[OSM-talk] Barriers of Entry
emacsen at gmail.com
Wed Sep 14 19:07:30 BST 2011
I respectfully disagree with most of your email, comments inline:
On Wed, Sep 14, 2011 at 3:24 AM, Frederik Ramm <frederik at remote.org> wrote:
> There are barriers of
> entry in the form of entrance exams at places like universities, with the
> aim of assessing the likelihood of someone succeeding in their studies.
I have a bit of contention with this but I'll accept the fact that
there is a correlation between entrance exams and academic success,
but I will rebut this premise later in the mail.
> of course there are job interviews, where employers sometimes raise the
> barrier of entry so high that only one in 1000 can pass.
This is a simple premise that can be rebutted in this context. The
issue of an employer is that of limited resources.
When the resources are large and free, the problems become different.
> Barriers of entry are not always a bad thing; they might keep you and from
> entering a tunnel for which your vehicle is too wide, or they might make it
> less likely for you to spend years at university with little chance to earn
> a degree.
The argument that a university should have entrance exams to keep
people out in order to prevent them from failing out has firstly, a
misunderstanding of the education system, and a misunderstanding of
education in general.
For discussions of universities, I can only speak about the US. In the
US, institutions do have cutoffs for entrance, and will drop the
students they don't feel would "make the cut", but they do not then
automatically take the top N highest percent. What they do instead is
to look for a wide diversity in the students, with the intent that a
university provides oppotunity to a great number of students, and the
understanding that diversity builds
Maybe this is different in Europe, but in the US, this is how things are.
Now I want to address testing in general, and why testing itself is
flawed, especially in how it relates to OSM (since this is where the
mail ultimately leads.
The purpose of education is to educate, and to "build citizens", not
merely blast knowledge out and see who catches it. This is why we are
seeing programs which test different teaching methods, which show that
through different ways of approaching the material, we see different
results. Studies are showing that by using some teaching methods,
those students who did poorly previously now excel.
Knowing this to be true, educators can reform the curriculum to be
effective to different groups. How this relates to OSM later on in
> In OpenStreetMap, people sometimes point to our barriers of entry and
> blindly claim that they must be bad for us. The main page not welcoming
> enough, the editor too difficult, the path to signup too cumbersome, and on
> and on.
Some "barriers to entry" are not barriers to entry, but just barriers.
Using your analogy of a school, if we have a school where all students
walk to school, perhaps we should not place this school on the top of
Similarly if the front page is unfriendly, we should not say "Ah, well
that's the cost you pay for trying to use OSM", but rather work with
people to fix it.
> Now some of this might be true and I don't want to keep people from
> improving the overall OSM user experience.
This seems like you know your argument is faulty and reacting accordingly.
> However one has to keep in mind that what we're doing here is, and will
> always be, more complex than, say, taking your dog for a walk.
It depends on the task. Let's take the example of not walking your
dog, but baking a cake.
If I said to you "Bake a cake", you could perhaps do it. It might
require some instruction, but chances are you could do it.
Now let's say that you had to start from scratch. That means you need
to grow your own wheat, mill it, raise your own chickens, harvest and
process the sugar, etc. Baking a cake is hard!
But we don't do that. We simplify parts of the task which aren't
essential to the task itself.
And some people even use cake mix, further reducing the work required
in baking a cake.
Some geographic tasks are hard, but we can simplify them to the
components that people need to focus on, and make them easy.
Examples can be provided upon request.
> For OpenStreetMap, we don't gain anything from making it look like
> participating in OSM was the easiest thing in the world. You have to possess
> some skills, and you have to be willing to engage with other people, or else
> your contribution is unlikely to be pleasant for you or the project.
I feel like this is a reaction to something that you're not pointing
out except in the abstract.
Who will argue against community involvement? Not me. But I may say
that not every OSMer needs to be participating on the tagging list,
and can safely use the builtins in their editor.
> part in OpenStreetMap requires more sophistication than, say, tweeting about
> where you're having dinner.
I think a Foursquare-type app would be very useful. It would give us
some additional data we don't have today, such as a confidence level.
If everyone is eating at a restaurant every day, and somehow we could
capture that, it would let us know that the data is not only there in
OSM, but accurate today!
> A barrier of entry that discourages those who unlikely to make
> an edit without breaking 10 things, or those who can make an edit but are
> unlikely to answer a question from another community member about it, could
> be a healthy barrier.
So you want to add moderation? That certainly doesn't sound like you...
> Before someone misunderstands that, I'm not saying we should introduce some
> kind of entry exams or so. Anyone who wants to participate should be allowed
> to. But it should not be our aim to make the whole world want to participate
> because only a fraction of them can.
If you aren't arguing for moderation (which it seems like you are),
then maybe what you're saying is:
"I'm concerned about uneducated newbies messing up OSM, so we need
ways to prevent that."
and my response is:
There are ways other than barriers to do this. One way is that some
folks monitor all changes in OSM and fix obvious mistakes. This is
something that already happens, today.
Secondly, you're arguing for better user education. I know some future
version of Potlatch will have a tutorial mode, and I'm going to
continue to work on the OSM videos I made (and make more) and some
folks are going to do documentation.
Thirdly, maybe (and I'm not convinced of this), you might suggest
another DB we could use. So in the foursquare-type clone, it could be
in another DB, usable to OSM as another layer from which editors could
work, rather than being in OSM itself...
More information about the talk