On 18/08/13 22:06, Graeme Wall wrote: > On 18/08/2013 20:30, Richard Smith wrote: >> On 18/08/13 20:24, Graeme Wall wrote: >> >>> 1900s plural is the decade, just as 1920s plural is the decade. >> >> You may use that convention, but a lot of people do use 1900s to mean >> the century, so the ambiguity is still there. > > Not something I've noticed, usually if you are dealing in centuries you > refer to the 19th or 20th or whatever. That's what I do too. (Though even that is open to a certain degree of ambiguity -- in the ISO 8601 format for centuries, "19" denotes the 20th century. Fortunately the ISO 8061 century format is pretty rarely used.) > Just goes to show you have to be very careful in making sure your date > references are unambiguous. Absolutely. The whole subject is riddled with ambiguities. Just to take one other example, does 1 Jan 1700 refer to the day after 31 Dec 1699, or the day after 31 Dec 1700? I rather suspect my database has a mixture of styles, even though my convention is to rewrite dates to be in the former style. Richard

On 18/08/2013 22:44, Richard Smith wrote: > On 18/08/13 22:06, Graeme Wall wrote: >> On 18/08/2013 20:30, Richard Smith wrote: >>> On 18/08/13 20:24, Graeme Wall wrote: >>> >>>> 1900s plural is the decade, just as 1920s plural is the decade. >>> >>> You may use that convention, but a lot of people do use 1900s to mean >>> the century, so the ambiguity is still there. >> >> Not something I've noticed, usually if you are dealing in centuries you >> refer to the 19th or 20th or whatever. > > That's what I do too. (Though even that is open to a certain degree of > ambiguity -- in the ISO 8601 format for centuries, "19" denotes the 20th > century. Fortunately the ISO 8061 century format is pretty rarely used.) > >> Just goes to show you have to be very careful in making sure your date >> references are unambiguous. > > Absolutely. The whole subject is riddled with ambiguities. Just to > take one other example, does 1 Jan 1700 refer to the day after 31 Dec > 1699, or the day after 31 Dec 1700? I rather suspect my database has a > mixture of styles, even though my convention is to rewrite dates to be > in the former style. > Convention is to write 1699/00 or 1700/01 as appropriate though, like you, I convert to new date format to avoid the ambiguity. Mind you I haven't actually got back that far on my primary line so the situation only arises on the wife's ancestors. I'm currently stuck in Berkshire on the mid 18th century trying to work out where a miller came from who could afford the lease on a 3 wheel mill and a large farm. -- Graeme Wall This account not read, substitute trains for rail. Railway Miscellany at <http://www.greywall.demon.co.uk/rail>