services, for example to include a full text search engine for only the content in the Compendium. I'm very keen on the idea of small, domain-specific search engines. I'm also already bursting with ideas for other projects under the Smol Earth label, inspired by various "Whole Earth" projects, but I'm trying to pace myself. If the Compendium goes well, I'll try to launch some other things in 2024.

Announcing the Smol Earth Compendium