[OSM-talk] Commenting and thumbs up/down feature for changesets

Steve Bennett stevagewp at gmail.com
Mon Jul 25 08:27:41 BST 2011

On Wed, Jul 13, 2011 at 11:44 PM, Frederik Ramm <frederik at remote.org> wrote:
> I was thinking of voting up or down contributions, and yes, this could also
> lead to league tables that identify people with consistently problematic
> edits; but that would not be because of who they are, but because of what
> they do. Maybe I am the only one seeing a difference here but personally I
> have absolutely zero problem in saying something like "this person has
> consistently made edits that others in the project found sub-standard". This
> has nothing to do with hurling insults at anybody.

To continue drawing lessons from Wikipedia, there have been quite a
few attempts to build "trust metrics" into Wikipedia, evaluating the
likely value of a given change, based on previous rates of reversion
of changes made by that editor, and other factors. The change is then
displayed in a different colour, accordingly. Research papers have
been written, but none of these features have ever made it into the
production system. Why? I'm not sure exactly, but I think it basically
doesn't offer enough value: what ultimately matters is the content of
the change, not who made it.

Two thinks that Wikipedia has that OSM lacks, are good visibility of
recent changes (just click "view history"), and localised forums (talk
page). Although it's possible (if difficult) to get the history of a
given object in OSM, I don't know of any easy way to get a sense of
the recent history of an area. Undoubtedly there are third party
websites, but anything built into openstreetmap.org? And if you do
find an issue that you want to discuss, your options are only to email
the person privately, or to raise it on the appropriate talk-*@
mailing list, and hope they're listening.

Personally, I would have trouble marking many changesets "-1" without
doing extensive research. But there are a few where I've queried the
author, and in some cases found explanations that weren't obvious at


More information about the talk mailing list