Thanks for the comments. Implementing year/month/day is easy enough, but I thought that, for consistency, I should also do '1847 May 22', and that would take a bit of extra work. Frankly, I was ready to move on to another project, so I punted.

What I wanted to do was to draw Wolfram's attention to the issues with DateObject[] with a string input. It's a mess once you get beyond the simple dates since 1800 or so. For example, I found that if you do DateObject["10/25/1415", CalendarType -> "Julian"], the function converts the date, which it assumes is according to the Gregorian calendar, and returns a date some nine days earlier. This is just wrong. It is my hope that there will be a revision to DateObject[] that will incorporate my code, with improvements.

I will try to revise the package to include the year/month/day and replace the existing package, in my copious spare time.

Hi George, I was very surprised and pleased to see your post on the topic of historical dates (and by extension, historical places). Two important and complex topics. I've so far only skimmed over your post and notebook but will get back to you with some detailed comments soon.

I've a particular interest in the topic since I was involved in an effort a few years ago to describe the research infrastructure needed to better tackle dates and places in particular for early modern scholars. You can read about this here (open access monograph). This in turn led to some (unfortunately incomplete – we ran out of time and funding) exploratory work on EM Dates and EM Places.

Though I've long maintained a dabbling interest in the WL I've never given myself the time to properly learn it and move beyond the beginner stage. One of my ideas was indeed to go to the Summer School with a plan to create a better historical date resource – but now I'm very glad to see that someone got there first!

best regards, Arno