Grants:IEG/Wiki needs pictures/Midpoint
Welcome to this project's midpoint report! This report shares progress and learnings from the Individual Engagement Grantee's first 3 months.
Studying some critical issues, brainstorming how to build the backend architecture, asking to people how they expect a tool like this should be, lead us here with clearer ideas about what the actual wiki-environment offers as data and about how we want to implement our vision. This midpoint is the starting point for resolve all the task that we have collected so far, writing code.
Methods and activities
- The project starts as a confuse idea rapidly scratched in some demo code.
- It follows a brainstorming part when we considered all the possible ideas correlated and develop a more structured schema.
- Then we "played" with some critical problems (e.g coordinates or categories) to stress the environment and understand which would be the best solution.
- After a monthly skype call in February with Marti, the project started to focus on a more concrete target. A GitHub repository came up, with its issues section. Here is where we will work on the second part of the grant.
- Some spot presentation of the beta tool during other wiki events in Italy.
- A more structured campaign with dozens of it>4 users on itwiki, wikidata, commons and itwikivoyage. Contacted users were selected because they showed a very strong geographical "center of Interest". An explanation about the P18 (image) and P373 (commonscat) properties of wikidata, was followed by a presentation of specific tools related to "image maintenance". Results show that almost 33% of them showed some sort of interest and activity.
- So far we have explored with great details the idea proposed at the start of this grant. So we have produced some test to understand the libraries involved in our software architecture (you can find our conclusions here).
- After having understood that our idea was quite feasible as is, we started to spread it to our nearest community, the italian one. As you can see in the image above we've tried to touch any region of Italy, interviewing both interested users (this group in a systematic way), both casual users. From them we receive feedback about usability and about their interest in a tool like that.
- Finally we end up with a github repository with the code of the beta tool, in order to receive punctual feedback during the second part of the grant.
Basically our finances are focused on manpower, they are compensation for the hours we spent in front of the screen. We think that so far we are spent our time as planned.
If possible, maybe in the row "Analyze how each community face the problem and survey of all other existing image tools to better interface with them" the time spent will be in the end higher then expected but only because the analysis has expanded its target. Not only other wikis were analyzed but additional time was spent to interview and contact newbies about our tool and other existing ones. Alexmar983 has spent an additional 4h a month on this aspect, and he is preparing some wikimetrics when the tool will be finished to target potentially involved users.
What are the challenges
- The quality of data: focusing on coordinates we have found misalignment (between Wikipedia versions and Wikidata), duplication (two or more pairs of coords for the same entity: which is the right pair?), poor quality (the precision isn't always the same) and simply wrong data (coords point to a different place).
- The environment is evolving: specially in Wikidata; this is a good trend but is difficult to build upon this stable services.
- Some API requests can't be solved in a short time (e.g Extension:GeoData).
What is working well
- Grants:Learning patterns/Git repository for software this is our way to produce free software and get feedback
- Grants:Learning patterns/Understand local volunteers' characteristics and behavior to develop an effective program although with a more one-by-one approach
Next steps and opportunities
- Develop the code following the studies done in the first part of the grant.
- Testing the application with different amount of data and different devices.
- Select an event or group of people (wikipedia platforms, chapter events) for testing in real the final version of the application.
AlessioMela: Being an IEGrantee as a tool developer is a good opportunity for having a number of feedback over the average. I was surprised by the effort that the WMF staff involves to follow the grant path. I've also understood that this process is a way to have more interactions with communities.
Alexmar983: I am positively surprised about the flexibility of the WMF environment. Sometimes I have to impression some passages are presented a little bit too rigid (deadlines, grids), but in the end it is not the case. Also, if your are committed to a grant, it is amazing how many spin-off ideas can come out of it. The "learning pattern" area of met is interesting, it is still too vague and I plan to update some those page with my experience. It should be promoted more.