WMDE UX/Newsletter

Issue 10, 16.04.2018

Previously in... the UX office (and elsewhere)

It seems that we roll with the quarterly check-ins rather than the monthly news. So here is a nice and heavily stuffed newsletter from your UX team. This time we are happy to introduce and recap the making-of of quite some new features: Check out our involvement in the resourceful Advanced Search, the heavily discussed Client Editing, the long awaited File Importer and the specifically researched Rollback Confirmation.

Besides our feature-involvement we joined yet another on the road activity: the WikiIndaba 2018 taking place in Tunis and bringing together the African Communities and User groups. To get a better understanding of the users we gathered our magnifier glass and detective hat and conducted research in place.

A big bullet point on our operational goals is striving for a learning organisation. This is why this issue is featuring two topics on this: in a very practical approach we are giving insights on how templates help to spread knowledge evenly and easily accessible. In another note, we look into how to sort things out with the ANT. 🐜

Besides all these interesting matters there is a last and definitely not least announcement I cannot be happy enough about to repeat: We have two new positions in our team while keeping the familiar faces! Charlie has been a working student with us for more than two years and is now officially employed as a UX designer. Jan being a UX designer and researcher so far will from now on take over more management tasks on top of that and be the team lead of the UX team. Whooop! 🎉🦑🎉

Templates, Templates

In the last months, I worked quite a bit with templates: Lists or Boxes to fill out to define features or a data need.

I think this is something that is helpful for us. For me, many things talked about were unclear and only some people seemed to know what a feature was supposed to do and particularly, what it should do for the users. So it helps to communicate this. It also helps to communicate only the essential things. We are a big organization but we sometimes have many problems of interfacing with other teams and disciplines because things “are complicated”. That is why the feature templates are one page and not more.

An alternative is the oral history of decisions. Looking at the stories we tell each other and the complexities we convey in them it is also a method we rely a lot on. As main mode of communication and decision making, it is tool I find great for smaller teams with relatively few connections to others. However, in a bigger organization, such written and standardized communication devices are important. Without doubt, they are still insufficient, like any other tool, to convey all of it™. Every tool has it’s hinterlands, so have templates. ~Jan

Advanced Search

Some of you may have seen it already, some of you may be using it themselves, the advanced search feature! Deployed on de-/ar-/he-/media-wiki.

What’s new is that you can now filter for languages, categories and subpages! (Not all of these are deployed yet, but they’re in the making). There are still a few UI bugs to be fixed but all in all we can slowly see “the end” nearing which is really exciting \o/ because we’re pretty much in time! ~ Charlie

Rollback Conformation goes AdminCon

This wish has finally been picked up again by us after being dormant for quite some time. After Conducting 4 in-depth user interviews + the research that was done at the WikiCon and Wikimania last year, we started coming up with solutions to the problem.

To give a quick recap: In deWP users with active sighter rights can rollback all latest edits of a specific user without a comment. This is called “kommentarlos zurücksetzen”. It is different from a revert, which can only ever revert one edit and asks for an edit summary before saving. Rollbacks are only supposed to be used to fight vandalism, but vandalism hunters are few. This means every active sighter sees the rollback link behind every latest edit, although they never or very rarely use it.

The wish is on position 7 of the 2017 wishlist survey and states that users will often misclick when trying to thank someone for their edit and instead accidentally rollbacking it without leaving a comment. Understandably they would like to avoid this uncomfortable situation and so they asked for a confirmation prompt, similar to the one you get when thanking someone.

To make sure we meet the needs of active sighters that don’t need the link and vandalism hunters who need it to work efficiently we decided to take our solution ideas and questions to the adminCon this year. This will be the first time that we try an experiment, where we do a session remotely with a session-partner from the community who will be present there to help us with the presentation and the discussion with the audience.

Here is a short list of our main findings: The community members present liked the format The person documenting the questions and feedback needs to be on site This format only works for non-complex topics since communication is very hard when not seeing the faces of the other speakers and they are on a different medium then oneself We were able to gather more insight and get some feedback successfully ~Charlie

Client Editing

“Client editing? That rings a bell. What was that again?” - I’m glad you asked! We want to make data from Wikidata more accessible from Wikipedia so it can be used in the infoboxes which already contain primarily structured data. Many communities already make their own templates and scripts so they can make use of the data in their articles.

The idea is to make it easier, so even someone who doesn’t know how to write templates or use Lua can use the data and most importantly, edit it directly from the local Wiki without having to go to Wikidata to change the data. *Mind blown*

We’ve just finished another feedback loop on the prototype, this time in the german-speaking community so unfortunately all the feedback is in German if you were interested to read it.

We got mostly positive feedback on the actual discussion page and very mixed feelings on the discussion page of the Kurier where I had asked for participation in the feedback loop in the first place. The discussion there is a really interesting read IMO so if you have 45min over lunch or on the train from work, give it a read. I find it interesting because it gives a good insight into the german-speaking community’s mind and their opinion of Wikidata. Obviously this is just a excerpt of the whole range. My finding was that there are a few people that scream loud and are fearful of Wikidata but it seems like way more people did not agree with their (often unfounded) reasoning and actually argued against them in a very objective way. I was very impressed and happy!

Next steps is a meeting to clear up all technical constraints and get a shared understanding of what the MVP is. Coming soon… ~Charlie

File importer

The File Importer Extension is something that was wished for in the Technical Wishes survey quite some time ago (2013 if you wanted to know in detail). Now we are happy to say that the mock ups for the MVP (minimum viable product) are fully delivered and currently in the making. Keep your eyes peeled when importing locally stored files from certain wikis to Commons will be a no brainer. ~Hanna

Research at the WikiIndaba 2018

When designing new features, pages or workflows it is the task of a UX designer to keep the different user groups therefore different user needs and several possibilities how to look at the newly introduced part of the wiki world in mind. So when I heard about the WikiIndaba first, I did not know what to expect at all, which was the first spark to give it a closer look. As especially Wikidata has a universal approach I set myself the goal to figure out how far the knowledge about Wikidata has spread, what expectations there are and why people might be interested in Wikidata. To give you a better idea of my findings I put them together and created two fictive people. Doing things like that is called creating personas. It should help to sympathize with a specific user group and therefore build for the user needs in the first place.

Meet Rana

Rana is a young women from Egypt contributing to the Movement for about 5 years now. Besides editing she mainly does workshops on how to edit on Wikipedia. She does so in a university context mainly. In her workshops Rana teaches editing in the wikitext layout on the desktop version. Some of the participants cannot bring their laptop which means they use their phones. To follow the workshops they switch to the desktop version to be able to follow along in the wikitext layout. It’s a great show of dedication going through this experience.

Talking about dedication: Rana also told me about her experience with Wikidata. She was really happy to have attended the Wikidata workshop as she finally understood of the capabilities Wikidata has to offer. She wrote her first query (search request) which will save her a lot of time in the future. As she is doing editathons to fill the gender gap she can now easily check Wikidata for missing articles in her Wikipedia. So why did I think she is dedicated? She mentioned aside that she already has over 600 edits on Wikidata, which is remarkable considering the fact that her workflow so far is: clicking in “Random Item” → seeing if there is something to translate to arabic → translate or click on the next random item. Thank you Rana!

Meet Lionel

Lionel is a well known member to the Wikidata Community and the Communications Team. His quest is to fill the gap of African cultural heritage in Wikidata and in return have Wikidata related sister projects benefit from that. Lionel is an active member of the Cameroon User group which has a strong bond of volunteer developers. Together with his fellows he organizes editathons and besides that develops for Mediawiki himself. Right now we investigate of setting up a stronger collaboration with Lionel and the Cameroon Community to establish potential pathways for our take on Participatory Design. ~Hanna

ANTs sorting things out

In the last few months I have been reading a lot on Actor Network Theory. This is a method that is infamous to tell that things have agency. Oh noes, postmodernism strikes. However, if the computer calculates and the pot boils water, you are already on the team (things still don’t have intent!) Anway, an important part of ANT is how we use “things” to structure the world. Famously developed in anthropological studies analysing how scientists transfer an experiment to measurements to tables to diagrams to comparison of diagrams to findings to scientific truths aka “facts™”.

This relates to some interesting work on what categories do (or what we do with them). Since Wikidata has a lot of instanceOf I find it very interesting to ask about the history of other instanceof systems.

An interesting instance of such work and its consequences is the ICD, a classification system for medical diagnoses: It defines illnesses and disabilities. It also un-defines them: Asperger’s, for example, was recently removed from the DSM (a close relative to the ICD). According to it, people with the same symptoms now fall into another classification – which is insofar interesting (and possibly devastating for people in that (former) category) since there is an active community around Asperger’s.

As far as I understood Leigh Star’s ideas, categories are a necessary part of life and particularly of structuring our modern world. However, each categorization necessarily hides some things and shows others – with consequences in the lives of people and thus have a moral part them which we can not escape. ~Jan