Persona 1:

This user uses their computer mostly casually and recreationally, and sometimes productively. They are only ever working on one or two projects at a time, and as such they already have a pretty good mental model of how different items relate to each other.

The human mind is actually pretty good at ascribing meaning to things, and when presented with an array of options the power of recognition is pretty apt at filtering relevant items out from the irrelevant. This persona doesn’t need the computer to group things for them and they are probably not going to spend time inside a management interface mapping their already existing mental model into the computer. What will benefit this persona is a computer that is able to tap into this mental model, allowing the user to access items immediately, eliminating the mundane, tedious actions of going through the file browser, or web history etc. It doesn’t actually matter if the groups created by the computer have fuzzy boundaries or are otherwise innaccurate, because this persona will still be able to recoginse the important.

Also for this persona it is not so important for groups of relations to persist. What I mean is that while individual relationships must exist for the technology to work, the concept of a ‘working set’ doesn’t need to be exposed to the user. There is no real need to be able to manipulate sets outside of the currently active one. Instead what is important is to be able to recreate the set (as oppose to retrieve from a data store) when the user needs it. You can determine when this will be because the user will access one of the items (from GAJ for example) and you can recreate the set around that file on demand.

Persona 2:

Recognising the fact the the human mind can only deal with a small number of things at any one time, this persona describes people who have to deal with a large number of things. For example a person who works with many clients and each client involves a number of subprojects would fit into this persona.

Here, due to information overload, the person is unable to rely on themselves to remember everything they need to know; they need to offload some of the burden onto the computer. This persona would benefit from a management interface with persisting data. They will allocate time solely spent on organising items into related sets. This is not so much of a burden as it is for Persona 1 because we know they do this anyway with to-do lists, journals, diaries, PIM tools and a whole range of ad-hoc solutions. Here something like ReflAction Journal is definitely useful.

Sorry if this is already obvious to you, I’m really just getting my thoughts down.

I also don’t think that these two personas are mutually exclusive. It certainly couldn’t hurt to include something like ReflAction with Persona 1 – after all they don’t have to use it. And I’m sure Persona 2 would benefit from the focus concept just as much as Persona 1. I just don’t think there’s a Zeitgeist be-all-and-end-all solution for everyone.

Also regarding placing the most relevant items on the bottom, I did this because:
1. It is expected that this list will have fuzzy boundaries, kind of like how Google search returns millions of results but only the first couple are useful. If the list went top-to-bottom, you would have to define some arbitrary cutoff point for the last item, whereas if it goes bottom-to-top the list can extend upwards indefinitely (ignoring screen space limitations).
2. The purpose of the menu is to act as a ‘quick-launch’. The user is most likely to click on the item that is most relevant, therefore it makes sense to position this item closest to the button. Otherwise they would have to click the button, then move the mouse all the way to the top to click the most relevant item. This erodes the convience of the menu, when the entire purpose of the menu is convienience!

Looking forward to future posts.

Here is a video of the ReflAction Journal (a UI of my PhD-work system ContextDrive, which also influenced and is influencing Zeitgeist and GAJ) prepared for our workshop presentation at CHI 2010: http://reflaction.info/wp-content/uploads/ReflActionJournal_CHI2010.mov. I’ll blog about it later this week. The video demonstrates appropriating calendar items as working-set/activity labels in the way, I had earlier described also on my blog: “The experience-trace implementations ContextDrive and Zeitgeist can not only capture when a calendar item has been created or modified, but, more importantly, what time span it refers to. Thereby, calendar items are turned into labels and container items for what actually happens during the time span in question. Plus, calendar items can be appropriated, for example, to serve users in relating their preparatory and follow-up activities to the represented “events”. The ReflAction Journal facilitates this via drag’n’drop.” Your focus button relates to the vertical “now” line in the video, but so far, we didn’t use it in the ways you are suggesting. But I think, I’ll change that.

Maybe, you’ll find it interesting to read through the section “Journal User Interfaces”, starting at page 4 of this paper: http://reflaction.info/wp-content/uploads/PersonalExperienceTrace_prante-etal.pdf, from where you can see that there are not only calendar-item labels in the context pane within the below part of the ReflAction Journal, but also task/to-do labels. And yes, we are working towards integrating this kind of stuff in Zeitgeist, the Activity Journal, and other Zeitgeist-based UIs.

I’ve put together a little mockup that builds upon the task pooper concept:
http://www.users.on.net/~cmoschou/images/augmented_task_pooper.png

It’s fairly simplistic, but I suppose it needs to be, otherwise it just becomes another management tool that people will never use. What do you think?