Evaluate latest-changes query

[nrenner]

latest-changes seems to be faster for area queries than achavi. Evaluate if this query produces the same result and indeed is faster (and why - maybe because not using changed?) and could be used instead.

[adiff:"' + last_week + '"][bbox:' + bbox + '][out:xml][timeout:22];
way->.ways;
(
    .ways >;
    node;
);
out meta;
.ways out geom meta;

https://github.com/tyrasd/latest-changes/blob/fb878927d2681777f619123f0a8ad3a05d6cdf8c/js/site.js#L47

[mmd-osm]

changed is more expensive, as it has to evaluate every change in the given time frame t₀..t₁. The query above simply returns the difference of the query, once executed at t₀, and again at t₁. If your time frame is rather large, this method might miss some changes that have a zero net effect (e.g. an edit which gets reverted very quickly).