The apps should use tags instead of (untranslated) hashtags in edit summaries.
Description
Status | Subtype | Assigned | Task | ||
---|---|---|---|---|---|
Open | BUG REPORT | Seddon | T358960 An unclear tag appears in the edit description when using the application | ||
Open | SNowick_WMF | T360164 Android and iOS apps should use tags instead of hashtags |
Event Timeline
We will be moving all event_comment values that were sent with app edits to Special:Tags, will track new tags in this ticket once we hammer out a consolidated set of appropriate tags.
Cross-referencing: there was recently a short discussion about these edit summaries on English Wikipedia too, with a couple of responses from WMF folks: https://en.wikipedia.org/wiki/Wikipedia:Village_pump_(miscellaneous)/Archive_77#Why_is_"%23article-section-source-editor"_appearing_in_edit_summaries%3F
Following are new values for Android and iOS app edits revision tags (Special:Tags).
Next step tickets (to be created by me and @Seddon) will be linked here:
- Task for adding tags to mobile apps extension. Tags will be registered everywhere (all wikis/langs)
- Task for replacing/migrating within the iOS and Android app
- Data validation/QA
Edit tags | Edit Type | Current event_comment | Platform |
app-suggestededit-description-add | Article description add | #suggestededit-add-desc | Android |
app-description-add | Article description add | Android | |
app-suggestededit-description-change | Article description change | #suggestededit-change-desc | Android |
app-suggestededit-description-translate | Article description translate | #suggestededit-translate-desc | Android |
app-diff-rollback | Diff Rollback | #diff-rollback | Android |
app-diff-undo | Diff Undo | #diff-undo | Android |
app-suggestededit-image-caption-add | Image caption add | #suggestededit-add-caption | Android |
app-suggestededit-image-caption-translate | Image caption translate | #suggestededit-translate-caption | Android |
app-suggestededit-image-add-top | Image Recommendation add* | #suggestededit-add-image-top OR | Android |
app-suggestededit-image-add-infobox | Image Recommendation add* | #suggestededit-add-image-infobox (both are used) | Android |
app-suggestededit-image-tag-add | Image tag | #suggestededit-add-tag | Android |
app-suggestededit-rollback | Patrol Task rollback | #suggestededit-patrol-rollback | Android |
app-suggestededit-undo | Patrol Task Undo/Revert | #suggestededit-patrol-undo | Android |
app-section-source | article section edit | #article-section-source-editor | iOS |
app-full-source | article full source edit | #article-full-source-editor | iOS |
app-select-source | Article select edit entry point | #article-select-source-editor | iOS |
app-talk-source | talk page full source edit | #talk-full-source-editor | iOS |
app-description-add | article description add (EN and Wikidata) | #article-add-desc | iOS |
app-description-change | article description change (EN and Wikidata) | #article-change-desc | iOS |
app-diff-rollback | rollback | #diff-rollback | iOS |
app-diff-undo | undo | #diff-undo | iOS |
app-talk-reply | talk page new reply | #talk-reply | iOS |
app-talk-topic | talk page new topic | #talk-topic | iOS |
app-image-add | Image Recommendation add | iOS |
In the idea of unifying user experience, when implementing the tags, you should reuse the format Growth uses, with similar pages to point users to.
Hey @Trizek-WMF do you mean the text format having spaces instead of dashes? We are using dashes like the Editing Team (see Edit Check). We can also point folks to a page.
I like the idea that all the proposed tags start with a common prefix, making it easier for filter writers to treat them as a group. I'm not that familiar with tag usage, but I'm wondering if there's some written standard for how this should be done. And if not, then maybe this is a good time to formally establish one so all future projects like this can just do it the same way instead of inventing their own way.
I see two broad categories of tags emerging; those that represent user behaviors (mw-replace, mw-rollback, etc), and those that represent essentially internal markers for UX experiments, such as the ones proposed here. A little bit of planning now to nail down syntax, semantics, and namespaces will probably yield good dividends later.
I maintain SPI tools (https://toolhub.wikimedia.org/tools/toolforge-spi-tools). One of the features I have is showing a matrix of which sockpuppet suspects used which tags. It's a bit of a mess because there's so many tags and most of them aren't relevant for my purpose. If there was a way I could easily filter out whole sets of tags to ignore, that would be useful. Having some sort of standardized convention for how tag names are constructed would help with that.
No, it's not a matter of dashes or spaces; just using tags with a link that clearly goes to a proper explanation page for experienced users. GRowth does this for suggested images: https://www.mediawiki.org/wiki/Help:Growth/Tools/Newcomer_Tasks#s-image
In a certain way, the standardization work started.
For Suggested images, Growth highlights the audience with Newcomer task and then the task type with Suggested: add images. We did it It allows patrollers to have several options, as they can either select all Newcomers tasks, or specific tasks. We did it for each tag: https://www.mediawiki.org/wiki/Help:Growth/Tools/Newcomer_Tasks.
I'm currently working on changing Editing's tagging as well, for Edit Check. The goal is to follow the same idea of granularity of tagging and to provide more meta information to the other users.