I’ve been mapping each field separately, but joining the nodes at the edges. Then I’ve been using barrier=hedge to show boundaries. Is this better than the first example, where a gap is left between each field?
I think first its important to say that there are no right and wrong ways: in usually depends on what one’s particular interest in mapping.
I have mapped farmland and hedges by drawing large areas of landuse and then adding the hedges as individual features. This is fine, if, as it is for me, it’s the hedges that are of interest. Amongst other things I’d like to see maps rendered so that different types of field edge are shown (like in Wainwright’s books), but also to do things like compute the length of hedgerow in a given area. Here’s an example of what I’ve done:
I personally prefer methods which share nodes, i.e., the edges of the fields are the midlines of the hedges. Other approaches look nice in current rendered views, but may be misleading in other applications. Just mapping the fields shows the field edges in a render, but is not useful for instance in routing, where the application will not know that hedges etc exist.
IMO either of the 2 ways of providing a continuous block of farmland are better than splitting into fields and leaving a gap. It looks terrible and what is the landuse between the fields if it’s not “farm”?
The advantage of 1 area per field is that it can show a field boundary that is not necessarily a hedge, but I wouldn’t know which way is best when you come to add tracks, gates, cattlegrids etc. Perhaps they are both the same.