We develop a SMS Gateway product that interfaces with applications and SMSCs. As unicode support in phones becomes more widespread, we are also seeing more unicode messages being sent from applications. Also we are starting to receive unicode from the users.
Sending and receiving unicode is not a problem for us. The problem is that the UCS2 encoding used in SMS can encode ANY 2-byte unicode character but the phones only display a small fraction of this set (which varies by software package installed for a particular country / region).
So we have no idea what characters can be displayed by a certain phone. It is very hard to test this by "brute force" because not only do you have to send a huge amount of messages to the phones to test it (close to a thousand to test entire range and manually verify it from the screen but also getting a Thai phone in Finland is not too easy (yes, we have gotten chinese etc phones but not all of the versions).
In Greece for example, we ran into a large batch Nokia phones that use Unicode composing as the default if the phone UI language is greek. This is pretty annoying, especially if the receiver does not have a phone with greek unicode support. GSM charset only has capital greek letters (no lowercase) so it's understandable that someone will want to use unicode but not reasonable that unicode sending is the default mode.. In our software we ofcourse send any received unicode to the application but again the application does not know exactly which characters are supported by the phone when it sends a reply.
I know this question has been asked many times. I will still ask again..
Is there or will there be available a listing of unicode characters supported in different language versions of Nokia phones? The lack of this information is making value-added services development in users native tongue very hard in non-latin1 environments.
I realize that for many languages, like Greek, it is easy to assume that everything in a commonly used 8bit codepage charset (like ISO-8859-7) is supported. However I'd like an indication/confirmation that this is so.. For chinese, thai etc this is a TOTALLY different story and ANY info would be great.
Next I would like the same for WAP and MMS of course..