The CardiganBay page has a proper overview of the reasoning behind this new wiki engine.
But I found myself writing a tweet-storm today highlighting what I think are the most important elements / ideas. So here it is, as a kind of shorter manifesto :
Cardigan Bay is clearly aimed at the #toolsforthought #digitalgarden space.
But it's not entirely like other tools in this area.
Why is that? A quick thread about the major design goals for Cardigan Bay, which explain why it is like it is \1
1) Cardigan Bay comes out of the Wiki tradition. And its primary use-case is to manage and bring order to ThoughtStorms : An old wiki of 5000+ existing markdown pages.
Which has a lot of outdated stuff, and is in sad need of cleaning up.
Some decisions are about backward compatibility with this existing data.
Others, eg. the "one large text box" philosophy stems from the need to restructure existing text, breaking 1 block into 2 or bringing 2 back into 1, which is harder in a more "structured" editor.
2) Another inspiration is JuPyter notebook, and also SmallTalk.
CB is intended to be a "smart notebook" which can embed live computation and explore simulations etc.
It's still very embryonic and a long way from where I want to get. But embedding code is already possible.
3) Cardigan Bay is intended to be usable on mobile screens.
While the UI is still very clunky / unfinished. And certainly not a slick native or mobile first UX. It is technically POSSIBLE to use CB fully on a tablet or phone.
I do it all the time.
This again explains some of the UI decisions.
I had an outliner based wiki back in 2014 https://github.com/interstar/OWL
The main reason I abandoned it is that while the outliner UI was great on a desktop with keyboard, it was awkward on a tablet.
In future, I will certainly go back and explore an editing UI that's more sophisticated, structured and intelligent. And may bring back more outliner type functionality.
But only once I've figured it out for mobile. Right now CB is equally meh-but-usable, on everything. :-)
Another thing. Both the JuPyter inspiration and the "usable on mobile" commitment, have pushed me towards thinking of the UI as a 1D sequence of cards or cells rather than wanting to explore 2D layouts.
CB is militantly 1-dimensional.
4) Finally, Cardigan Bay should be usable offline, on your own machine. And store all your data locally. So that you have full ownership, control, privacy etc.
To publish a wiki you export a flat site.
Now in future there might be more cloud / hosted options. I'm interested in a way of integrating with https://remotestorage.io
But these will always be optional and secondary.
CB should always be usable offline & locally.
So that's why Cardigan Bay is the way it is.
Obviously some of the flaws are just good old fashioned laziness and stupidity too.
And hopefully they'll get improved over time ...
But these 4 big design decisions / goals : being wiki, running code, being mobile friendly and under your control, I'm pretty committed to.
And they'll shape Cardigan Bay for the foreseeable future.
Even as I work on improving UX and interop.
And if this vision of #toolsforthought excites you, then, be aware there's a long way to go, but get involved : https://github.com/interstar/cardigan-bay
Everything is fully open source and intended to remain so forever.
There some more stuff on AgainstDirectManipulation explaining what I don't like about it, which also feeds into the design of Cardigan Bay.