[HOT] Gentle grump
ray at ganymede.org
Fri Mar 6 20:07:39 UTC 2015
On Tue, 3 Mar 2015 15:12:21 -0500
john whelan <jwhelan0112 at gmail.com> wrote:
> Just for the heck of it I ran JOSM validation on a tile I was mapping
> before touching it. It turned up duplicate buildings, crossed
> buildings, lots of highways separated by a few inches etc.
> Do we need an idiot guide? A sort of this is how to provide the
> maximum benefit for the least effort.
Speaking as an idiot, I would say that the answer to this is yes.
Perhaps you think I jest....
> Mine would probably run along the lines of for Africa the convention
> is only the following values of highways are used for minor highways:
> path, track, unclassified, use highway=road if you are uncertain.
> Someone will probably have tagged the secondary and primary highways.
> If possible use JOSM especially for buildings. Please map buildings
> as building=yes do not assume it is a house.
As a 2-3 times per week mapper (who wishes I could do more), it can get
frustrating. Lots of projects point to the Africa roads page but that
page is hard to interpret for any particular context. There is a lot of
And I hate to say it but I use ID and it drives me nuts. This may be
from browser/js/platform issues. I am using Firefox 36.0 on Ubuntu
14.04 LTS. But I have looked at JOSM and it is somewhat bewildering and
I have no idea how long it would take to get over the first humps of the
learning curve. For now, my annoyances with ID are tolerable.
If one was able to look at a task and see what tags where being used
and how often within just that task, this might help the "African
> People use maps to get from one place to another, if the highways are
> joined up then routing software such as comes as part of OSMAND can be
> used. Look for highways around settlements that connect to other
> <Crtl><arrow> in JOSM will navigate vertically or horizontally making
> scanning easier.
I should see if there is a cheat sheet for JOSM. It would be nice to
know what control-shift-elbox-J does and all that. Of course, these may
be platform specific (eg Windows keys vs Linux keys vs MacOS X keys).
> I assume that most of these errors have crept in because JOSM
> validation was not used. I suspect that the immediate feedback from
> JOSM might assist our less skilled mappers to improve their skills.
> Cheerio John
There needs to be validation on input and obviously both ID and JOSM do
some, but can validation be done on the server? This would be better,
especially if the results can be communicated to users. A HOT task could
have a "Validations" tab. I, for one, would like to see the things that
have been already fixed in data in that task. It would let me know when
there are things not to do. If I am going to make a mistake within a
task's maps, it is at least a bit likely that others will make or
have made similar mistakes in the same context.
Again, seeing the phrase "JOSM might assist our less skilled mappers",
I have to wonder what you are thinking about here. Any sentence with
both "JOSM" and "less skilled mappers" in it is going to lead to bad
things. JOSM might be easier than it is, but I am not even very sure of
that. Sometimes complex tasks require complex tools. One just hopes
that there are options between the "very-simple-but-also-brain-dead"
tool and the "amazingly-powerful-but-shockingly-unintuitive" tool. I am
not saying that this is what JSOM and ID are, but hopefully you see my
So, grump back at ya. :-)
cheers - ray
More information about the HOT