Anatomy of a report

[barbaracassani]

The purpose of an application might be planned by its author, but at the end, I believe, the users will run away with it (ideally) and it will be the "feel" of the application, its UI, how it positions itself relatively to the user (i.e. virtuous and with a high entry barrier like wikipedia, or playful and accessible like Foursquare) than will determine its diffusion and the way it's lived day by day.

I think it might add to the discussion to explore some of the choices that Yip will unavoidably have to make in order to achieve a brick and mortar (in the internet sense!) existence. There are many but for the moment being I'd like to focus on the anatomy of a report. It might be useful for everyone involved to get an idea of what it will all be about.

Prescriptive vs freeform:

Which elements will form a report?

• a title
• a text
• tags?
• will the content be editable? by the author?
• reputation of the first reporter? but what if a report is edited by someone who is not an author?
• location - either via geolocation or added manually. Will it be possible to add in some way details about the scope of the reported news (i.e.: local area, city, region, world maybe...?) or will it be entirely determined by the geographical spread of all the reports pertinent to the same event?
• multimedia content: pictures, videos, sounds...
• time: start time, end time maybe?
• external links
• internal links? I think it's worth exploring a way to connect report to the same event in a visual way.
• will it be possible to "re-yip" an existing report to increase its importance, in a twitter / facebook fashion?
• will it be possible to upvote or downvote a report?
• will there be fields that allow the user to rate: importance, sentiment (positive / negative), veracity, usefulness...
• is it worthy to create an "aggregation" feature where content can be, simply, rearranged, like dragging and dropping placeholders, creating "event groups" a bit like the circles in Google+?

I am posting this laundry list only to spark a debate. I am willing to contribute my own opinions to the discussion - I favour immediateness over completeness, broadly speaking. But others might disagree and even if they agree, the way in which such principle applies to the various choices I've listed it's certainly widely arguable.

[padolsey]

I plan on adding a draft to this repo (probably in content/ux/report.md) where we can work towards a solid wireframe for a typical report.

---

(all imho)

reputation of the first reporter? but what if a report is edited by someone who is not an author?

The original reporter will/should be highlighted in some way. When the list of collaborators starts to grow I imagine the significance lended to any single author to diminish -- i.e. to not be such a large part of the UI.

location - either via geolocation or added manually. Will it be possible to add in some way details about the scope of the reported news (i.e.: local area, city, region, world maybe...?) or will it be entirely determined by the geographical spread of all the reports pertinent to the same event?

I imagine any information about the diffusion of the report itself will be in a separate interface/section of the report. The actual location of the event should be primary.

time: start time, end time maybe?

Yes, some events will have a planned end-time. Others will be open-ended.

internal links? I think it's worth exploring a way to connect report to the same event in a visual way.

Do you mean for sub-events? I've played with the idea of having an event within an event. This is mentioned briefly in the Glastonbury case study [not yet finishied].

So, as part of a developing emergent event, there may be something else that happens as a result of or within the scope of that initial event. For this we could take advantage of three different types of relationships that events can share:

• Causal -- X was caused by Y (e.g. a fire causing a riot)
• Integral -- X happened within Y (e.g. a single performance in Glastonbury)
• Topical -- X is topically related to Y

I imagine some events may be more complex than what is accommodated by these relationships though. We'll need to keep developing the case studies to expose such things...

is it worthy to create an "aggregation" feature where content can be, simply, rearranged, like dragging and dropping placeholders, creating "event groups" a bit like the circles in Google+?

Honestly I'm not sold on such UI sugar. I think there will need to be a way to define relationships between events though.

[barbaracassani]

Some reflections in no particular order:

• From my laundry list above, there might be a blatant omission: categories. In truth, it's something that we might or might not want. It seems to me that, paradoxically, they are slightly against the ethos of Yip: producing the seismogram of a determined moment in time (and its immediate neighbours), with all of its opportunities and obstacles. Adding categories sounds nearly dirty - like we were reducing the platform at a mere "yet another listings website" like Upcoming back from the grave. At the same time we can't forget we aim to fulfill practical needs and the spectrum of possible events would be ill-defined by tags alone. I.e. if I am looking for something to do with a date, that would encompass too many possible tags (fun, entertainment, cultural, bars, concerts, art exhibitions, etc) for those to be useful. So, as cringe as it might seem, it's a category "What's on in town" I would be looking for. Inasmuch as I am fascinated by seeing the territory as a unit, and a closure of the Northern line, a Nobi new opening and a fire in a suburban street all intertwined onto the same fabric.
• I am thrilled by the schema of event relationships you suggest. I really think we are onto something there: how various events or reports interlock it's what can make Yip have a truly documental value and not only a practical one. Upon further reflection, I also agree on the doubts about what you call "UI sugar". People like playful UIs but our concept of aggregation / relationship (exactly like the importance / gravitas discussed elsewhere) must be kept, for immediacy's sake, as much as under the hood as possible.

[barbaracassani]

So, the question(s) that Yip will be answering is "What's happening?", "What happened?" but the most basic atom of "happening" - the single report, sub-event, event - can be read as incidental to his surroundings, on the various different plans:

• time: what happened before / after
• space: what happens in the proximity
• mental space: what happens that make me feel / think similarly

I think that the "relationships" you describe might be a very important focus in guiding the architecture and the UI - in fact, this is the very core of the project imho.

[barbaracassani]

Adding to the laundry list:

• The social aspect. Will it be possible to send a report to one or more user (as an independent concept from the instances of the software "in a box" for closed circles of users). Would this make Yip uncomfortably close to any other recommendation engine?
• The possibility of tagging other people (celebrities? other users... ?)
• ... which bring me to yet another distinction: journalism vs journaling. We are basically looking to cater for both. It might be interesting to attempt cataloguing differences and common points.

[padolsey]

RE: Tagging / Classification / Categories

... too many possible tags (fun, entertainment, cultural, bars, concerts, art exhibitions, etc) for those to be useful. So, as cringe as it might seem, it's a category "What's on in town" I would be looking for. Inasmuch as I am fascinated by seeing the territory as a unit, and a closure of the Northern line, a Nobi new opening and a fire in a suburban street all intertwined onto the same fabric.

It's an interesting point you raise. I think it needs to be solved, but I don't think categories would serve it well. We can have other classification means though. Maybe user-curated lists (like you have on Twitter)?

Tagging can be very powerful too. Enough tags and the search mechanism will be very powerful. I could do a search like:

Sentiment: VeryPositive
Popularity: Local
TextSearch: Entertainment AND Food

The TextSearch would look through tags (weighted highly), description text, title text (also weighted highly) etc.

I imagine if the events are well titled, well described, and have ample tags, then we won't need to worry. IMHO straight-up exclusive categories seem very limiting.