-- Leo's gemini proxy

-- Connecting to zaibatsu.circumlunar.space:1965...

-- Connected

-- Sending request

-- Meta line: 20 text/gemini

Gemini aggregation, curation, etc


I wrote previously in my pikkulog about how while acdw was starting to feel frustrated that new content did not appear in Geminispace quickly enough - relative to the "endless firehose" expectation that the mainstream web, and social media in particular, have forced into our brains - I was starting to worry about the fact that new content was appearing too fast, and on too many different servers - relative to the "it's not that hard to read almost everything" expectation that a few years of participating in Gopherspace had forced into my brain. This topic has been bouncing around in the back of my brain ever since, and was brought to the surface again today when Meff mentioned that "Geminispace is exploding", that they "haven't really been reading everything on CAPCOM", and, most importantly, that "Geminispace is starting to have a lot of content, but other than GUS, a lot of the content seems (to me, at least) difficult to find and organize. I'd love to see more curation of content". The curation thing was well-timed, because I had been wondering about this ever since I stumbled upon a little section of iolfree's capsule which, as far as I can tell, is a little index of pointers to places around Geminispace they find interesting.


acdw's post

See my pikkulog entry from 2020-07-17

Meff's post

iolfree's list of links


Whether we're talking about Gopher, Gemini, or the web, it's well worthwhile thinking about *how* we discover content, *when* we discover, from *whom* we discover content and *why* we discover content. In line with a lot of earlier discussion in Geminispace, about blindly copying ideas from the web and/or from commercial services, I think it's important to be careful and mindful in building tools for content discovery and aggregation, so that we end up with an online experience that's healthy and beneficial. Some miscellaneous thoughts on this follow:


Gemini as a FOMO-free zone


Fear of Missing Out (FOMO) is a bad thing. Advertising funded services deliberately foster it because they want your eyeballs on their ads as frequently as possible, but it serves no purpose in a free, non-commercial environment. The idea that reading everything which happens in Geminispace and responding to everything of interest in a timely fashion is important, or desirable, or even *possible* is, IMHO, a bad idea and we should try to resist getting it into our heads. In earlier days it made more sense, when we were trying to encourage the growth of a fragile seedling. It makes less sense now, I think, and if current rates of growth continue or even increase it will very soon make no sense at all. We shouldn't design or build tools, services, workflows etc. which resolve around this idea.


You should be in control of what you read and when


For many people, content discovery on the web is largely achieved by following links which appear on their Facebook, Twitter, Mastodon, whatever feed, or following links they find on link-sharing sites like Reddit, HackerNews, etc. This is an easy way to be assured of a never ending stream of constantly new content (which *might* be something you want, although I encourage people to question this), but depending upon specifics it can also effectively mean handing control of what appears on your screen day after day (which can very directly influence your mood, your beliefs, your biases, and more) to a large group of strangers. It also means that you read stuff when it happens to appear in one of these sources, which may not be when you are in the mood for it, and may not when you have the time to dedicate to reading it closely. I think it makes a lot more sense for each of us to take a more active and mindful role in deciding what we read and when. Of course, some degree of "outsourcing" content discovery is unavoidable, but perhaps it's best to limit it to carefully chosen, trusted sources, and to decouple the sharing of links from community discussion of said links, to remove the incentive to read everything as soon as you become aware of it.


CAPCOM and Spacewalk are the beginning, not the end


I setup CAPCOM as a "public aggregator of all the things" early on in the "fragile seedling" phase of Gemini's life, and Spacewalk was setup at roughly the same time. Both of these are still functioning pretty well, and I still check CAPCOM daily. It's the centre of my personal Gemini experience. I also check Spacewalk a few times a week, and stick my head in at a few other places too, like gemlog.blue. For small spaces, services like CAPCOM and Spacewalk (and their Gopher equivalent, Bongusta), work extremely well. They are vitally important, in fact, for bootstrapping new spaces.


But as Geminispace grows, and more and more feeds are added, these kind of public firehoses are going to become less and less useful. The sheer quantity of new content each day will reach the point where it becomes onerous to try to read everything. The proportion of content which any one individual reader genuinely finds interesting or engaging or pleasant or inoffensive will necessarily get lower and lower (nobody honestly likes reading *everything*!). Continuing to use them the way people are using them now is eventually going to violate both of the principles above to some degree. The current scenario probably still has many months of life left in it, but I'm not convinced it has many *years* of life left (unless growth of the space slows right down and it never gets much bigger than it is now - which very well may happen).


Fortunately, both CAPCOM and Spacewalk aren't *just* public firehose feeds - they are software tools you can run yourself, to make your own personal aggregators. You can share your aggregated results with the public, or you could lock them away behind a client certificate so nobody but you knows which Gemini content you like to read. I hope this approach becomes increasingly common as Geminispace grows. I could be wrong, but I don't think many folk are doing this yet.


Sharing is caring!


For a little while I was getting myself pretty excited thinking about a shiny new way for me to find new content, which aimed to combine the best features of Atom or RSS based approaches (like CAPCOM) with simple change detection approaches (like Spacewalk). Maybe something will still come of it, but I cooled down on the idea a little when I realised that I preferred a much simpler, more organic approach which didn't generate any extra automated traffic (not that automated traffic is fundamentally *wrong* - on the contrary it can make wonderful, valuable services possible - but I think it's better to generate it sparingly and carefully): which is simply to discover new stuff in Geminispace when other people I already follow talk about it and link back to it. I very much like the idea of using my own personal instances of CAPCOM and/or Spacewalk to follow the output of a relatively small number of carefully chosen and trusted authors, and relying on the idea that if somebody outside that circle writes something really good on a subject I care about, somebody else in my circle will hear about it through *their* circle and spread the word. This should work quite well as long as people have at least slightly non-overlapping circles. But it only works if people do, in fact, share the content they find and like, so let's all try to do that!


Final random thoughts


Multi-user Gemini hosts with an organising principle - like breadpunk.club for people interested in baking bread, or tanelorn.city for people interested in writing science fiction or fantasy - are a great idea that facilitates straightforward content discovery, and I hope we see more of them.


There has been a little discussion on the Gemini mailing list just recently about the idea of bringing "webrings" to Gemini. This also sounds very promising.


Whatever content discovery solutions we end up building, let's do our best to keep them simple, decentralised, human-scaled and user-empowering!

-- Response ended

-- Page fetched on Fri Mar 29 00:56:04 2024