>>>We have an application which negotiates with license plates. It is focused to be international. It currently supports Canada, United States and another country. I would like to know how many characters you are usually reserving for such a field in your application knowing that it could be used to store license plates from all over the world.
>>
>>Well - http://en.wikipedia.org/wiki/European_vehicle_registration_plates#Common_letter_and_digit_systems_between_countries
>>
>>If you *know* the country that the car is registered in then I suppose it would be possible to build up a Regex library with one or more entries for each country.
>>
>>Don't see a country in Europe listed that has more than 8 letters/digits. Don't recall seeing anything longer in non-European countries...
>
>I like the regex idea - and the patterns will need changing as-you-go, ie start with a few countries and expand. Keeps one busy
>
>Also, the format may change in some country over time
>You need at least the country of registration - maybe the state as well - and the country (via the language set) should give you LTR or RTL reading order
>
>
>I'd start with nvarchar(16) - maybe even nvarchar(12) - should be sufficient and the length can be changed over time
>
>ps: I would not volunteer to write the regex patters for arabic, chinese, japanese

:-}
Regex for Kanji - wouldn't know where to start....