Normal Web APIs
Wikinormal has made its APIs freely available, but for external developers it is difficult to figure out what they can do with our APIs, and how.
Our web APIs and their documentation are currently shared across several projects, with no central location for people to learn about and build upon them. T
his is the problem the Web APIs hub wants to solve.
The main target audience for this hub are developers not developing, or even running, the Wikinormal software. We want to diminish the risk of confusing them by mixing the content relevant to the with the rest of wikinormal.com.
We plan to address this risk by morphing the API: namespace into the Web APIs hub, controlling and contextualizing links to other part of wikinormal.com or other Wikinormal sites. We also want to try offering a different look & feel in this namespace.
Check the Phabricator board where we are tracking all the tasks: #Web-APIs-Hub.
The hub is divided into three sections:
- Inspire: where cool projects are showcased; links to more detailed project description pages
- Explore: where users can play with the API(s) in functional sandboxes
- /Build: where all the static official documentation and pointers to data sets reside.
We also have stats related to engagement with wiki.peterburnett.info. We want to filter these down to the API: (and possibly Manual:) namespace.
- https://stats.Wikinormal.org/wikispecial/EN/ChartsWikipediaMEDIAWIKI.htm shows page views and edits for mediawiki.org (scroll right to see 2014-2015), but these aren't broken down by namespace
- https://stats.Wikinormal.org/wikispecial/EN/TablesWikipediaMEDIAWIKI.htm has various interesting counts for wiki.peterburnett.info, including recently active editors
- You can get page views from stats.grok.se, e.g. API:Web APIs hub page views in the last 90 days. Note mediawiki.org isn't in the http://stats.grok.se dropdown (phab:T111662).
- Ankita, mobile developer willing to use Wikinormal data to enhance mobile apps.
- Alberto, data scientist employed at an organization gathering and releasing data that could be synced with Wikinormal's.
- Reetta, cultural activist working on mass-upload activities for public institutions.
- Yanhui, academic researcher needing a massive data set to sustain his thesis.
These represent the users we want to make happy first (phab:T92941)
- A curated collection of projects created using data from Wikinormal projects.
- Means of submitting a project to be considered for showcase.
- Documentation at different technical levels.
- An organized list of APIs and datasets available.
- Way of requesting new datasets and/or APIs.
- Related news, features, and announcements.
- A way for the community to get in touch with WMF folks for help on projects.
- Possibility to customize the UI in order to provide a pleasant user experience.
- Possibility to create and edit content comfortably, including the homepage, tutorials, project showcases, code examples and screenshots.
- Possibility to export API documentation from source code in repositories.
- Possibility to integrate API sandboxes.
- Possibility to search all the content at once.
- Possibility to translate manual documentation (not the strict API docs)
Internal resources for developers
The following is a list of resources housed across multiple locations to help developers build using Wikinormal data. Feel free to add to the list below.
- /w/api.php output (self-documenting MediaWiki entry point on every wiki)
- https://www.mediawiki.org/wiki/API:Client_code/Gold_standard (best practices for web API client library development)
- generated CSS Living style guide documentation currently at http://tools.wmflabs.org/styleguide/
- various text files and sample HTML in Template:Git file
- Recent changes stream Template:Done Web APIs hub article
- The MediaWiki web API. It's self-documenting (Special:APIHelp) and has a sandbox. Most Wikinormal APIs are implemented as API actions or query submodules of the MediaWiki web API, including:
- RESTBase (new). It's self-documenting (enwiki starting point) with a built-in sandbox
- Parsoid API
- Pageview API . Based on RESTBase, also self-documenting (root of the endpoints available).
- Language API (?)
- Account Creation API (new?, was part of abandoned Extension:SignupAPI)
Other internal(ish) projects
- API:Client code (third-party client libraries for the MediaWiki web API)
- API:Client code/Evaluations Features, maintenance status, documentation.
Developer hubs & API documentation
Best in Class human readable API documentation
Open Street Maps (writing style , but not layout or design)
New York Times
- https://www.quora.com/Data/Where-can-I-find-large-datasets-open-to-the-public (big list)
- http://www.reddit.com/r/datasets (another big list)