Thoughts On The CDG, And How You Can Help FightThe Discoverability Crisis

Recently the 1 week release of CDG post came out by the author, and they asked for feedback. So here we are.

CDG one week later

One problem I have always found, as have many others who try out gemini, is difficulty in finding interesting content. There are many contributing factors to why, but lets just say the problem is an ongoing issue that has no real easy solutions. Search engines and feed aggregators help out a lot and are great places for newbies to start. Many capsuleers also have created public collections of links they find worthwile to share which is helpful when you stumble upon them.

The Discoverability Crisis

However; unfortunately, still much awesome content and useful services slip through the cracks of many peoples awareness. Either they aren't indexed well, mentioned often, or properly advertised by the author. Some great content on the geminispace will be missed because its hard to find. Sometimes you need to be lucky to stumble across them, and have the foresight to save and share the links.

There is a great post I saw a while back about how the best thing you can do for the small net protocols is to create your own public bookmark indexes/link collections featuring cool stuff. If everyone did this things would be much easier to find. After all, Whatever site you think is mundane and obvious, someone else may have never heard of.

The Small Net Discovery Crisis by Marginalia

To help answer this 'discovery crisis', a brand new service has been born't known as the Collaborative Directory of Gemini. Instead of many individual users maintaining their own seperate public list of links, the CDG acts as a hub for all users to build up a singular, well catagorized database linking the best content gemini has to offer. Anyone is free to submit a link, though the host is the one who has the final say in curation.

Power Of FOSS

As a bonus, the CDG is liscenced under the Creative Commons Attribution-Sharealike 4.0 licence by the CDG Contributors. The code itself is released under GPL. These are very permissive liscences which make the FOSS nuts like me happy. This means that if you don't like how things are structured and think you can do a better job, then you can build off the pre-existing (and continuously updated) builds of the directory. This is great for ease of preservation.

It would be cool if there were a text file at the root that contained every single URL listed in the order of the full directory tree. One text file you could download to have a offline directory.

Also I would like two navigation links at the bottom of each page one to go back a step in the directory and one to go back to the CDG root. This would be awesome!

efforts in submitting

I found out about the CDG on the 18th and since then have contributed a sizable portion of the URLs that are in now. I was very excited about the project from the get-go, and having a great repository of my own links to feed in. I don't think poor mbay could keep up with the rate of my submissions for a while there lol. Sorry about that. Things will slow down now that I have run out of my pre-existing list of bookmarks. It was fun to see how the CDG adapted and changed over time to better fit the links. I always like seeing growth and change.

Its very difficult to get a good catagorization system going. Too little catagorization/directories and things become big list of links with not much in common besides one or two themes. Too much catagorization/directories and it becomes a bitch for the user to navigate the maze of subcatagories within subcatagories within subcatagories, not to mention as a submitter it becomes more difficult to tell what should go where.

Theoretically better systems have been spoken about like the dewy decimal system. In reality, practicality usually wins out. Whatever works best with least effort is good enough for now especially when the project is first starting out.

Overthinking things and needlessly overcatagorizing is probably not the best option. Trying to be too rigid with the rules will make the whole thing brittle and clunky. Whatever makes intuitive sense is probably the right choice.

If you were to ask me for a constructive suggestion though, I like how the textfiles.com archive is catagorized the most and perhaps the CDG could benefit from studying how they do it.

gemini://gemini.spam.works

> Does a glog devoted to a specified topic belong in the corresponding topics/ category?

I say yes. If the glog revolves around one topic then it should go in topics.

> Currently "Gemini-first" content is separated from mirrors of websites and so forth. My thinking was partly that if you come to the archive hoping to discover what interesting content you can find on Gemini, you'll be disappointed to find it full of copies of things you could have found on the web. But I'm thinking of relenting on that and moving mirrors into the topics/ hierarchy where this makes sense, in particular for the contents of mirrors/archives/.

I agree fully with relenting. Most of the archives would be be better suited in a more apropriate catagory. While I like all the mirrors and archives being in one place, they don't really have much to do with eachother besides that they are archives of content produced elsewhere. Most people aren't going to care if its a mirror/archive as long as they are interested in the content. Plus, if the pages are properly converted to gemtext, its like a whole new experience from the web version so theres a novelty factor (at least for me). I have never felt disapointment from stumbling across a mirror/archive, instead its like concentrated information distilled into its most essential form free of the bloated web crap.

> A few links to individual glog entries have been added. I'd like to just blanket-ban that practice, on the basis that the Directory would quickly be overwhelmed if it became standard practice. Also because we already have the aggregators to handle gemlogs, and I see the Directory's purpose as orthogonal to theirs. But for now I've been avoiding deleting submissions, since it's happened a few times that my first reaction was that something didn't belong in the Directory, but trying to fit it in led to adding some sensible category. And there are a few cases where it would be a shame to have to delete the link, e.g. the only information on Lagrange for Android in Geminispace seems to be in skyjake's glog posts. Still, I'm leaning towards a purge.

Admittedly I am a bit biased seeing as how many submissions are individual logs from my 'glog', however I do feel that blanket banning would not be good. Mostly for the reasons already listed, some individual logs are more important in informational value to where they deserve to be featured. One-off guides, recipes, and scripts are sometimes are found only as individual logs. So, I do believe that topically relevant logs should be allowed in the GDC.

I do feel that it does leave open the issue of people using the archive to spam and over-promote, as time goes on we will see if it becomes more of an issue. For now a blanket ban is not necessary as long as everyone plays nice and isn't a dick about it. If it does come though, it would be easy as pie to just fork a version of the directory before-ban so no biggie. The power of FOSS!

> Ill host my own directory, with blackjack and hookers!~

Even now, I refresh the 'recent additions' section multiple times a day just to see if anyone else has contributed something new. Its always fun to see a new URL pop in the feed. Also sometimes mbay will restructure everything and a old link I submitted will shoot back up to the top for one reason or another. I make it a game to spot every revision and difference :P

Ok well thats all I had to say on the matter. Hope the project continues to do well. If you come across something neat, useful, or cool looking, do your part to share it! Together we can end the discoverability crisis!

Back To Logs