Meeting started at 2026-09-08 13:00:13 UTC

The Meeting name is 'Fedora Docs Team Meeting - 2026-09-08'

!hi

Petr Bokoc: Petr Bokoč (pbokoc)

!hi

Christian Strauf: Christian Strauf (cstrauf)

!hi

Venefilyn: Freya Gustavsson (venefilyn) - they / them / theirs or she / her / hers

!hi

!hi

Hristo Marinov: Hristo Marinov (hricky) - he / him / his

!hi

Jocelyn Gould: Jocelyn Gould (korora) - she / her / hers

!hi

Eli Ridge: Eli Ridge (egret)

!hi

Peter Boy (ServerWG, Docs): Peter Boy (pboy)

Big crowd today, that's great to see!

My cats apparenlty want to join the meeting as well, so like, 15 cats are all saying !hi

We have a couple today so let's get through them

The threads are always in Project Discussion under the tag `#docs-team`: https://discussion.fedoraproject.org/tag/docs-team/337

I'm hoping this is going to work better than the meetings board because we keep forgetting it even exists and to clear the "meeting topic" tag anyway so it's not very useful

I'll also update the room description or make a sticky post to highlight this

And another thing - YOU ARE BEING WATCHED

yeah, I think aurelien can see everything we send in most matrix rooms

#docs:fedoraproject.org being an exception I think

He actually added us recently and asked us if we're okay with it, hence the ticket

makes sense

Thanks for still working on this!

Sorry for the long delays. Was/am sick ATM.

Should be alright next few days.

We need a consensus what to do with the data.

Here's what we have and my thoughts:

🔥

!hi

Justin Wheeler: Justin Wheeler (jflory7) - he / him / his

- I have a CSV with 680 lines which contains page / subcat names and the latest modification date.

- Of these, about 100 have been modified within the last 7 years.

- My proposition: blindly delete everything older than 7 years, only vet anything that's younger than 7 years.

The only problem: someone *cough* needs to check whether what the script collected makes sense. I checked a few by hand and it looked sound but it's likely it (a) entered subcats that are not 100% docs centric and (b) missed stuff. The whole "everything is linked to everything" paradigm of a wiki makes stuff hard to automate.

In any case, working on this wiki thing is not a rewarding thing. ;)

Right. A page can be in multiple categories and there was never any curation, so a page might be tagged as belonging to the Docs Project while also being still relevant to something else. Damn.

I currently see no easy way to solve this with automation.

Pages that are ONLY in the Docs Project category and none else should be okay to delete outright, if there's a reasonable way to filter those out.

"Older than 7 years" might not be an accurate enough criteria for deletion.

do we have a way to clone the entire thing somewhere before we do deltions just in case we nuke something that someone still consideres releveant?

I can't really think of a bulletproof way to do that.

There's also gonna be a bunch of user pages that we should probably leave alone.

In this case I'd upload the CSV to the issue on forgejo and propose that we deal with all pages manually but vet the newest pages / subcats first and work our way to the oldest. I really have no idea how to automate the decision whether or not something is important or even within our realm.

Okay, yeah, let's start with that.

Before uploading I'd modify the script first to collect the URLs as well which should make it easier to just click on stuff.

I didn't want to overload the output.

As soon as I'm back on my feet I'll have a look.

Okay then!

An idea, but—

We could also consider deleting in batches

Petr Bokoc: We should definitely make a decision whether to continue manually editing / deleting stuff or not.

That gives us a chance to go through them in smaller chunks and gives time for people to notice that a Wiki page is gone

If they really needed it

The stuff without edits for more than seven years is highly unlikely to be worth publishing still

I mean, that's kinda what we've been doing, just manually, and it's a PITA

That's what we've been doing anyhow. Just to stay sane. :)

I mean automated batch deletions, if that's what we were already doing, then disregard!

You can't automatically delete. Page titles don't give you any indication whether you're (a) entitled to delete the page or (b) whether it's unused.

There seems to be no way around manual vetting.

Justin could unleash an agent at it :-D

Not even through MediaWiki API?

lolol

Maybe Justin is the agent...

I use the mediawiki API to collect the pages. But the API doesn't help in the decision making process what's still needed and what's not.

Or maybe I misunderstand you.

I guess we can continue this in the ticket so we can get to other topics, I probably need to catch up. But I'm suggesting we use the API to delete pages in scripted batches, particularly the ones without any edits for more than seven years. The ones with more recent edits might warrant a closer look

Manually going through all of those pages one by one is… a lot

Let's discuss this in #docs:fedoraproject.org outside of the meeting.

And deletions can always be undone

Why doesn't Element have a tinfoil hat emoji, I'm filing an issue when the meeting's over

ACK 👍🏻

There're reasons not to do it via the API.

ack

Okay, moving on then

We're technically still in Announcements :) So, anyone else have anything to announce?

3.. 2... 1

💥

I'll start with the updates for the Beginner's Guide for F45 at the end of the week. If anyone wants to join in, you're welcome to do so, please contact me on #docs:fedoraproject.org to coordinate who's doing what.

It's not complicated, should be pretty smooth sailing.

Also: please discuss on #docs:fedoraproject.org whether you like the idea of changing the backup solution from kbackup to Vorta in the BG. Cheers!

So, I had an action item last meeting to open an issue for this in Tickets and I totally blanked on it and only noticed now, sorry about that. I'll do it... tomorrow, probably, I have meetings all day today

I have a ticket that I would like to bring up, but I think this could be left for the open floor

I have a ticket that I would like to bring up, but I think it could be left for the open floor

I'll ping the channel when I do it and I'd like people to have a look. The idea is to add a couple of simple questions to the Change template and ask Aoife to tell every Change owner to fill them out. The questions are meant to help us (relnotes writers) understand who's the target audience of each change and what exactly it does, how does it impact users, packagers, etc. The important part is to keep them simple so the people answering them don't have to think too hard. Because when people think hard they go on weird tangents and stuff and then the info is useless.

So that's about it. 5 questions max and we'll ask Aoife to keep pestering everyone to actually answer them this time.

I like targeted questions more than generic "write moar docs!!" sort of requirement for Changes

I wonder if this needs FESCo input/approval or not 🤔

<del>more text for the localizers either way</del>

We’d already had a discussion with FESCo about this two years ago. FESCo was a bit reluctant, but agreed to the changes.

Nah the questions are just for relnotes writers. Then we'll take them and turn them into an actual relnote. Which we've already been doing, but we usually have to dig the info out of the change page and/or upstream changelogs. For localizers the amount of text will be more or less the same, it's just meant to make it easier for writers.

FESCo needs to agree, of course.

Well the release notes should be based on information in the change proposal, I guess.

Since it touches the Changes process, I am wondering what the right way to make sure we have buy-in from FESCo et al for making sure this is part of the process, or if Aoife running the process is enough

Yeah, they are. (Usually - a lot of Changes are stuff like a new version of LLVM or some other tool or language, so we just link to upstream relnotes/changelogs)

I think, we need some specific info frome the change proposal. Sw we have to include 2-3 questions therein.

A while back I checked. The vast majority on proposals hat nothing inside the documentation sections.

A while back I checked. The vast majority on proposals had nothing inside the documentation sections.

Yup :) There are even two sections, one for relnotes and one for documentation, and they tend to be both empty often. That's why I mentioned we'll have to make sure people actually fill it out this time.

One issue we discussed bach then: There is to less guiding, what to write in the documentation.

Therefore the idea, ti include 2.3 specific questions.

One issue we discussed back then: There is to less guiding, what to write in the documentation.

Yeah. The questions are supposed to guide Change owners into telling us what we actually need to know. They often have no idea.

Anyway

Probably. we sjhoiuld gut together 2-4 questions together with Aoife, and then open a ticket with FESCo

Probably. we shoiuld put together 2-4 questions together with Aoife, and then open a ticket with FESCo

Peter Boy (ServerWG, Docs) Any news on this?

Well, not yet. I'm still look for a way to get in contact with join.

MatH and Ankur Sinha (FranciscoD@fedora) (he/him/his) were the two people I would normally ping for questions about the Join SIG. If we are not able to get any engagement from folks in Fedora Discussion or Matrix, we might need to rethink what we are putting on the homepage

But independend from that, we should collect ideas, what information and links in the 3-6 boxes are helpful to make contribution and to find an entry into contribution easier oder helpfull.

I am actually in the join SIG, but I have no clue what we're talking about

I tried MatH among others,l. He seeems to be offline

we've tried for a few hours on a different issue and we got nowhere

I think I have his phone number somewhere

I know Mat moved back to Austrialia, but I've not heard from him in a while.

So, maybe we should think of some other pages to put in the Docs homepage then?

🇵🇱 | Nik Nikovsky: I tried to contact join in the Matrix rooum, but w/o any feed back. When we discussed this a year or so ago, Join Team was to fill it the header box the basic info and basic "motivating" links.

Other than Join SIG, is there a list of the other pages proposed?

Not yet (besides docs, of cxourse)

Well, that's one thing, but we also need a set of clear guidelines for any future additions. Right now it's not always clear where to put links to what.

I see, I did not hear back from him for a few days

Well that explains it, dropbears got him

must be the outback spiders

Agreed. But that may change over time.

🇵🇱 | Nik Nikovsky: Could we discuss this later via direct messages? And could you try to get join aware of this?

(DM because it s easier to keep copntext.)

Guidelines would be good. Based on what we want to put there now, we could form up some guidelines for homepage inclusion

*Notes Justin watching the clock*

I think if you make/bump the ticket in the join/Fedora-Join repo, it will bring attention

*Notes Jocelyn noting Justin watching the clock*

I guess that answers the old "who watches the watchers" question

Okay, we're out of time

Thank you everyone for coming and I'll see you back in #docs:fedoraproject.org!

Thanks for chairing!