> None of your suggestions will work with heterogenous environment. Even
> if Jaybird defaults to UTF8, all Delphi apps will default to NONE,

Will they?

DB Express, EMB provides for last years, already uses Unicode, even before Delphi switched to it (they used COM BSTR as main type in pre-Unicode Delphi).
Even then in conenction setup GUI there as field to specify charset, probably for table names.
http://www.danielmagin.de/blogimages/tib/8.png

I think same would go for 3rd party DBX providers for IB/FB as well.

same for Borland .Net Data Provider
http://www.compress.ru/Archive/CP/2004/6/13/fig10.jpg


Also, DBX would have a published property for that
http://docs.embarcadero.com/products/rad_studio/delphiAndcpp2009/HelpUpdate2/EN/html/delphivclwin32/SqlExpr_TSQLConnection_LocaleCode.html

BDE is outdated and not to be described.
ADO and OLE DB provider would probably work via UTF-16 COM BSTR again

Interbase Express has the charset explicit, so probably good programmer would set it
http://studmaster.ru/st/delphi/delphi7secr/chapter18/pic1.gif

Unified Interbase has published property CharacterSet

FIB+ makes it a 1st class citizen by
TConnectParams = class(TPersistent)
property UserName: string;
property RoleName: string;
property Password: string;
property CharSet : string;
end;

IBObjects has in TIB_Connection the property CharSet: string;

So... in native Delphi that was accounted for for a while.

----

> they
> won't be able to read that UTF-8 characters.

They would not have to.
Server would convert them to 1251 or anything. If possible, of course.

> The only solution is to ditch the NONE completely, leaving it for
> backward compatibility only.

Why not.
NONE is not supposed to work in heterogeneous environment, until every app would make peer-to-peer support tricks.