Frequent Update

In this page the process of frequent update of the OSM-GB dataset according to the OSM frequent updates is explained.

The frequent OSM updates makes new dump files available in Planet OSM website, as well as making diffs updates. In OSM-GB, the updating process needs to compare three datasets: The previous OSM dataset, the current OSM-GB dataset, and the updated version. During a last run of the OSM-GB processing algorithm, most of the features may have been kept unchanged, but some others are either flagged or debugged, even some may be removed. The same features may have been kept unchanged, changed or removed in the new version of OSM. Some new features may also been added in the new version. The problem is how to match or synchronize the feature differences between the three and to do the frequent update on the OSM-GB.

Using the diffs is the resource-saving alternative to reloading the whole OSM and re-applying the algorithm on it. While we try to make the result of diffs update as close as the result if it would be applied on the whole new maps, those two results may not always be equivalent, which is the cost of having a quicker update workflow. That is why a less-frequent whole-update is suggested anyway.

OSM is the source of crowd contribution to the map, so it is important to make sure that the changes in the new version has the priority to be reflected in OSM-GB than the changes made by the previous round of the OSM-GB algorithm. However, after the changes are reflected, the algorithm shall be re-applied to the new features when applicable. In some cases, applying the rules to the "adjecent features" is also recommended.

The following scenarios can be considered: