There is some discussion of Gardner's statements in this WR thread, but rather than further diverting the discussion of the "holy shit slide", I thought I would start a new thread. Feel free to merge, if desired.

On one of the slides Gardner showed in Hamburg (and presented to Wikimedia UK) there is a timeline of the image filter. It says, in part:I believe that both of these statements are demonstrably false and self-serving.

Here is what the WMF board asked Gardner to do:Note that the Board does not suggest that users will need to opt in to such a filter. I have bolded a few phrases which suggest that this is an opt-out system (i.e., it is in effect unless one opts out of it) not, as Gardner suggests, that users will need to "opt out of seeing images". The proposal presented to the community does not satisfy the requirements as set out by the board, but that is unsurprising.

What Gardner refers to as a "community vote" in her slide was a disingenuous and confusing "referendum". It was neither a referendum about having an image filter nor a community vote to assess support for an image filter. Its apparent aim was to determine the importance of various features in a piece of software that had clearly been designed without any community input. As stated in the preamble of the "referendum":It is entirely disingenuous to make any claim of support or rejection on a poll which states that the respondent's opinion about having an image filter is irrelevant to its implementation. The inclusion of the question simply caused confusion and much unnecessary discussion.

(My apologies for the earlier misspelling of Gardner, now corrected.)

Wow, so you're saying that Sue Gardner is lying?

The image filter was always conceived as an opt-in filter (i.e. you have to opt in to filtering images, and the default display includes all images). That was also the recommendation from the Harris report.

The referendum didn't ask whether people thought it would be a good idea because the board had already decided in late May that it wanted to do this (controversial content resolution). The referendum only asked whether people thought it was an important feature to offer readers, and most did say yes to that. There were some geographical variations, as can be seen in Appendix 2 of this page on Meta.

Here's the relevant text from the Harris report: It is not explicitly stated that the filter is either opt-in or opt-out. The comparison to Google's search options is worth noting, since Google offers "safe search" by default. It can be turned off with a click, and the preference will be saved if one is logged in to any of Google's services.

Harris stated that the report would present an analysis of common practice on other websites, but failed to do so. It is interesting to note that in the same slide that provided the quotes from Gardner, she has a footnote stating that "To develop his report, Robert Harris gathered together 40 studies, white papers and other research from 15 countries around the world originating in 15 countries and covering attitudes, regulations and practices in 76 countries". I am not aware of an analysis of this research, nor even so much as a bibliography that has been made public.

I'm sure you will agree that asking people to rate the importance of a feature is not at all the same thing as asking them if they support the adoption of a system that would incorporate such a feature.

I don't think you even need to log in. Google preferences can be saved in a cookie.