Either would be fine.

I like this. I think keeping the non-JSON pages and translations on-wiki (at least, the ones for individual TrainingSlides), but making the database the only source for the Library and Module configurations that are currently implemented as .json pages makes sense.

Sure. I think it needs more explanation of how you'd tackle the problem of having this new system alongside the existing mode of training content. How will new features be rolled out in a way that will be backwards compatible? Also, how will removing content work?

I worked with Esther on her Outreachy project to improve the docs for Programs & Events Dashboard, and I can say with confidence that she would do a fantastic job on this project. She's really thoughtful and thorough when it comes to reviewing existing docs and identifying the right changes to make. She picked up on all the technical things necessary for working with complex wikitext pages quickly, she's self-driven when it comes to diving into code to figure out how things work, and she's great with finding the right people to ask when she needs input. Working with Esther was a pleasure, and I strongly endorse this proposal.

@Ujjwalpathaak I'm not going to do full reviews of any proposals before the final submissions, but if you'd like me to look at a specific aspect, I can do that.

I've just switched it back. Hopefully it's working again now.

Ah yes, today I switched it out of P&E mode into Wiki Education mode, so that's probably what broke it.

I've just sent the data to @Rmaung. It includes every record in the database, which should stretch all the way back to when the first training module was added.

Those edits were just made a few minutes ago, and that event page has not updated again since they were made. The system currently estimates the next update in 9 hours, and after that update it should show up.

Excellent. Your description lines up with the terminology discussion we had today for "sources" vs "citations", and the ambiguity of "references". It's nice that your parser already can provide the citation count, which is what we'd use.

I've put this on my todo list. I'm on vacation next week but might be able to do it the following week.

@Isaac awesome, thanks! I agree, this data makes it seem clear that HTML will be a significantly more accurate data source for ref counts... at least, for current revisions. I can imagine that accuracy would degrade in some cases when you look at old revisions if there are very many cases of citation-generating templates that were deleted or changed, since we can generally only get HTML based on old wikitext rendered with current templates. But my intuition is that it wouldn't make much difference and would still outperform the wikitext-only approach.

@Maryann-Onyinye done!

@Maryann-Onyinye done! thanks!

@dcaro I just disabled the cron.

When was this screenshot from? I deployed the update that switched it to LiftWing about 3 weeks ago, and it shouldn't be hitting the old ORES server anymore. I just double-checked, and the Article Finder feature is behaving as expected for me with requests going to LW.

This is fine to archive. I think we set up the Gerrit repo as a mirror of GitHub, or perhaps we were considering a fork, but we've never actively used it.

Do you have timestamps for when this occurred? The database went down at some point recently, which might or might not explain the problem, if these happened when the Dashboard was basically completely un-operational.

lolol! thanks @matmarex!

Wiki Education has been getting a lot of reports about this from student editors this term. (We've heard directly from students or instructors about 20 times, which is the tip of the iceberg because most people who hit this bug don't report it to us.)

No objection to impactvisualizer as the name, if that avoids a possible problem.

It's been in good shape lately, so I'm closing this; we can re-open this if we starting facing problems again.

In T346390#9173859, @Maryann-Onyinye wrote:

Hi @Ragesoss

Kindly complete the proposal submission next steps:

• Please add your project here https://www.mediawiki.org/wiki/Outreachy/Round_27 under “Ideas for Project” and help promote it via the mailing lists. You can simply reply to the thread here: https://lists.wikimedia.org/hyperkitty/list/wikitech-l@lists.wikimedia.org/thread/IPGTE5XVAZ32DO3NWDSDQW2A3A4NZB5X/
• Lastly please submit your project proposal on Outreachy's website via this link: https://www.outreachy.org/outreachy-december-2023-internship-round/communities/wikimedia/submit-project/

If you have any questions on the next steps, don't hesitate to reach out.

Thank you!

I've been hearing about this from several different users (student editors in Wiki Education programs) in the last week, including a report that it's happening repeatedly for the same user. The error screenshots I've been sent show the "No stashed content found" message.

@Maryann-Onyinye I'd like to list this as a second Outreachy project for round 27. Could you lock this one down? (The people currently subscribed are interested parties and should all retain access.)

@srishakatux thanks! please do add @Aminehassou to the mentor group.

Thank you!

internal might be fine, but since it's only 2x the default limit which we were already hitting, I thought wme would be safer.

I fixed that truncated portion of the project description.

Fixed!

Thanks for reporting this @jhsoby! I'm sorry I couldn't get to it sooner; I was on vacation last week.

Thanks! This was a project that I haven't (yet?) gotten far enough along to do anything, and isn't running in production. If I ever take it up again, I'll be sure to use Lift Wing instead.

🎸 rad! thanks @Tgr !

Production secret is ready for use: P49567

In T334142#8986187, @vaughnwalters wrote:

see T336900#8983092 for testing for this. This ticket is ready for product sign off once we can confirm that QTE/T334142 is no longer synced with the dashboard.

In T317707#8950246, @Daimona wrote:

Yup: when enabling or editing an event registration, we run the Dashboard validation before saving the event. In particular, if the event is being created, we don't have an event ID yet.

In T317707#8938682, @Daimona wrote:

@Ragesoss Hi, one thing I just realised: when passing the dry_run parameter, in particular to the confirm_event_sync endpoint, we may not have an event ID (e.g., if it's a new event). The endpoint works just fine if I omit the event_id parameter when dry_run is true, but I thought maybe you wanted to clarify that in the endpoint documentation. Thanks!

I think this will be a lot easier for organizers if it can accept either a course slug or a URL. There's no obvious way to get the slug except by cutting it of the URL, but the pattern should be easy to handle; it's just [domain]/courses/[slug] or [domain]/courses/[slug]/home.

The pageviews (unlike other stats) are estimated based on all the views that happen between the first edit by an event participant and the most recent time the stats got updated, not just through the end of the event. So increasing view counts after the end of the event is normal.

Are you referring to the 'Activity' tab? That tab does not show tracked edits, it pulls the latest edits from any event participant from any of the tracked wikis. However, any edits after the end date will not be reflected in the cumulative stats or anywhere else except the Activity tab.

I think this proposal is not sufficiently focused on the main objective of improving the user experience for users who are setting up a course/event for which they want to track only a specific set of pages. I don't think React upgrades or converting between HAML and JSX are very relevant to this UX, because the current workflow for setting up a course is all in React already. Refactoring to move queries from templates to controllers is also not very relevant; that's a code quality issue that could be done at any time, and should not affect the user experience.

Hi! This looks like a proposal that is primarily a design project, rather than a UX development project that also includes a design element. It will not be realistic to both learn the necessary programming skills and implement a new design in ~80 hours. It also won't be a competitive application without an accompanying track record of preliminary contributions to the Dashboard codebase. I usually encourage prospective applicants to explore the codebase and make several contributions first, before attempting to put together a proposal, as a strong proposal needs to demonstrate an understanding of the codebase and how the proposed project fits into it specifically.

This is a well-thought-out proposal. One thing that could improve it would be to include some preliminary research about what options might be appropriate for replacing the jquery UI library for the survey system (and what the tradeoffs might be in terms of whether it's worth it to do so).

Akanksha8800: sent.

In T328698#8707329, @Fedepoi159 wrote:

HI! I have experience with React and I would love to start contributing to this project. Can I please get the Slack invite so I can get involved in the community?

Thank you!
Federico

@eren023 check out the 'students and interns' doc linked in the description. That has info on how to get started.

In T329796#8620406, @matmarex wrote:

You can see the problem in the accessibility tools in Chrome developer tools – apparently the markup indicates that the entire message, including other buttons, is a label for each button.

In T329796#8620359, @Tgr wrote:

A side effect of T102973: OOUI-ify the OAuth authorization dialog, maybe?

A fix that improves the error message is ready, and will be live after the next deployment.

How long was it down, that you noticed? It's up for me right now.

This is complete, and it's been deployed. Now the CSV for 'editors data' includes the global ID.

Thanks! @dcaro I have some longer term plans to reduce the data requirements — basically, eliminate the need to store revision data long-term, which is the bulk of the disk usage — but it will require some significant architecture adjustments to get to that point.

I tested this on the example from above, the track listing table from https://en.wikipedia.org/wiki/Renewal_(Billy_Strings_album), but it isn't fixed even after purging the cache, making a null edit, or previewing the table with a non-null edit.

This makes perfect sense. I've opened an issue for it on GitHub: https://github.com/WikiEducationFoundation/WikiEduDashboard/issues/5195

Thanks Boyani. That's a potentially interesting idea, but I don't know whether any open source text-to-speech engines are good enough to provide a good listening experience — especially for the sometimes-technical content of the training modules. I'm unlikely to prioritize a project like this, although if you know more about how this could be implemented, I'm interested to hear details.