[krisoft]

> That is correct, but what would that help anyone in general (for any historic reporting)?

The question is rather what are the error bars with these observations. When someone says "my car broke down last week" everyone understand that that is not a precise timestamp. On the other hand if you hear "according to the telemetry the crankshaft seized at 1696606533 unix timestamp." you know that they are talking about a very precise moment in time.

The problem comes if you take the first kind of description and convert it to a unix timestamp you imply precision where there was none found originally.

So when the wikipedia entry says "According to Songshi, the official history of the Song Dynasty (sections 56 and 461), the star seen on May 1, 1006, appeared to the south of constellation Di, between Lupus and Centaurus." Do they mean that Songshi wrote down the date according to his local convention which can be converted with a high confidence to our current date system as "May 1, 1006"? Or did they just erroneously implied more accuracy than what they have?