Okay, here’s an idea: the main site will be a wiki, but text and git based.
Oh, let’s back up: we are going to build a wiki of paracosmic knowledge, to assist mages, from this world and others, combat the forces that would destroy knowledge, and therefore our connection to the All.
To most non-mages, it seems like another video game and movie database.
With that out of the way…
I really, really want to run a Semantic MediaWiki; it is great, all these neat things you can do, make neato queries and searches, all that jazz. But it comes with the cost of MediaWiki, which is increasingly not the tool for the job, unless that job is a Wikipedia project.
I could enumerate all the things that peeve me about MediaWiki (and I should probably make a list like that online, to reference), but here is the cart before the horse: I want to separate the “semantic operations” from the dataset.
I don’t want to duplicate Wikidata, or all the other databases out there with factual content in them. I want to use them as is, but also introduce our own data that we can’t store elsewhere, into our processes.
So when I want to make an infographic showing the family tree of MUDlibs, I want to build the structure from Wikidata, include contemporary active instances from Mudconnector, and include our feature/antifeatures from our own wiki.
There is a field of software that does this kinda stuff, one that comes to mind is Metabase. The appeal of pulling in multiple data sources is strong, and while we can do that with MediaWiki (or my much more comfortable WordPress), I feel that we would be building a solution to fit the quirks of those established systems, and I’d rather explore how to make data very portable, regardless of where it “lives”.
For implementation I am thinking Hugo or another static site generator to produce both our web-accessible content, but also to generate the raw data feeds for ingestion into other systems.
I’ve resisted such a system for sooooooo long, because it is not friendly. But the world changed a lot lately, and my opinion of approaching its users has changed in kind. A crude summary would be: folks need to learn to use computers in more ways than filling out forms online.
When we look at a system like MediaWiki versus one like a static site created via git, it would appear as if this very technical requirement would dry up pool of collaborators. But where am I coming from? I am not Wikipedia. It all likeliness, this art project will not be seen by many people, and the ones that do will know what git is.
But even more than that, being able to log in and edit a page directly isn’t the goal for every editable page on the web. There are a lot of folks that see an issue, but don’t want to fix it because it requires a lot of effort, or they are unsure of the exact fix, or they don’t really know the process for changing a thing, or they are impostors in complete control of their otherwise beautiful bodies and therefore do not believe they have the answer.
Of course we want feedback from all those folks, and MediaWiki even allows for anonymous visitors to edit the site. But we should all know for all practical purposes, anonymous editing is not really allowed to us. Again, we are not Wikipedia.
So let’s come at this a different way; let’s go out of our way to give folks to interact with us. Let’s put forms and links to contact methods on every page, so we can give people a way to interact with our data, which is the world’s data, in a way that works for them.
But that’s not all! If they stick around for a bit, tell them what else happens, maiki!
For the folks that interact with the project, they become part of it, and a driving principle of multiverse adventuring is self-education; our goal is to teach everyone, each others, strangers, that one friend you are sure could geek out hardcore if they found their thing. Everyone learns!
So I am feeling better about proceeding along this path. Also, this is just text, ya’ll. If we scare folks off with text, we are in bad shape.
Gotta run, thoughts?