5th
So I had this idea. The stats pages are great but they show you the stats for a lat/lng bounding box, meaning that my stats for an area will be different from your stats for it since we will have different view sizes, different origins and different bounding boxes. So the idea was that for the stats page, rather than use a bounding box, I’d reverse geocode the origin, get the location as a text string of the country and administrative area (posh term for town) and use that to find all the attractions (and thus stats) within that area.
Sounded like a sound plan, until I tried to reverse geocode all 44,000 attractions in the database and hit on Google’s usage quota. Not that that in itself is a problem, but thinking of the future, I don’t want to build something that relies on using a service that might run out of quota, cos then no one gets their stats until the next day.
So that stats are staying as they are for now. I might make it so that the bounding boxes snap to a larger grid or something in the future, I’ll have to think about it some more.
I am however still going to reverse geocode all of the attractions, just over a number of days to keep Google happy. This data will be stored and used to display on each attractions page just so that you know where it is and so that you don’t confuse all the St Mary’s Churches that there are out there (and I’m sure it’ll have some other future uses).
