[OSM-talk] Funding of three infrastructure projects : Nominatim, osm2pgsql, Potlatch 2

Matthew Woehlke mwoehlke.floss at gmail.com
Mon Aug 3 14:14:13 UTC 2020

On 02/08/2020 06.05, Simon Poole wrote:
> Extending this a bit further, you could just as well say, given that all
> current and actively maintained general purpose editors require 1-2
> FTEs, the OSMF should simply block all non-iD editors and tell the
> developers to either work on iD or go home.

For OSMF *funding* purposes this might happen, but telling volunteers 
what they should or should not volunteer to work on should be a hard no-go.

> iD is branching out in to more and more niches, reducing the
> breathing space for anything else massively and other editor use has 
> effectively been stagnating for a long time. While people will 
> automatically try to start listing special use cases that can "only"
> be done with editor XX, the problem is that these are special cases
> and unlikely to be worth spending a couple of $100k on per year
> (virtually or real) for the small number of users that will remain as
> iD gains more and more features.

There are a few things iD does "better" than JOSM¹, but it is *far* from 
feature parity... and one use case which I consider *absolutely 
essential* before it could be considered a JOSM replacement is the 
ability to load and save local files (notably including shapefiles and 
geotiffs) and work on non-OSM layers... and I'm not sure that will ever 
happen. JOSM isn't "really" an OSM editor, it's a GIS tool that 
"happens" to have really good OSM integration. (Note also that these 
features are *mandatory* for doing imports from other GIS data.)

I've been using JOSM a lot lately, and AFAIK iD is quite some ways from 
matching even some of its more "basic" functionality. Angle constrained 
ways, lane view, way smoothing features, ability to mirror content 
(symmetry), and more. Relations are *much* easier in JOSM. Heck, just 
*selecting things* is much easier.

I'm not saying iD is *bad*. It's a very nice editor *for its 
capabilities*. It's great for making *small* changes or introducing 
someone to OSM editing... but there are a lot of use cases still where 
JOSM is just a far superior tool. Maybe in *5-10* years that will 
change, but I'm not going to hold my breath on it overtaking JOSM in 1-2.

(¹ iD can 'square up' individual nodes and does a passable job with 
*mostly* orthogonal shapes with the odd 45° angle. There are ways to 
work with those in JOSM, but generally speaking if you try to square a 
shape with a single 'wild' node, JOSM turns the whole thing into a hot 


More information about the talk mailing list