resync

It is requested that an edit or modification be made to this protected page.
Administrators: Please apply <nowiki> or {{Tl}} to the tag after the request is fulfilled.

---

Fulfill this request after checking that the code is safe, beautify and escape unsafe characters

Doc James

Can this be resynced with https://www.mediawiki.org/wiki/MediaWiki:Gadget-Global-ImageStackPopup.js to get the new version of the gadget? Bawolff (talk) 17:49, 9 October 2024 (UTC)

Hm, I’m not sure how I feel about that use of mw.track(), to be honest – especially the inclusion of wgPageName feels like it would make it easy to follow which pages people visited? Is there a precedent for this? (I only found a handful of mw.track() uses in other gadgets here or on enwiki, and nothing near that level of granularity.) Lucas Werkmeister (talk) 18:37, 9 October 2024 (UTC)

The results get displayed at https://grafana.wikimedia.org/d/IlK0cZbSk/gadget-stats?orgId=1&refresh=5m&var-prefix=gadget_ImageStackPopup&var-metric=All&from=now-24h&to=now It shows where the gadget has been used (the track only triggers after user interacts with gadget) but not who is using it. The primary goal is to try and figure out on which pages the gadget provides value and which it doesn't. For example in order to be able to refine templates and figure out which changes work and which dont (this is especially true on wikipedia where different articles will integrate the gadget into the page in different ways. Just having overall stats on usage doesn't really provide much insight into what is working and what isn't with regards to the usage of the gadget). Bawolff (talk) 20:40, 9 October 2024 (UTC)

It is not tied to any specific user. So how would this be different than us collecting data on pageviews for a specific image? How about we collect it by 24 hour periods? Doc James (talk · contribs · email) 20:44, 9 October 2024 (UTC)

@Bawolff, @Doc James: Sorry, I didn’t see your replies before now. So how would this be different than us collecting data on pageviews for a specific image? It’s certainly different in temporal granularity – having the gadget usages pinned down to the exact second reveals more information than pageviews’ once-per-day data. (I thought there was also a difference in reporting the exact number of accesses even if it’s very low, but apparently pageviews does that too. Ouch.) If you collect it by 24 hour periods, I think it would be okay (or at least, not worse than pageviews, even if I don’t like the new pageviews behavior I discovered either ^^), but is that doable with mw.track()? (It might be possible to make Grafana only show “binned” data, but IMHO that would not be sufficient, because anyone can still access the underlying Graphite data directly. I would want the unaggregated data to be protected.) Lucas Werkmeister (talk) 15:04, 18 January 2025 (UTC)

I am happy with values collected by day. In fact do not want further granularity beyond that. Doc James (talk · contribs · email) 21:30, 18 January 2025 (UTC)