= Сообщение: 1101 из 7127 ====================================== FTSC_PUBLIC = От : Michiel van der Vlist 2:280/5555 14 Jan 14 13:03:47 Кому : Björn Felten 14 Jan 14 13:03:47 Тема : CHRS kludge implementations FGHI : area://FTSC_PUBLIC?msgid=2:280/5555+52d529ce На : area://FTSC_PUBLIC?msgid=2:203/2+52d35980 = Кодировка сообщения определена как: UTF-8 ================================== Ответ: area://FTSC_PUBLIC?msgid=2:203/2+52d54a06 ============================================================================== Hello Björn,
On Monday January 13 2014 04:12, you wrote to Nicholas Boel:
NB>> If it's at a level 4, it's supposed to let other software know NB>> there are multibyte characters in the header or message.
BF> As if UTF-8 in itself doesn't say that, you mean?
UTF-8 has some redundancy. Not all byte combinations are "well formed". The presence of non "well formed" strings is a good indications it is not UTF-8.
In your case he following strategy seems to work:
o When it is a "well formed" string assume the encoding is UTF-8.
o Obviously a single byte in the range 128-255 is not a well formed string. In that case assume it is Latin-1.
BF> Anyway, unfortunately the CHARS kludge is usually not applied to BF> the header.
What do you mean by "usually"? We do not see many examples of non ASCII in the headers in Fidonet do we? Far too few to apply the concept of "usually". Your JAmmNNTP does not, my Golded does.
BF> I would love to see an editor that can do so --
Look at the header in this message and look at the kludges that tell you what editor was used. You are looking at it.
BF> unfortunately I haven't had time to make JamNNTPd make it happen, but BF> maybe eventually I'll find the time. 8-)
