Last week, I finished up my three weeks at Google CSSI, a summer crash course into web development. It was quite enjoyable; our third week was spent frantically putting together a web app in roughly four days. My group made Double Crux, a web forum based off trying to facilitate online use of the CFAR technique of the same name.
You can see the Double Crux site yourself here.
Here’s a few screenshots of the site:
We tried to do two cool things with the web app to help facilitate the Double Crux technique: split-screen and recursion.
One thing in discussions that can get annoying on places like Facebook and Reddit is needing to scroll way up to see what other people said—comments stack linearly. On the forum, we made the decision to divide the screen in half: your points would stack up on one side, and the other person’s points would go on the opposite side.
It sounded good, but this didn’t work out as well as I had expected. For one, our UI basically forced people to write out their cruxes, title and content, one at a time. But in practice, what seems to actually happen in Double Crux (at least for me) is that I generate the reasons for why I believe something (i.e. the crux titles), but the actual details (i.e. the crux content) come after delineating them.
So one thing that might work better in the future would be if we encouraged people to do a quick brain dump of the reasons they believed one side, and then prompted them to enter in details in a second step.
The idea behind recursion was that you’d be able to take cruxes and turn them into new full-blown split-screen discussions. While both the idea of starting new discussions from sub-cruxes and the Double Crux format were good on their own, our execution of putting the two together didn’t work out well.
The most obvious one is a lack of continuity between the parent discussions and its child discussions. Sure, you can click on Recurse to take you to a sub-discussion, but there’s no real signal that you’ve gone “deeper”.
Without any slick animations or the nesting format of Reddit, the connection between one discussion and the next is broken. It feels like you’re just browsing two disparate disputes rather than two related ones.
One way we could have made things better would have been incorporating some form of a minimap to track how the tree of points and discussions were developing.
There’s some additional readability issues that are still unaddressed. There’s a lot of empty space on the site that’s just not really serving much purpose. The current crux boxes still seem too big, meaning you can’t fit that many on the viewport before you need to scroll. I think that in the future we want to prioritize quantity over quality when it comes to how many cruxes we fit on-screen.
The screen issue seems fairly problematic to me. We want people to be able to easily add new points without having to scroll all the way down (like they do now), see most of the cruxes at a glance, and also keep track of what the central question (aka title) is even about.
And all of this isn’t counting things like a better way of handling profiles, tagging for discussions, search functionality, and other things in the way of general accessibility.
In conclusion, there’s a lot of work left to do. An improved attempt at making Double Crux a usable online tool would (aside from having well-intentioned people who know what Double Crux is about) need to:
- Allow cruxes to be easily entered and edited / elaborated at a later time.
- Visually represent recursive sub-discussions and allow for quick travel from node to node and branch to branch on the discussion tree.
- Optimize the UI to show more points at once, and figure out how to better allocate space on the screen.
Realistically, at its current stage, I don’t expect many people to actually use our site for Double Cruxing. However, there are some useful design lessons here, and I also think I learned more about what it’d take to make Double Crux a viable online option.
But at least it looks pretty.