We are in Norway for a week. It’s bliss with Edgar. It’ll be his first 17th of May today.
When re-reading about the PARC NoteCards and how they wanted it to be very general I thought how that there is no such thing as a general tool, even a knife is made in many different ways for different uses.
I also thought about all the interesting spatial hypertext type projects there have been and, of course, they are not for all users, but for the specialised user they are also not as popular as they should be. The key–and this is the key to my PhD thesis–is that they must be a view of other, core text/information, not a separate ‘thing’.
This then turned my previous post about how to include nodes/elements on its head. The view should not ask the user to include individual elements but ask what category to show, so it will be live, but also allow the user to create new elements in this view, primarily headings but also glossary terms, notes and drag in images and documents to cite.
I further thought about the KJ-Ho brainstorming and thinking method and how Author should support that so that we use Level 2 headings the Dynamic View by default and use Level 1 as Category headings. Hence the different interfaces for those two.
There could be the approach that on entering the dynamic view two side panels would appear but nothing in the middle, on the canvas. The left panel would be about layouts and the right panel about elements:
However, making it easy to have any and all types of nodes (headings, glossary terms etc.) visible at all times would be easily messy. If the user turned on headings first and then glossary terms, where would they go?
This is why I think we need to provide quick access to views of each type with either a mixed-view. I thought of cube-flips to make the sides mean something different though I was concerned about the feeling to the user at having something jump out of the screen but this looses the tangible aspect:
It seems at this point that supporting different views for citation analysis and brainstorming would make sense but citation analysis can benefit from having more than just the citation documents on screen–headings for groupings (or notes) and glossary terms to map concepts to seem integral.
One way to add items would be to shrink the canvas and allow the user to place the new category, such as adding headings here:
This kind of interaction seems needlessly heavy though.
I have also experimented with columns but not sure how to make this free-form:
The primary use cases for this are
Or so I thought (see below).
Because it’s so clearly split for this application, perhaps pinch in should provide a TOC view (as now) and make it freeform? If we can provide a TOC view with level one headings being categories when thought of as KJ-Ho, as shown below, AND a way to put free headings (just created, then that could be interesting. I just tested using this format for TOC instead of a single scroll and it works beautifully and does not feel odd at all.
In this view, non-attached Headings, created in the Dynamic View by double clicking, can be anywhere on the screen. The user can scroll left and right as required and maybe up and down but that could get messy. Any heading not attached would not show up in the word processing view.
The user can double click to type a level 2 heading but to make a level one/category heading we should test with maybe having a + to the right of all the categories or maybe a button to click when making a heading? Likely we’ll have a button or a graphic like this:
This leaves pinch out for non-heading items, such as Citations and Glossary Terms. The initial view here should likely be citations since the glossary terms are external entities. If there are no citations in the document the pinch our should do nothing, same as pinch in does nothing if no headings.
In this mock-up, which is expected to be used full screen mostly, we have only the timeline view showing only citations.
Judging by the timeline mock-up above, glossary terms are not always necessary or useful when doing citations analysis. Should the glossary terms be in their own view? Maybe, but they need to be linked and linkable and visual with the citation documents they are supported by.
Perhaps glossary terms should be accessed by clicking on them in the text, perhaps via cards? This way they can be accessed in a unique way, away from the other views but still benefit from the dynamic view.
For these reasons I think it’s prudent to improve the pinch in for TOC to be in columns and support adding headings in that view and Citation Analysis in the timeline view and leaving glossary terms to produce a view based on clicking on them.
I have a feeling the interactions could be more self-evident, if not intuitive, with some sort of use of the cube metaphor.
Maybe use the cube for different layouts of the same data?
Maybe make entry to timeline a different one, maybe on the back of the document, and leave glossary terms for the pinch out?
Note, this led to my next post.
This was the morning after Edgar was a climate warrior and got a haircut. #extinctionrebellion