but also, some projects that were not so useful, sure. But we learn, move on, we're not the first group of people to make mistakes : )

_______________________________________________ Wikimedia-l mailing list, guidelines at:

New messages to: Wikimedia-l(a)lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l,

Does it make sense to have an "Incubator team" ("Bell Labs" if you will), whose core competency is to nurture small projects? When projects are mature and need to switch into maintenance mode, they move under the umbrella of a different team. On Wed, Feb 24, 2016 at 5:06 AM, Brion Vibber &lt;bvibber(a)wikimedia.org&gt; wrote: wrote: _______________________________________________ Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines New messages to: Wikimedia-l(a)lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe>

Brion, there was a very constructive, heartfelt session on models of bottom-up open innovation at this year's WMF All Hands. You can find extensive notes from this session on the Office Wiki ("Embracing skunkworks") which I encourage you to read and that I'd love to share publicly in a more readable format at some point.

There are obvious tradeoffs between allowing more flexibility on the one hand and making sure we have a reasonable budget plan, accountability to donors and stakeholders and appropriate resource allocation on the other hand, but I believe this model would work much better than the current

at least for projects that are not core initiatives.

Skunkworks is what got us revision scoring, EventLogging, countless initiatives by TechOps and innovative MediaWiki extensions. Dario On Tue, Feb 23, 2016 at 6:14 PM, Yuri Astrakhan &lt;yastrakhan(a)wikimedia.org&gt; wrote: > Does it make sense to have an "Incubator team" ("Bell Labs" if you

> whose core competency is to nurture small projects? When projects are > mature and need to switch into maintenance mode, they move under the > umbrella of a different team. > > > > On Wed, Feb 24, 2016 at 5:06 AM, Brion Vibber &lt;bvibber(a)wikimedia.org&gt; > wrote: > > > On Feb 23, 2016 5:52 PM, "Dan Andreescu" &lt;dandreescu(a)wikimedia.org&gt; > wrote: > > > > > > but also, some projects that were not so useful, sure. But we

> > move > > > on, we're not the first group of people to make mistakes : ) > > > > Yep... High-tech organizations call it "failing fast". > > > > -- Brion > > > > > _______________________________________________ > > > Wikimedia-l mailing list, guidelines at: > > https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines > > > New messages to: Wikimedia-l(a)lists.wikimedia.org > > > Unsubscribe:

-- *Dario Taraborelli *Head of Research, Wikimedia Foundation wikimediafoundation.org • nitens.org • @readermeter <http://twitter.com/readermeter> _______________________________________________ Wikimedia-l mailing list, guidelines at:

New messages to: Wikimedia-l(a)lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l,

If I remember correctly, I think that's how the Content Translation project started -- it was someone's personal project, which got more people and attention because it's a great idea and showed real success.

On Tue, Feb 23, 2016 at 7:53 PM, Yuri Astrakhan &lt;yastrakhan(a)wikimedia.org&gt; wrote: open of rather So glad to see this being discussed in the open with smart folks like Brion, Dario, et al. 3D support would be most welcome – we’re in a holding pattern with Smithsonian 3D GLAM projects in DC because of that shortcoming in Commons. It was never known whether anyone was paying attention or going to put 3D on the radar screen. (https://phabricator.wikimedia.org/T3790) At my keynote talk at Wikiconference USA, I said one of the things WMF “must do” is multimedia and interactivity. Your work on the interactive graph is a great step. Brion, myself and others have been working on collaborative video. Brion’s ogv.js work is a great example of skunkworks type projects having a huge impact. And if 3D is given a priority, the three areas would be a great collective step towards Wikipedia continuing its revolutionary work. Best of all, they would be technologies developed in service of content and community needs, and not simply created for tech’s sake. An organism reacts to the change of its environment by redistributing Yes, and in a recent meeting you mentioned Bell Labs as a model. As someone who worked there, it’s a very good ideal to shoot for. https://www.mediawiki.org/wiki/Discovery/2016-02-16_Discussing_Knowledge_En… Thanks for opening up this discussion. -Andrew _______________________________________________ Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines New messages to: Wikimedia-l(a)lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe>

I would like to clarify a fairly major premise of this conversation: namely, the comment I made that Yuri quoted in the very first message. When I say that the hierarchical organisation of the Foundation is something that is preventing us from doing better, I was not thinking of how we develop software. Indeed, I suspect that peoples' tendency to bring things constantly back to "does it improve the measurable speed at which we right code" is symptomatic of the problematic dynamics here. What I was thinking about was how we pay attention to organisational hiring, to how we promote, to how we treat people, what empathy we have and how we value empathy. I have consistently found the Foundation to lag in all of these regards. It is not good at making sure that the recognition of employees is fair and treated equitably (be that who gets called out in presentations, who gets given opportunities, or who gets raises). It is not good at making sure that how we hire is fair. It is not good at making sure that concerns of employees are given weight. All too often the people marginalised by our approaches are the people marginalised outside the Foundation, as well; women, people in "non-technical" roles, people in roles that we code as "support work" (and guess what tends to correlate with a role being coded as support work?) All too often the work marginalised by our approaches is the work that Doesn't Product Code (again: guess who tends to do the heavy lifting on things like organisational health and process and structure?) As an organisation I have found the Foundation overly rigid and resistant to the most conservative change around these problems; particularly I think of efforts to improve unintentional bias in our job descriptions. Basically, unless you as an employee go out and do the damn work yourself, for free, with 0 recognition of the emotional and temporal cost of that work, it doesn't get done. The organisation as a whole is not interested. Switching to a flat organisational structure does not, in any way, solve for this problem. In fact, in some way it makes it worse, because it makes us *think* that we have solved for the systemic and hierarchical power dynamics that make it difficult for low-level or marginalised people to get things done, or people doing marginalised work to get things done, when we have only shifted them. To pick on someone, I pick Trevor (sorry Trevor. For reference this is an entirely hypothetical example and Trevor is lovely): Trevor's voice is given a lot more weight in the organisation than mine. Trevor has a lot more influence than I do. Trevor has a lot more influence than most WMFers do! Crucially: this *isn't because he's management*. This was the case even *before* he was management. Because: 1. He's been here a really really long time and so knows everyone. 2. He's an Engineer, and we give engineers more weight and cachet than we do, say, administrative staff or people in "support" roles, even though those people are both as-smart and have an equal interest in the organisation's success; 3. His background matches what we strongly correlate with Authority Voices. If we switch to a flat organisational structure where nobody has a title, or..whatever, all of these things will still be true. We will switch pronounce systemic biases or uneven power dynamics Done, and we will have achieved something that's actually worse than not doing anything at all. Because now, we still *have* all those problems, we just think we're done and don't have to put any work in and can't talk about it, and nobody has the responsibility for continuing to fix things. The Foundation I would return to is not an organisation with a flat structure. In fact, it could be an organisation that looks a lot like this one, because I don't believe reporting lines or titles have as much of an impact on dynamics as we think they do. What *does* have an impact is how we recognise the value of emotional labour, how we recognise our implicit biases and advantages, and how honest we are with each other: not just in terms of what we *say* but in terms of how we *listen*. In other words, the litmus test for me is: what happens when the socially and politically weakest person in the organisation has an idea? Anyway; I don't particularly want to go into a long drawn-out conversation, just correct the initial, fundamental misunderstanding. Hopefully I've provided a bit of food for thought along with that. On Wed, Feb 24, 2016 at 3:50 AM, Pau Giner &lt;pginer(a)wikimedia.org&gt; wrote: project iterations anyone a consider. wrote: yastrakhan(a)wikimedia.org&gt; example Commons. 3D skunkworks collective they needs, more someone https://www.mediawiki.org/wiki/Discovery/2016-02-16_Discussing_Knowledge_En… https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines <mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe> _______________________________________________ Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines New messages to: Wikimedia-l(a)lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe>

In other words, the litmus test for me is: what happens when the socially

Hey Yuri, [Responding offlist because I'm linking to officewiki]

On Wed, Feb 24, 2016 at 5:38 PM, Yuri Astrakhan &lt;yastrakhan(a)wikimedia.org&gt; wrote: socially An something the how having "idea draw My email opened with "I think reducing things to engineering terms are sort of indicative of the problem here". I'm not talking about code. I'm not talking about designs. I'm not talking about software products. And thinking about it in terms of engineering projects, which is what we do as an organisation a lot, will not be helpful. If it did, then after several years of insisting that we are primarily a tech shop, we would hopefully not still be having conversations about structure and direction! What I am talking about is ideas generally. They might be about software products. They might be about social products, a la the teahouse. They might be about how to tweak our process by which we interact with the community. They might be that our hiring process is kinda weird and here's this one cool way we could look at improving it. They might be that the break room snacks _suck_ (again, hypothetical: they're fine. Sorry, Facilities). In any case, the litmus test is just that; a litmus test. Our structure should be designed cognizant to these problems, and then pass the test, but not be designed *specifically* to pass the test. And the designer idea seems pretty hyper-optimised just for the test. https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines <mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe> _______________________________________________ Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines New messages to: Wikimedia-l(a)lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe>

Guillaume, the idea may come from anywhere, shouldn't we post the process on meta? Or is this WMF specific, e.g. "I want my favorite cereal in the cafeteria" proposal? :)

I would like to clarify a fairly major premise of this conversation: namely, the comment I made that Yuri quoted in the very first message. When I say that the hierarchical organisation of the Foundation is something that is preventing us from doing better, I was not thinking of how we develop software. Indeed, I suspect that peoples' tendency to bring things constantly back to "does it improve the measurable speed at which we right code" is symptomatic of the problematic dynamics here. What I was thinking about was how we pay attention to organisational hiring, to how we promote, to how we treat people, what empathy we have and how we value empathy. I have consistently found the Foundation to lag in all of these regards. It is not good at making sure that the recognition of employees is fair and treated equitably (be that who gets called out in presentations, who gets given opportunities, or who gets raises). It is not good at making sure that how we hire is fair. It is not good at making sure that concerns of employees are given weight. All too often the people marginalised by our approaches are the people marginalised outside the Foundation, as well; women, people in "non-technical" roles, people in roles that we code as "support work" (and guess what tends to correlate with a role being coded as support work?) All too often the work marginalised by our approaches is the work that Doesn't Product Code (again: guess who tends to do the heavy lifting on things like organisational health and process and structure?) As an organisation I have found the Foundation overly rigid and resistant to the most conservative change around these problems; particularly I think of efforts to improve unintentional bias in our job descriptions. Basically, unless you as an employee go out and do the damn work yourself, for free, with 0 recognition of the emotional and temporal cost of that work, it doesn't get done. The organisation as a whole is not interested. Switching to a flat organisational structure does not, in any way, solve for this problem. In fact, in some way it makes it worse, because it makes us *think* that we have solved for the systemic and hierarchical power dynamics that make it difficult for low-level or marginalised people to get things done, or people doing marginalised work to get things done, when we have only shifted them. To pick on someone, I pick Trevor (sorry Trevor. For reference this is an entirely hypothetical example and Trevor is lovely): Trevor's voice is given a lot more weight in the organisation than mine. Trevor has a lot more influence than I do. Trevor has a lot more influence than most WMFers do! Crucially: this *isn't because he's management*. This was the case even *before* he was management. Because: 1. He's been here a really really long time and so knows everyone. 2. He's an Engineer, and we give engineers more weight and cachet than we do, say, administrative staff or people in "support" roles, even though those people are both as-smart and have an equal interest in the organisation's success; 3. His background matches what we strongly correlate with Authority Voices. If we switch to a flat organisational structure where nobody has a title, or..whatever, all of these things will still be true. We will switch pronounce systemic biases or uneven power dynamics Done, and we will have achieved something that's actually worse than not doing anything at all. Because now, we still *have* all those problems, we just think we're done and don't have to put any work in and can't talk about it, and nobody has the responsibility for continuing to fix things. The Foundation I would return to is not an organisation with a flat structure. In fact, it could be an organisation that looks a lot like this one, because I don't believe reporting lines or titles have as much of an impact on dynamics as we think they do. What *does* have an impact is how we recognise the value of emotional labour, how we recognise our implicit biases and advantages, and how honest we are with each other: not just in terms of what we *say* but in terms of how we *listen*. In other words, the litmus test for me is: what happens when the socially and politically weakest person in the organisation has an idea? Anyway; I don't particularly want to go into a long drawn-out conversation, just correct the initial, fundamental misunderstanding. Hopefully I've provided a bit of food for thought along with that. On Wed, Feb 24, 2016 at 3:50 AM, Pau Giner &lt;pginer(a)wikimedia.org&gt; wrote: _______________________________________________ Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines New messages to: Wikimedia-l(a)lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe>