In the past few years we've worked really hard to bring UploadWizard to a usable state, and now that we have a Multimedia team framework to actually do some good work on it, and we have the impetus to start making the upload pipeline way better, let's do it the right way: Let's make this project the organized, well-engineered machine it should have been in the first place.
I would particularly emphasize code conventions being followed, jshint conformance, an eye towards an MVC-like structure (nothing strict, but as a general goal...), building toward libraries that can be used elsewhere, and a variety of different possible interfaces.
Ideally none of the upload steps will be entangled unnecessarily with other steps, and we will be able to untangle them when necessary and put them in different pages or tools.
Ideally we will use the brilliant work- and complexity-reducing technology (oojs, oojs-ui, jQuery promises, moment.js) that other projects have started using in recent times.
It is an absolute *must* that we build tests for this system as we are building it, if we want this to be a generally useful set of tools for other projects. Same for documentation.
The leaders in LOC for active frontend projects deployed on the cluster are* VisualEditor, UploadWizard, Flow, and MultimediaViewer (in no particular order). Let's start bringing the upload pipeline projects closer to the high standards that other projects are holding themselves to.
- According to an unscientific survey based on guesswork
See https://www.mediawiki.org/wiki/Requests_for_comment/More_general_frontend_uploading_tools for more.
Version: unspecified
Severity: normal
See Also:
https://bugzilla.wikimedia.org/show_bug.cgi?id=62513