If you start with 2 and would like to arrive at 5, you'll have to add 3. This is a solid rule with no exceptions. Choosing the best city representing a region is according to the tz-Theory file a process governed by a solid rule, but it looks really silly if there is in a China database-part no reference to Beijing. Choosing Shanghai because it happens to have a larger population than Beijing is following the rules far too strictly. The same can be said for Viet Nam: Saigon is larger, but capital Ha Noi would be a more logical choice. Yes I know, we discussed this already from 2000-09-26 to 2000-10-05. But, look at this example. Africa tz-file. South Africa: Cape Town is the legislative capital, Bloemfontein is the judicial capital, Pretoria is the administrative capital. Guess what? The tz-database has Johannesburg as representative city. Why? According to the UN Population Count website (http://www.un.org/Depts/unsd/demog/ctry.htm) Johannesburg has fewer inhabitants (712507) than 'winner' Pretoria (854616), but Pretoria has not been chosen for the tz-database. So there exists an exception from the solid Theory-file rules. Why not Beijing and Ha Noi as other useful exceptions? "Eastern Turkistan (Turkestan), Uyghuristan, Uyghuria, Chinese Turkistan, Xinjiang, Sinkiang, Uyghur Autonomous Region." - Whatever. http://www.ccs.uky.edu/~rakhim/et.html (WWW Virtual Library for Eastern Turkestan.) Beijing time, China time. I get the impression that 'Beijing time' is used by Chinese sites with an English section, whereas 'China time' is used by some American sites. For those who are interested: a nice overview of the pinyin & tongyong (amongst others) romanizations: http://www.romanization.com/ Regards, Oscar van Vlijmen 2001-09-25