Since many of the aspects of the FINCH project have been created already by many different teams, a single unified extension would be ideal to start migrating that code into one unified place
Placeholder for extension proposal and discussion — https://www.mediawiki.org/wiki/Extension:FINCH
See also : T78362: Migrate PageContentSaveComplete into editing specific analytics extension
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Invalid | None | T75693 Capture browser size and screen size for web users | |||
| Duplicate | • Gilles | T76039 Hackathon: Generic viewport metrics tracking for other projects/teams | |||
| Declined | None | T75806 Capture the devices types and classes users visit the site and app with | |||
| Declined | None | T75798 Capture session length of users | |||
| Duplicate | None | T53854 Log page loads without JavaScript | |||
| Resolved | • Nuria | T58575 Browser and platform stats for logged-in vs. anon users for security and product support decisions | |||
| Resolved | None | T1384 Capture Javascript support level for web users | |||
| Declined | None | T76814 Capture network performance information from navigator.connection | |||
| Declined | None | T76809 Rename WikimediaEvents extension to FINCH to aggregate tracking information from non-feature specific sources. |
How is this different from Extension:WikimediaEvents? It seems more data is desired, but that could probably be added to the existing extension.
@Mattflaschen I don't have any major opposition to that. @dr0ptp4kt suggested a separate extension might be useful for non-foundation installs, that could find this information helpful. Additionally I would like to keep the FINCH metrics rather simple and broad, individual features will instrument many more things specific to their needs. My hope was to keep the things in FINCH pretty tailored to being usable for Product and Design (rather than starting to instrument EVERYTHING!!! which can happen over time.) decision making about users technical experience of the site, rather than about specific features.
That said, if you think this should be part of WikimediaEvents it seems like it could be a workable solution…
WikimediaEvents can be used by non-Foundation installs freely. There is an appropriate FOSS license and a note to that effect at https://git.wikimedia.org/blob/mediawiki%2Fextensions%2FWikimediaEvents.git/b552e154a4cbe92de97fd538b0b32bd0e3401be5/README#L11 . If the name still seems like an issue, maybe the extension could be renamed (again) (although the name 'FINCH' is not very clear to the uninitiated, that's not inherently a problem).
Additionally I would like to keep the FINCH metrics rather simple and broad, individual features will instrument many more things specific to their needs. My hope was to keep the things in FINCH pretty tailored to being usable for Product and Design (rather than starting to instrument EVERYTHING!!! which can happen over time.) decision making about users technical experience of the site, rather than about specific features.
I agree instrumenting everything is a bad idea (since no one will actually do anything useful with the data). In general, extensions should do their instrumentation in the specific extension (e.g. GuidedTour or Flow) rather than WikimediaEvents, so WikimediaEvents ends up being mainly for general purpose things like user preference changes (already implemented), screen size, and JS support.
So currently WikimediaEvents tracks:
PrefUpdate — which makes sense for FINCH
PageContentSaveComplete — This should likely be moved into a seperate #editing team analytics package, which I think @Catrope has already started.
I think you're right, its a good call, what is the effort level in renaming the extension? We'd have a head start with it already running on the site, should make it easier to migrate in the code from mediaviewer and zero.
Eh, I think you're inappropriately stepping into technical implementation. What does it matter to you where the code lives? Is the code's repository name really important?
Respectfully, I think you should focus on describing what you're actually trying to accomplish and leave the technical implementation to others.
If you really want to get into the technical details, why do you think a single extension is a good idea? It sounds like it would very soon become ill-defined and bloated.
@MZMcBride "stay in your lane"? My recommendations are based on conversations with developers. Designer welcome advice from engineers. I would hope that we'd all be secure enough in ourselves that Engineers could do that same for Designers.
@Jaredzimmerman-WMF: Not exactly, but this task is quickly losing focus and its place in the hierarchy is starting to look pretty silly.
This task is tracking some great ideas (T1384, T58575, T76814), but none of them are actually blocked by renaming some silly extension to an even sillier name. (Is FINCH even a real term? The English Wikipedia says it doesn't exist.) The extension name isn't important, the functionality and features are. I think brainstorming ideas for ways to make better (or at least vaguely more informed!) decisions is fantastic, but this task itself is a weird collection with the wrong summary, in my opinion.
@MZMcBride, I feel like this conversation that you and I are having now is the only thing derailing this task. https://en.wikipedia.org/wiki/Finch is a bird, you can read about https://en.wikipedia.org/wiki/Darwin%27s_finches as well to think about the ties between how FINCH could help our products evolve. What's in a name? a lot actually. Here is a short read that might help you think about things in a different way http://www.newyorker.com/tech/elements/the-power-of-names
Hey @kevinator, can you please estimate with the team. Please call out what we would need research's help with as well as what we can't do without uniques.
-Toby
Asking after nine months of silence on this task (and basically all other FINCH tasks) and no edits to mw:Extension:FINCH since December 2014:
Is FINCH dead? (Or who could answer that question?)
If so, it would be nice to reflect that in the project description. Updating its wikipage is welcome in *any* case, dead or alive. Thanks.
Maybe ask @DarTar . I'm pretty sure it existed as a project, but not an extension (hence this proposal to rename WikimediaEvents to it, which has not been done).
Boldly declining given how long this has been.
This project has been lost without a leader and budget (of resources in general, time and money) since we lost design leadership. On the Design Research backlog board, we put all of our tickets having to do with Finch in our "no resources" lane. It is not in our future goals. However, there are lots of good ideas here and could be useful someday... Not sure what people want to do with the content.
@aripstra: Thanks. I've updated the FINCH project description.
Do you know if the open FINCH tasks should potentially be in the backlog of any other related project?