The first Jermory survey is under way. It is a real churchyard — All Saints at Besthorpe, in Norfolk — and it is being recorded plot by plot, stone by stone, by the people who look after the ground and by us alongside them. There is no formal parish contract behind this. It is an informal arrangement with the volunteers who keep the churchyard, and that suits the stage the work is at. We wanted to build against a site before we tried to build for one.
The choice early on was between designing the tool from a list of features we imagined a custodian would want, or taking a single ground and letting its awkwardness set the requirements. Besthorpe is the second option. Most of what follows is something the churchyard asked of the software before we thought to ask it ourselves.
Putting the ground on a map
Georeferencing sounds like a solved problem until you do it on the ground. A churchyard is not a blank rectangle. It has a lychgate, a path worn off its original line, mature trees that break a clean satellite fix, and a boundary wall that is not quite where the old plans put it. We set out survey control — fixed points we could measure from and come back to — and tied the aerial imagery to those points rather than trusting the imagery on its own. The gap between a careful fix and a lazy one is the gap between a family standing at the right grave and standing one row over.
That work fed straight back into the tool. We added a way to hold several reference points at once and see how far the imagery has to stretch to meet them, so whoever is doing the survey can tell when a fix is honest and when it is being forced.
Boundaries that follow no grid
Burial grounds are not laid out like street plans. In the older part of Besthorpe the plots curve with the path, share edges at odd angles, and in a few places overlap in the record because two burials were once fitted into a space meant for one. Drawing those boundaries meant tracing what is actually on the ground — kerb sets, changes of level, the line where the mower stops — rather than dropping a rectangle and moving on. The plot editor came out of it slower and more forgiving: you can follow an irregular edge point by point, and correct it later without starting the shape again.
Reading the stones before another winter
The other pressure is time. A lot of the lettering at Besthorpe is soft limestone, and every frost lifts a little more of it away. We are photographing and transcribing the inscriptions now, while they can still be read, and keeping each photograph beside its text so a later reader can check the transcription against the stone. Where a line has already gone, we record that it has gone rather than guess at it.
None of this is finished. We have not put a number on the ground yet, because the count keeps moving as plots are merged and split, and a figure quoted now would only have to be corrected later. What we can say is that the survey is real, it is happening, and it is teaching us what a burial ground actually needs from a tool. What we learn here shapes what every other ground gets.