[Talk-GB] [Osmf-talk] Live OSM discussion in ~45 minutes (7.30pm UK time)

Frederik Ramm frederik at remote.org
Thu Jul 27 11:42:57 UTC 2017


Hi,

On 26.07.2017 23:58, Ilya Zverev wrote:
> While I was dismissive of her arguments four years ago, now I see that 
> all of her points were valid, and are still valid. 

I think that it is possible for an insider of OpenStreetMap to look at
Monica's work and see some valid points in there. But try to switch off
your background knowledge and look at her work. What sticks with you is
something like (quoting from a 3rd party web site that introduces the talk):

"She looks specifically at the case of how "childcare" was not approved
as map category within OpenStreetMap."

This comes from her work massively exaggerating the issue for effect,
and being extremely sloppy with OSM background research.

Reviewing her talk, the OSM part begins with her showing group photos of
past SotM conferences claiming "these are all men". Which clearly isn't
true (you just have to zoom in on the picture). Maybe I'm putting the
bar to high by measuring this with the "science" yardstick, but it feels
wrong to me. Do you want future scientific papers to quote "according to
<source>, no women have attended large OSM gatherings before 2013"?
Because that's what she says.

She then goes on to equate the number of different values in the
"amenity" key space with the importance of something (arguing that
because you have different amenity values for bars and pubs it is clear
that this is an important distinction); this is not tenable as just
slightly more research would have shown, there is no correlation between
the importance of something and the number of different key values in
the amenity space.

She then claims that "amenity=swingerclub" was the (1) most recently (2)
accepted (3) voted on (4) approved amenity - not a single one of the
numbered points is correct as far as I can see from the Wiki history
(but I invite readers to double check, I might have missed some page
renamings?).

Going forward, she gives listeners the impression that a successful tag
proposal was a requirement for being able to tag features, which is
plain wrong. At the very least, a non-misleading, non-sensationalist
presentation would have to mention that

(a) anyone can tag anything they find important,
(b) this *may* be influenced by editor presets (which didn't feature
swingerclubs at the time and don't now)
(c) what appears on the *map* is a different issue again, and
swingerclubs weren't on the map then and aren't now.

(As a tiny nod towards the actual subject of this thread, point "b" was
addressed in Andrew Hall'S "Wikimedia Research Showcase" presentation.)

She then goes on to discuss the amenity=childcare proposal, which had
been voted down in 2011. As you can see from
http://wiki.openstreetmap.org/w/index.php?title=Proposed_features/childcare&oldid=789581
the proposal itself had been framed sloppily; it claimed to be
applicable to all age groups ("Example: 0-6") but didn't explain in how
far it was meant to replace the existing amenity=kindergarten or just be
for after-school/after-kindergarten care. A total of 9 people voted
against the proposal; most because of this technicality, and two because
they would have preferred amenity=social_facility.

Did those 9 people vote because they "were ignorant" or "didn't care"?
Maybe, but in my eyes the fault lies just as much with the proposal
itself; the confusion with "kindergarten" and the question of whether
"social_facility" would not be better didn't come from nowhere and they
should have been addressed, the proposal refined, and brought to vote in
a better shape.

Do voters have a duty to pass a badly done proposal when it is for a
good thing? Or are they right to shoot down a badly written proposal?
The "post mortem" on the page says "Voters have either not grasped this,
or have considered the fact of overlap sufficient to reject the proposal
without taking the time to propose a proper alternative." - but is it
the voter's responsibility to propose a proper alternative?

Monica Stephens makes the proposal sound less confusing in her talk -
she explicitly claims the proposal was for childcare for kids that are
"not of kindergarten age", when the proposal explicitly lists "0-6" as a
valid age example. So her listeners will not be able to understand the
confusion.

She then says "OpenStreetMap is a democratic society where people vote
on which amenities will appear on the base map" which is, of course,
wrong in several ways (see my a/b/c list above).

In criticising the "against" voters, she picks out a few that have
spelling mistakes and adds a prominent "[sic]" after each "refered" or
"usefull" - something that may be scientifically correct but speaks of a
desire to belittle these people for whom English is not their first
language. She doesn't quote any of the "against" votes that say that the
overlap needs to be explained, she only quotes those who believe the new
thing is identical to kindergarten. And the correct tally of 9 "no" and
5 "yes" votes becomes, in her talk, "voting ended and was 15 to 4". Just
sloppy?

She then proceeds with some anti-German slurs, claiming that "all but
3... or 5 ... of the brothels in OpenStreetMap are in Germany, the rest
are in Amsterdam". Now this "American values are the right values"
attitude is something I could go on about for a while (are more children
harmed by brothels or by guns) but I'll save that for another time; I
have counted the objects tagged amenity=brothel in OSM at the beginning
of 2012 and found 510 in Germany and a total of 825 world-wide, so I
don't know how she counted but apparently it wasn't all that important
to her. Just a little harmless fun at the expense of all those German
and Dutch perverts, right, let's all have a good laugh? At the same time
there were 16,693 amenity=kindergarten and 51 amenity=baby_hatch mapped
in Germany, numbers which might have served to put the whole thing into
perspective - sadly her listeners are denied that piece of information
which a responsibly scientist should have shared.

She concludes that "OSM is dominated by male contributions" (which is
correct) "and excludes the other 1/2" (which I'd argue with). She says:
"Women cannot really map their local community; their local information
is particularly excluded from this base map and from what features are
(inaudible) in OpenStreetMap."

This is a very broad, I'd almost say outrageous, claim, and not at all
supported by the evidence she has provided, even if that evidence were
factual.

She proceeds to claim that "... all of these options for child-care,
day-care, have failed in OpenStreetMap, continually". Again, not at all
supported by any evidence. She again claims that "swingerclub was
democratically approved without a single opposing vote in 2012", when
indeed no vote on that tag has taken place, ever.

So, to close this off, Ilya I think you are doing OpenStreetMap a huge
disservice by taking a talk that is so full of false claims, so biased
and misleading, and publicly say that "all of the points are valid".

There is a valid point in that it would be desirable to achieve gender
parity in OpenStreetMap and that this would make for better discussions,
better results, a better map. But almost every other point made in that
talk is at least exaggerated for effect, if not blatantly false.

I'm afraid I have now wasted two hours of my life doing what Christoph
warned of, namely heightening the visibility of Monica's work by trying
to point out the flaws in it, and I agree it would be nice if we could
ensure that if researches criticize OSM in the future - and there's
certainly a lot to criticize - they at least get their facts right.

I want to live in a society where everyone is free to say their opinion,
but I don't want to live in a society where everyone can claim facts
that are simply and demonstratively wrong and not be called out for it.
There's too much of that out there already.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00'09" E008°23'33"



More information about the Talk-GB mailing list