On Friday December 15 2017 16:07, you wrote to me:
MvdV>> You may not be able to read it, but you managed to properly MvdV>> display and quote it in your reader. Complete with a correct MvdV>> @CHRS: UTF-8 4 kludge. Congratulations!
KE> Well I dont know what I did, but some time ago I did some tinkerimg. KE> What surprises me is that almost half the exotic characters do resolve KE> to the proper glyphs in my terminal emulator.
Not bad considering that most console implementations only support a limited subset of the Universal Character Set.
KE> Things like UTF-8 4 can be manipulated with much hassle. But I still KE> maintain, that in the xlat implementation in Golded it should be UTF-8 KE> 2. As that is the type of translation table that is used.
When writing this message, there is no translation used at all...
KE> If the CHSRS character family is fixed to a translation table type, KE> then that type is redundant and has no use.
You come to the right conclusion with the wrong logic. The level parameter is one of those things that looked like a good idea at the time but that with the knowledge of today could have been done without. FTS-5003 recommends to ignore the level parameter on incoming and only use the name of the identifier for processing. For backward compatibility the level parameter should still be written on outgoing messages. For the time being...
Your logic is wrong because the starting point is wrong. The "if" in "If the CHSRS character family is fixed to a translation table type" is not satisfied. The level parameter is NOT fixed to any translation table type.
The level parameter simply denotes the type of encoding. No more, no less. Level 1 and 2 denote a one byte encoding. Level 3 is reserved for two byte encoding and Level 4 denotes a (variable length) multy byte encoding. It has nothing to do with any translation process. So when the text is encoded is UTF-8, the level is 4. Independant of how the encoding came into being.
The level parameter is redundant because the identifiers are unique. There is only one level associated with any identifier. When the identifier is Latin-1, one knows everything one needs to know to properly display the text. Same for UTF-8. The level parameter is a relic of the past and only there for backward compatibility. It may disappear one day. Or not...
KE> My aguments sound familear, we must be a yearly pet peeve.
That happens in the dark days before the festival of the Return Of The Light. ;-)
