[OSM-talk] Thoughts on OSM design, and looking forward and back
david at frankieandshadow.com
Wed Feb 24 16:19:45 GMT 2010
I'd like to say a few words on the home page and editor.
1. Home Page: while I think Steve's proposal addresses some of the
criticisms of the way the home page functions, I don't think it takes a
holistic view of the project. What someone coming to it will initially
see is essentially a "me too" for Google maps: it offers a service not a
I do think it is essential to have _a_ map on the home page, but I don't
think it need take up the whole page. It need only be a representative
map, and for the first time visitor well zoomed out area, so detail is
low and it doesn't need to be that big. Once someone searches, clicks on
the map, drags, presses the relevant button or whatever, we could go to
a page like Steve's where you can also get search results and other
direct services like export. But I think on the home page we would do
better to have a smaller map and more information visible without
clicking tabs links or buttons
- a (brief) introduction to the project and link to more,
- how to get involved + link,
- especially links to all the services, products, projects and
innovative ways people have based things around the project that aren't
hosted on the site as well as those that are,
- contact info for who can help provide services based around OSM (or at
least an indication that there are such people and a link to where you
can find out about them),
- and space for a prominent "Report a problem" button.
At the moment, Mapnik rendering *is* OpenStreetMap as far as the casual
visitor is concerned, and I'd rather see that dominance reduced (not
taken away, as it is a really good showcase for the outcome of the
project, but it is only one), not emphasised even more.
2. Editor: Potlatch (and JOSM) address a different market from a
feedback system, OpenStreetBugs or whatever. The latter only works if
there is enough context on the map to make an observation about the
content. If you're starting on virgin territory, that's not nearly
enough. There's a place for both kinds and both kinds need to be
improved. I find it hard to envisage a system for near-virgin territory
editing which doesn't need at least some of the kind of graphics
manipulation you need in products like Adobe Illustrator; but that's far
too hard for someone who just notices an error in a well mapped area, so
an alternative "point and say" type interface is definitely needed for
these people. Off the main stage, I think it would be helpful for those
who are acting on the information such a system provides to have a means
of seeing and tracking it, which can be more complex than the reporting UI.
OpenStreetBugs corrections in my area seem to fall into three
categories: 1. "my street/village is not there" which is usually not
helpful as it hasn't been surveyed yet, 2. incorrect changes: someone
goes down a street every day and thinks the map is wrong. But they
haven't actually gone and looked for the purpose, or they don't
understand the signs, 3. helpful, valuable corrections. Sometimes 2 and
3 are hard to distinguish and need a visit. If someone new does make a
change in my area, I usually make a point of checking it if I'm doubtful
about it - and many times it does turn out it was my error, but very
often not: the original survey was looking in detail and that often
beats someone's casual memory.
But then OSB is a rarely used tool as no one really knows it is there.
I also think a feedback system needs at least the option of someone
providing a contact or for them to receive info back - either a thank
you, we've corrected the problem (so they get a nice fuizzy feeling of
contribution) and/or a question to clarify their contribution (which
I've needed more often than not for OSB contributions but have no way to
do for anonymous entries, which is most because that's the default).
Formal registration is way OTT though.
More information about the talk