On 12/8/15 2:26 PM, rick@... [firebird-support] wrote:

>
> I've used FB for years with ISO8859_1 charset and collation.
>
> Now I designing a new DB for a big app update in FB3 and wonder about
> this charset and collation.
>
> For test purposes I am using UTF8 and UNICODE_AI_CI and I like that I
> can now don't have to worry about casing comparisons nor accents or
> other special characters.
>
> In fact though, all the data in will be in the regular english-us
> alphabet with maybe a little in spanish for things like "japapeño",
> but still all single byte chars.
>
> From what I read, only one byte is used for storing these characters,
> so the same amount of space would be used as compared to ISO8859_1.
>
> What I am wondering before I get too far along, is there any downside
> to using UTF8 and UNICODE_AI_CI?
>
> Regards,
>
> Rick
>

UTF-8 uses 1 byte for code points 0-127 (the basic ASCII Characters),
and two or more bytes for other characters (like ñ), so you may take a
small hit is data size. Code pages like ISO8859-1 represent all
characters in a single byte (for those they represent).


--
Richard Damon