[OSM-talk] A plea for meaning ful changeset comments

[Tobias Knerr]

Frederik Ramm wrote:
> I could imagine dropping the mandatory changeset comment, but when left
> empty, display a pop-up that explains why changeset comments are
> important and ask the user to reconsider. (Indeed that dialog could be
> shown whenever the changeset comment is less than 15 characters or so.)
> And of course that dialog must not have a "don't ask me again" feature.

I believe that people will only provide truly useful changeset comments
if they do so voluntarily. Not to mention that some react badly to
rules, and will rebel against something they *would* have done
voluntarily when they are forced to do it.

Therefore, I think that annoying people with permanent dialogs is highly
counter-productive.

Instead, I suggest the following course of action:
- remove requirement to fill in the changeset field from editors
- improve the tools that *use* the changeset comments

Why? Because mappers who regularly use features like edit histories
themselves will actually experience why changeset comments are useful.

For example, getting rid of those "big" edits from the history on
osm.org would improve the usefulness, and thus acceptance, of changeset
comments far more than any mailing list thread could.

Tobias Knerr