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

[Tobias Knerr]

Ulf Lamping wrote:
> There are people who actively watch out "their area" what changes there.
> That's fine and valueable. But IMHO it's *their job* to make sense of
> the changes, not the mappers job.
[...]
>> Don't be fooled; the small changeset comment that you enter when
>> uploading stuff *will* be read by many people.
> 
> I don't think so. Do you have numbers?

Even if only a single person tries to understand your changesets, it
will take them far longer to figure out what you have done than it would
take you to write it down.

If you think that watching one's area is a valuable activity, it makes
sense to add changeset comments. With tools like OWL, it's now actually
feasible to look at edits in an area - but if mappers don't write down
what they have done, it would take a lot of time to understand the edits.

There's another situation when changeset comments are useful, and that's
when I try to fix broken data (and find out why they are broken), as
that involves browsing though the elements' history. With changeset
comments, it's easier to identify suspects among the edits.

> First of all, you probably need better diff tools (I mentioned that
> before :-), not better changeset comments ...

There are excellent diff tools for source code. That doesn't stop people
from adding version control messages. Of course, I'd love better diff
tools, too, but they don't operate on the same level of abstraction as
changeset comments.


Finally, I want to point out that I don't want rules forcing people to
add changeset comments. If people resent the act of adding the comments,
their comments tend to be nonsense anyway.
I just want to say that, yes, there are people who read those comments,
and it would be great if more mappers added them - voluntarily, because
they have decided that it is a useful feature.

Tobias Knerr