2011/1/16 Eero Tamminen
<oak@helsinkinet.fi>
If it's a www-forum, that's a Browser problem, not OS one I think.
I think all mainstream browsers do their text handling on their own as they
need different text layouting than what widget toolkits provide and need to
be cross platform (and different OSes use different unicode standards).
HTML and email standards both include settings for encodings, so if Atari
end works according to standard, the other ends shouldn't have trouble
reading your posts.
Atari end on needs to run the text through encoding translation for which
there should be a lot of example code available as open source.
Netsurf doesn't have a problem displaying sites since it handles all that internally and using it's own engine but what about sending information? It will have to translate and it will have to translate from a standard. That standard can't be the atari one since no one cares enough to include it. Highwire on the other hand I believe does the translation through NVDI/Speedogdos/Vdi. So you need to fool it a bit to display greek since it doesn't have the encoding needed built-in. Again sending information is the problem. Based on what standard will it do the translation? And how does NVDI handle true type fonts? Remember that when the last version of NVDI was released those standards weren't as widely adopted as today.
I think it's better done through translation when sending & receiving
the messages (which the corresponding frontend program should automate).
Using 16-bit character encodings or UTF-8 would mean that text editing would
take more memory and be slower than when done with 8-bit Atari encodings.
I think that makes the added complexity of unicode<->8-bit encoding
handlings (which these programs anyway need to handle) worth the trouble.
. Eero
You are kind of overcomplicating things. Imagine the simple situation of transfering a text file to a windows/linux/macOSX pc for further editing. The pc will open it and see meaningless symbols. Infact what you need to do is pass the text file through an encoder program that will convert the text to either ISO or unicode. ISO encoding are 8-bit, just like the Atari ones and so you don't need more memory. Most applications will understand an iso encoding, at least for now. (I am talking about ISO 8859-x ones).
Regards,
Christos