Wikipedia:Wikipedia Signpost/2016-02-24/Op-ed

The Signpost
← Back to Contents
View Latest Issue


Op-ed

Backward the Foundation

Contribute  —  
Share this
By Oliver Keyes

When I joined the Wikimedia Foundation there was an operations engineer named Ryan Lane. In most respects, Ryan was just a standard operations engineer, but every Friday as work wound down he'd crack open a big archive drawer full of drinks and make everyone cocktails.

It was a nice opportunity for mingling: legal, administration, engineering and global development would all come around and have a cocktail (alcoholic or not) and chat. It broke down barriers between departments, sequestered as they are on different floors, and let visiting remote employees meet people they might not otherwise have encountered.

Then Ryan left and nobody bothered organising it any more, and the function went away.

When I joined the Wikimedia Foundation there was an Executive Director named Sue Gardner. In most respects, Sue was just an ED, but she instilled values of transparency outside the organisation and transparency inside it. She created an environment where you might disagree with a decision, but you could still respect it, because you were there for it. She created an environment where everyone, even executives, were answerable for the work they did and did not do. She created an environment where dissent was expected and valued rather than classed as unprofessional.

Then Sue left, and guess what happened next?

Culture is a fragile thing, much more fragile than we expect. When we're in the middle of it, good or bad, it just ... fades into background noise. It's taken as a given. And then people leave and you slowly realise both how valuable their presence was, and the fact that the things they were doing aren't anywhere in anyone's job description, or things you were hiring their replacement with an eye to.

When Ryan left, we lost alcohol. And hey, I can deal with an absence of alcohol. Tech and alcohol haven't exactly been the healthiest of friendships. But when Sue left, we lost a lot of our transparency, internally and externally. When Gayle left we lost a boundless love and fierce determination to make us do better and comfort us when we didn't. When Anasuya left we lost steady counsel, an awareness of the width of the world and knowledge of the multitudes it contains.

We hired for none of these values. We tasked for none of these values. And so we have, organisationally, none of these values. The things that always distinguished the Wikimedia Foundation as a workplace are gone, and replaced with an environment that prizes unanimity above confidence and lacks accountability for organisational failures.

Because of that, I am leaving. I don't know what things I did that nobody will organise now, but I do know that I am not looking back. This was a good place to be – I wouldn't have spent half a decade at it otherwise – but it has tarnished and rusted every day of the last year and a half.

To my friends in the wider movement, I would ask you to keep insisting that the organisation do better. Insist until your keyboard is worn down, insist until your lungs give out, insist until the next Board election and the next opportunity to make the people holding the job at the moment actually do it.

To my friends in the organisation – and there are so very many of them, so many wonderful, glorious, loving people – you owe your people trust and respect and protection, and sometimes that is shielding them at your expense. But sometimes it is getting out while you still can, so as to set an example that leaving is a thing that can be done. If you wait to leave until you have pulled everyone out, you'll be consumed by it. I would not wish that on any of you.

If anyone wants me, I'll be at Rapid7, a company whose employees like being there, whose work is interesting, and whose managers are accessible and answerable. Turns out being a security/privacy nerd who likes data is, in fact, remunerative. If you're reading this: I'm sure you can work out where to find me if you need me.

Oliver Keyes is a Research Analyst at the Wikimedia Foundation until March 18. This article was originally posted on the author's blog and is republished with his permission. The views expressed in this article are his alone and do not reflect any official opinions of this publication.

← Previous "Op-ed"
Next "Op-ed" →
+ Add a comment

Discuss this story

These comments are automatically transcluded from this article's talk page. To follow comments, add the page to your watchlist. If your comment has not appeared here, you can try purging the cache.

I am guessing that "opsen" is short for "operations engineer"? It took me a few minutes (and failed lookups in Wiktionary and Google) to figure that out. — Mr. Stradivarius ♪ talk ♪ 05:15, 25 February 2016 (UTC)[reply]

Ooh, good catch; I'll tweak it (the inside baseball references made a lot more sense when it just lived on my blog). Thanks! Ironholds (talk) 05:16, 25 February 2016 (UTC)[reply]
That would be a good thing to add to Wikitionary. I can't seem to find any sources for it though. Gamaliel (talk) 05:43, 25 February 2016 (UTC)[reply]
I also couldnt quickly find a source for it, but I did see it used as a twitter hash tag with this meaning. I added a redirect & definition on wikitech. John Vandenberg (chat) 01:28, 19 May 2016 (UTC)[reply]
  • Thank you for your truthfulness about your personal opinion. I know none of the details of what has been going on behind the scenes, but (I guess like many others) I am alarmed by the shenanigans, and hope for a fresh start with the culture of open collaboration that you describe. I wish you well in your career and hope you will always strive to create openness and transparency wherever you are. Chiswick Chap (talk) 09:25, 25 February 2016 (UTC)[reply]
    I hope so too! And in the meantime I have to rewrite the article on Aaron Burr. Thank you for your mind words! Ironholds (talk) 13:06, 25 February 2016 (UTC)[reply]
  • We hired for none of these values. We tasked for none of these values. And so we have, organisationally, none of these values.. That is very well put. Thank you for sharing your thoughts, and your hard work at the WMF. Husky (talk page) 09:29, 25 February 2016 (UTC)[reply]
    Thank you. It was...I'm not going to say "fun", obviously, but it was good to be able to so easily translate rather inchoate and nuanced thoughts into words. Ironholds (talk) 13:06, 25 February 2016 (UTC)[reply]
  • I totally agree that culture is fragile. I coded in a different era when it was OK if we had alcohol while out at lunch with our department, as long as we promised not to touch the code for the rest of the day. That was generally at goodbye lunches where the person had been bought by one of our competitors in Massachusetts (all of them defunct by now). I am pretty sure what we did back then would be considered illegal today. So are you headed to Massachusetts? Maybe you can help them fix what they have started 3x and been reverted. Good Luck, Jane (talk) 12:05, 25 February 2016 (UTC)[reply]
    Already in Mass! I'll try not to edit work-related stuff :D. Ironholds (talk) 13:06, 25 February 2016 (UTC)[reply]

"The things that always distinguished the Wikimedia Foundation as a workplace are gone, and replaced with an environment that prizes unanimity above confidence and lacks accountability for organisational failures." Strange, the impression I have from here (enwiki) is that the WMF for years and years completely lacked "accountability for organisational failures", of which there were many (as you should be well aware), and that only during the last years or so has there been any improvement in this. I don't know about unanimity, but one of the reasons for the many failures was overconfidence (why should we test or have a fallback scenario, this thing can't go wrong!) and arrogance from too many people at the WMF. Fram (talk) 15:15, 25 February 2016 (UTC)[reply]

I'm not going to claim it was all roses, but I'd argue there is a big difference in accountability and transparency between say, the last year or so pre-transition (where, sure, things were not perfect, but absolutely getting better) and the current situation. Ironholds (talk) 15:25, 25 February 2016 (UTC)[reply]
"the last year or so pre-transition" would be 2013-2014? The period that saw the introduction of VE and Flow, to name just these two? In what way was the accountability any better at that time? What, if anything, was actually "absolutely getting better" in that period? I seem to recall it as one long downhill course, ending one year later with the Superprotect fiasco, and only since then has there been some improvements (coupled with many same old same old problems, ranging from Gather to the WMF Board problems). Fram (talk) 15:33, 25 February 2016 (UTC)[reply]
2014-15, I was thinking of. Again, I'm not saying it was perfect, but I think that a lot of the feedback processes were improving - the proper resourcing of community liaison tasks helped a lot. I'm not expecting us to agree here, mind ;p. Ironholds (talk) 15:52, 25 February 2016 (UTC)[reply]
Community liaison? You mean the PR people/ablative meat-shields you have to get through before dealing with someone who might actually know what is going on?
From the outside in "the last year or so pre-transition" the WMF appeared to be a large part to be drifting while waiting for the new ED. Sure the management may have been better (at least reports suggest it was) but the lack of overall direction combined with the ongoing need to have software "successes" no matter what the cost was a problem. Still thats pretty much what you would expect given how much the ED hunt overan.©Geni (talk) 17:56, 25 February 2016 (UTC)[reply]
Speaking as a former liaison I don't know what productive conversation you expect to have when you open by referring to the other participants as "ablative meat-shields". Again: not saying it was perfect. Not saying it was good. Just saying it was getting better, and has recently got a lot worse. Ironholds (talk) 18:10, 25 February 2016 (UTC)[reply]
What Geni is saying is that the liaison absorbs the flak form the community - until they are worn away or ground down by it.
Now if the culture is such that the liaison is able to be a community champion inside the organisation, then the flak will, probably, abate - and good things will happen. However if the liaison is required to push the organisation's views on the community, and dis-empowered from achieving change on behalf of the community, then the job of liaison is neither pleasant nor useful.
I think consensus is probably that the latter has been more the case recently than the former.
All the best: Rich Farmbrough, 20:23, 25 February 2016 (UTC).[reply]
Some give the impression of not only being required to do so, but to be happy to do so, either because they share the same views or because it is more interesting to please the one that pays you. Some of the people I most respect at the WMF are two community liaisons, but some of the worst (remaining) people at the WMF I have had contact with are two other community liaisons. Good community liaisons trying to ease the tensions when some of the technical people (or usually their bosses) at the WMF have screwed up again are a very good thing; poor community liaisons only increase the distrust and problems between enwiki (or dewiki presumably) and the WMF, as happened recently with the Gather discussion. With VE and Flow, the good and the bad community liaisons provided an uneasy balance there, but it was often amazing to see how people with the same job description in the same (difficult) situation dealt with things in completely opposite ways. Fram (talk) 08:02, 26 February 2016 (UTC)[reply]
2014-2015 was the first year Tretikov "ruled", so to speak, and the year e.g. Erik Moeller left. How was this a "pre-transition" year? Perhaps you should better explain what "transition" you are talking about, and when it happened, as I don't seem to be on the same page. In any case, I have no idea how it was to work at the WMF before Tretikov and how it is now, but I can see that some of the most toxic elements (wrt relations with the communities) have left, that some very poorly thought out and implemented projects have been suspended or completely redirected (Flow and Gather), and that many of the problems we recently had are not directly caused by Tretikov (who is not on the WMF Board, unlike people like Jimbo Wales), apart from probably the Knowledge Engine communication disaster. From my position, many of the grumblings of the older WMF ranks come across as "damn, we no longer have the happy careless job where we could pursue pet projects without any accountability" and not the exact opposite, which you claim to be what's happening. It's very hard to reconcile these two views, and so far you have done little to enlighten us on what really is the problem now. Claiming that before Tretikov, there was more accountability and better community liaisons seems hardly realistic. Fram (talk) 08:02, 26 February 2016 (UTC)[reply]
I appreciate there's an information disparity here, and I'm not able to do much to solve it; I'm bound by an NDA and so can't give my perspectively entirely or hypothesise too wildly.
You may have misunderstood my point; I didn't say that there were better liaisons, I said the liaisons were one of the things improving, and I've felt over the last couple of years that things have remained pretty static on that sort of front (IOW, they are not worse, or better, more..stable).
I agree there have been fewer massive, public communications problems in the last couple of years, mind; I don't think that's because we've stopped making mistakes, I think it's because we've stopped communicating what we're doing. Even inside the organisation, a lot of people don't know what other teams are doing, or find out in the last minute, and a lot of the current projects are not necessarily editor-centric things (an example is the www.wikipedia.org redesigns, which have been tested and played with for a couple of months now but don't exactly have village pump messages plonked down about them). So: don't confuse a lack of public pratfalls for a lack of pratfalls. It could just as easily be a lack of projects, or a lack of transparency about those projects.
The "old guard" argument - well, frankly, most people in the organisation don't get to choose what projects get worked on, and never have. Unless you're a Product Manager or above or a particularly forward designer who can wow people with cool mockups, your ideas tend not to go anywhere. So the idea that people are grumbling because they no longer get to work on whatever they want: well, most of them never have. And if you look at the people who have been publicly expressing issues with the organisation's direction, you'll not find product managers: you'll mostly find people outside of engineering entirely, responsible for things like the Teahouse, and engineers or researchers, some of whom joined after, not before, Lila did.
So I think the idea of a marginalised old guard is a pretty inaccurate one, but also (more importantly) one that's disprovable with public information. Anna Koval had nothing to do with Flow; Siko had nothing to do with Flow; I did in the initial days but sure as hell didn't propose the thing. Neither did Anasuya or Frances. There's clearly something beyond a change in projects to explain the departures. Ironholds (talk) 21:12, 26 February 2016 (UTC)[reply]
Disprovable? People like Moeller and Jorm (and you) are old guard compared to Tretikov, and Moeller and Jorm were some of the people most involved with those failed projects (it isn't really clear what you have done after you have been taken of the Flow project). And you focus quite strongly on "they weren't our projects", but the more important thing in my post was the lack of accountability, which is just what you claim is happening now but from a wikipedia point of view was what happened before. I guess that there were problems at the WMF which didn't really directly affect wikipedia but which were seriously problematic for people like Siko, so I'm not saying that Tretikov was a faultless saint; I lack the information to accurately judge her work in toto. But I note that there were problems at WMF which have seriously been reduced since she was there, and which were at least in part solved by her reshuffling of some departments and getting rid of some people. And one of the things behind these earlier problems was clearly a lack of accountability, which is what you claim was better in the old days. And when I notice some of the people who were part of the problem then now posting about the problems now (not just you, but also and more ironically someone like Moeller), then I can't help but notice that lack of self-awareness and acknowledgement of their role in earlier, bigger failures is still rampant and that little if anything has been learned by some from the Flow, VE, ... disasters. Fram (talk) 07:55, 29 February 2016 (UTC)[reply]
While I can't speak for Siko the reason I'm leaving is that being in the organisation under Lila's leadership was so deeply unpleasant on a personal level that I've advocated HR educate people about common responses to psychological traumas for myself and my coworkers. I work on the Discovery team and I didn't know half the stuff posted about the Knowledge Engine until it was posted, so an argument that transparency is improving holds little water. Absolutely, she did a lot of things well (absent some resourcing problems I really like the internal re-org, for example). And absolutely, Flow and the VE were seriously problematic. But they're totally unrelated to my reason for leaving, and neither excuses nor justifies the chilling absence of discussion and toxic internal culture that marked her tenure. Ironholds (talk) 17:32, 29 February 2016 (UTC)[reply]

Just in case anyone missed it, the title is a reference to Forward the Foundation. --Izno (talk) 19:18, 25 February 2016 (UTC)[reply]

What do you think of The Signpost? Share your feedback.