What sort of outcomes are you looking for in such a meeting? Are you looking to meet with engineers about technical issues, or managers to ask about formally committing WMF resources?

@brion, this is broken off from T117193: Prepare MediaWiki Stakeholders' Presentation for January Developer Summit at @Qgil's request. We'll meet tomorrow to nail down what we want to talk about.

In T119403#1826003, @brion wrote:

What sort of outcomes are you looking for in such a meeting? Are you looking to meet with engineers about technical issues, or managers to ask about formally committing WMF resources?

Blatting some ideas from @Ckoerner here.

• People use MediaWiki!
• How can we bring them into the fold?
• What is the WMF stance on MediaWiki? Is it part of the mission or a by-product of it?
• Roadmap roadmap roadmap

More at my weblog which I'll refrain from posting here.

We'd all like to know WMF's stance on MediaWiki...

I read the blog post, and it is still unclear to me what is the outcome of the meeting you expect and, more specifically, what is your ask to the WMF. I could deduce it... but that is a risky exercise. I'd rather let the MediaWiki Stakeholders Group come with a specific ask.

In T119403#1839561, @Qgil wrote:

I read the blog post, and it is still unclear to me what is the outcome of the meeting you expect and, more specifically, what is your ask to the WMF. I could deduce it... but that is a risky exercise. I'd rather let the MediaWiki Stakeholders Group come with a specific ask.

I appreciate that. As Brion says, we'd like to know the WMF's stance on MediaWiki. Is MW just a piece of software that helps them create sites like Wikipedia? This would mean that making the software available for others to use is just an inessential by-product.

Conversations with other people in MediaWiki-Stakeholders-Group are on-going, and we haven't finalized what we would like, but, in the interest of keeping the conversation going here, I think the following are some good actionable items:

people

• Hire developers dedicated to non-WMF uses of MediaWiki.
• Give the developer relations team a broader scope so that developers who contribute to MediaWiki with non-WMF uses in mind are engaged on a regular basis.
• Talk to users who've set up sizable private wikis and find out what they're needs are.

direction

• Make sure there is a solid upgrade path and road map for the software so that MW is treated as an actual piece of software that can be used for a variety of purposes other than the WMF's needs..
• Ensure dependable APIs for developers of extensions that aren't used by the WMF.
• Clearly communicate deprecations (e.g. the page view counter) and provide reasonable fallbacks.

I've tried to make the above more results-oriented rather than just the whiney rant like I posted to my blog, but, as I said, these are just my thoughts and my initial response to your interest in a specific ask. I'll ask the others for feedback.

Pasting some Phabricator tasks which come to my mind:

In T119403#1839754, @MarkAHershberger wrote:

• Hire developers dedicated to non-WMF uses of MediaWiki.

Potentially related: T88265: Investigate sources to fund third party MediaWiki features

• Give the developer relations team a broader scope so that developers who contribute to MediaWiki with non-WMF uses in mind are engaged on a regular basis.

Potentially related: T85600: Who are the top 50 independent contributors and what do they need from the WMF?

• Talk to users who've set up sizable private wikis and find out what they're needs are.

T954: Prioritize List of Top MediaWiki Features Requested by Third Parties?

• Make sure there is a solid upgrade path and road map for the software so that MW is treated as an actual piece of software that can be used for a variety of purposes other than the WMF's needs..

Potentially related: T119711: Let update.php check for composer updates in extensions?

• Ensure dependable APIs for developers of extensions that aren't used by the WMF.
• Clearly communicate deprecations (e.g. the page view counter) and provide reasonable fallbacks.

T114384: Standardise procedures for deprecating public-facing code?

I'll ask the others for feedback.

Has that happened? :)

Aklapper writes:

I'll ask the others for feedback.

Has that happened? :)

Yes. We're going to be collating our thoughts over the next few weeks
so that @cicalese and I have something concrete to discuss when we meet
with the WMF to discuss this.

@Rfarrand, Any update on this?

Hello - I am unclear exactly what this event is or what actions are needed on my end, however I am happy to help if I can be useful.

What type of event is this? A private meeting? A public tech-talk / hangout? An IRC meeting? An in person meeting?

Who, specifically, needs to be there?

If someone can provide a definition of the event, a specific description of the event, as well as a list of people and their contact information who should be there, I can help set it up. Otherwise I think this still may need some discussion before I can help out.

This is a private meeting to be scheduled on Wednesday at the WMF offices, with the participation of ? members of the MediaWiki Stakeholders Groups and then probably Wes (Product), Robla (Technology), myself (Community Engagement), and probably someone with a double role WMF / 3rd party.

@Rfarrand, I had asked you to organize this, but maybe it is simpler to ask i.e. Praveena to schedule this meeting?

@RobLa-WMF, I believe that you are missing one valuable type of meeting in your "Good meetings" taxonomy: an introductory meeting to put faces to names and set the foundation for valuable collaboration. This is especially useful in a community that is normally geographically disparate when there is an occasion to come together in person.

The MediaWiki Stakeholders group comprises dedicated third-party users of and contributors to MediaWiki. We have volunteered our time and resources to MediaWiki in many ways from contributing code to developing, administering, and analyzing the results of a survey of third party MediaWiki users. We would like to share our perspective on the needs of third party users/developers such as ourselves and build a basis for a strong working relationship with the Wikimedia Foundation going forward. As this meeting comes at the end of the MediaWiki Developer Summit, we hope to also be able to provide our perspective on decisions and progress made at the event.

We have a lot to offer to help make the great software that is MediaWiki even better. We also have an interest in making sure that MediaWiki continues to be a viable platform for third party users.

Our request for a meeting was made publicly in Phabricator in accordance with the standard practices of this community. If a private meeting is more appropriate, please consider this a request for a private meeting.

@Qgil, makes sense to me to have Praveena do the scheduling, thanks

This works for me. I'll update the description.

@Qgil, what else do we need to do get a time for this meeting so that @cicalese and I make sure we're both available when the people from the WMF are available?

@MarkAHershberger: I didn't follow this task closely but as "unconference slot" seems to be the preferred option (instead of "meeting scheduled by Praveena"), the program lists Unconference parts.
T119403#1894641 suggests Quim, Wes, Greg, and someone from Strategic Partnerships to attend for the WMF part - want to drop them a quick email in advance maybe?

@Aklapper, thanks for the help. I appreciate all the pointers and patience as we try to make this happen.

Sending emails....

This is scheduled for 11:30 in unconference room 3

Etherpad is https://etherpad.wikimedia.org/p/steak_and_wine

Here's the etherpad copy:

WMDS session with MediaWiki Stakeholders group/WMF

Cindy talked a bit about MITRE and their wiki farms. There are a lot of different types of third party users, some are on shared hosting, etc.

• Upgrading is a huge issue even though we try and stay current with the most recent release

Purpose of this meeting:

• Third-party users bring benefits to WMF/MW, more users testing, write patches, file bugs, extend things, etc.
• Make sure the WMF knows that third-party users exists, don't want to get abandoned, some things look scary?

There's a paper agenda: is https://docs.google.com/document/d/1cCf07DiL4bXJ29qvDFESo1Z22HeFajxwS5ihTg_cu1M/edit?usp=sharing

• Side 1: https://drive.google.com/file/d/0B7UN7tgsOZ8qYklnaFRiM21LNkk/view?usp=sharing
• Side 2: https://drive.google.com/file/d/0B7UN7tgsOZ8qUVVqSmRvNnhld2c/view?usp=sharing

Main problems:

1. No overall guiding principles for MW core development
2. Compatability breaks
3. WMF does not track what is important for third party users
4. Third-party users don't have an effective mechanism to voice concerns to the WMF

Mark: Basically there is no roadmap/plan for MW

Mark has some proposed solutions (they're on the paper agenda too):

1. WMF should provide a vision/roadmap for MW core development
   • Helps third party users plan and stuff
   • Roadmap provides communication for future dev, release(?), and potential breaks in back-compat
2. (we need this internally too! lots of clear plans on individual work areas, but WMF does not not a good roadmap for MW core right now --brion)
3. see also https://office.wikimedia.org/wiki/Engineering/New_core_team (sorry non-WMF folks, no public page ATM)

[sidebar: MW users survey -- https://www.mediawiki.org/wiki/2015_MediaWiki_User_Survey ]

2. WMF should appoint a rep to engage with MW stakeholders so that the WMF know what is important to third-party users

Gilles asked mark: if you had a million dollars, what would you do?

• extension store
• content bundler
• something like the wordpress foundation thing

WMF needs to see MW as something besides the thing that runs Wikipedia
Dept of Agricuture, for example, runs a wiki to track Supplemental Nutrition Assistance Program (SNAP) regulations.

bawolff: lack of vision isn't limited to just third party/MW, it's the foundation entirely
matt_f: there's no one team responsible for planning a roadmap, (we do have archcom) - we need a long term roadmap, it's not a lack of vision.
bawolff: lots of ambitious things, but nothing tying everything all together. Don't know where we want to be in 5 years
cindy: we sell the concept of MW to people, but we want to know that we can support it for the next 5 years. We set it up, but might not maintain it.
brion: WMF doesn't have a MW core dev team, archcom is organized as very reactive. We don't have a good set of people to put together a roadmap. Most WMF teams are feature oriented, and have specific tasks, or are doing WMF scalability-type issues. Brion wants to fix the lack of a core roadmap. He also is willing to take on a point-person role at WMF. Don't know if we'll ever have a team to support third party users, but we need open communication
quim: when you say core, do you actually mean MW core (as defined by git repo), many new features are being developed in extensions.
mark: core means what you get in the tarball basically
mark: "I would be happy if MW would stay the same and be supported" <laughs>
quim: [said something and I got distracted]
mark: WMF should appoint someone to interface with stakeholders, get info from various users and pass it back to WMF.
brion: I'm the point person now.
quim: but brion is busy as hell
brion: For now I can help with collecting info and clarifying the roadmap
Ryan: WMF purpose isn't to maintain MW, it's a side-affect. It releases software because it believes in open source, but doesn't actually have developers supporting it, and when it does, they get dissolved. We need a MediaWiki Foundation (MWF), that raises its own funds, etc. The fact that WMF maintains MW, holds both WMF and MW back.
matt_f: if WMF is doing annual planning, should they also consult stakeholders?
cindy: we need a WMF point person, stakeholders are just one group reaching out to third-party users. We do need engineering resources, but [I didn't understand this part]
mark: Who's gonna create a MWF?
gilles: Creating a MWF should be very similar to Wikimedia chapters
Ryan: Look at WordPress and OpenStack models.
quim: something about WMF has dev relations team, ...., pgsql support isn't going to be paid by WMF ever
Mark: I made a point about joining IEG so I could help provide technical review of IEG requests
quim: ...
Ryan: MW is an official project of the WMF. (bawolff: so is Wikispecies)
gilles: something about fundraising, donors would be okay
Ryan: I am more than willing to help setup a foundation

[people raise hands to record names of people willing to help set up MW Foundation]

• Ryan Lane
• Matt Flaschen
• Addshore
• Tyler Romeo
• Legoktm
• Quim Gil (to be contact person on Wikimedia organizational side)
• Brion
• Greg G
• Ray Paser ray.paseur@armedia.com
• Mark Hershberger
• Cindy Cicalese

Ray: One possibility to consider is how Alfresco (open source document management) works. It's open source, but there's also a corporation behind that. They make a living providing a fully supported version

Greg-g: WMF technology roadmap should cover some/all? concerns. Something about service-workers, and we need a CTO-type position that is yet to be hired
Wes: How long of a roadmap do you want?
Mark: 5 years
[someone I don't know]: Can you have a roadmap for 5 years from now? Mark: 1 or 3 years would be fine
[same someone]: low hanging fruit is to create a roadmap based on what was already done
cindy: long term and short term goal, MWF isn't going to be set up tomorrow. Need roadmap in the interim for short term. Have not breaking things be a goal / or make it easier to fix broken things
adamw: Brainstorm funding model of MWF? Beg for WMF grants, but the hand that feeds you is the hand you love
gilles: Apply from grants from the users
Ryan: Some support model or something
[got distracted]
brion: very easy to break stuff when upgrading, it's mostly customizations [we need to improve our extension/skin interfaces so it's easier to make something that will survive an upgrade!]
tyler: better testing for third-party stuff -- virtual testing team?
lego: we can have postgres tests block merge if someone adds CI infra support
[same someone]: What features does WMF not support? greg-g: Everything that is not in WMF production.
ejegg: other things we don't test like alternative storage backends
Matt_F: different funding models, could have MWF hire people, or more like apache where different companies assign people to do stuff
quim: strategy process for WMF is going on, reply that WMF doesn't care is better than no reply
mark/greg: how do we do that? <looks at Wes>

Mark: Ryan Lane is the ED/CEO of the MWF. <laughs>
/end

As the meeting happened, can this task be changed to status "resolved" or what is left here?