Grants:IEG/Alt text tools/Midpoint

Welcome to this project's midpoint report! This report shares progress and learnings from the Individual Engagement Grantee's first 3 months.

Summary edit

We've spent time researching alt-text and tagging best practices
Invested considerable amounts of time learning new technologies to make our tool better
Have reached out to several physical spaces and are working towards setting up 3 Editathons in the NYC area
Reworking our tool for stability and usability

Assignment of roles and responsibilities
There was a bit of a late start as the lead programmer wasn't available for the project till September 1st
Time has been spent refactoring existing codebase
The dump parser which find images missing alt text hasn't run since the prototype was built in February is working again.
We are in the process of getting the thumbnail fetcher back up and running. It's a program to download millions of images and validating WMF thumbnail pipeline over a dozen servers. The downloaded image will be saved for metadata extraction and computer vision analysis.
We are looking into computer vision. While machine generated alt text has become available since the grant's writing, we consider them ill-suited for the encyclopedia similar to machine language translation. However, as pointed out in the grant object identification (cat, dog, building, etc.) would allow for users to select an area to specialize in.
About two days were spent building a colorblindness simulator in JavaScript to learn Canvas and color spaces.

As of today, we've spent time working on the existing prototype, to improve functionality, performance, and overall accessibility
Members of the community, here in NYC, have been contacted for support in setting up editathons
Research for the documentation has been done, and its production will be happening in the 2nd half of the grant cycle

As of right now, we haven't spent any of the grant's finances. We're currently planning our meetups, and working earnestly on documentation and the tool itself, so we anticipate that we will disburse funds within the next few months.
We don't foresee there being any substantial differences between planned fund usage and actual fund usage. I suspect that we may spend more time doing testing and error remediation that initial thought, but that's true with most software development.
Our fund spending has gone as planned. We figured that no funds would be spent in the first half of the grant cycle as most of the time has been spent planning and coordinating and not as much time "doing".

Time. It's very difficult to coordinate schedules, and set aside time in our individual lives to work on this project as a team
As 1/2 of our team is not a seasoned Wikipedian, the "cultural differences" in the Wikipedia community are somewhat difficult to understand. The established wiki-conventions (think the editor itself) and the utilization of "talk" pages and such, is difficult to wrangle for a newcomer to this environment. Naturally, as more time is spent here, the familiarity increases.

Wikipedia Community Involvement - We've found the Wikipedia community incredibly supportive concerning our efforts to help make Wikipedia more accessible to individuals with disabilities.
Expert Involvement

External Community Involvement - By leveraging relationships with the disability community at large, we've been able to increase overall project interest.
Organizing an Event Series

Texting - because sometimes emailing or leaving a message on a talkpage to your fellow Wikipedians is too slow.

Wikimedia Foundation staff has been incredibly helpful throughout the initial grant process
Wikipedia community has been welcoming and amazing to work with thus far
It's awesome to be a part of a project that will make one of the largest, most trafficked websites in the world a more accessible and inclusive place