> On Oct 6, 2025, at 5:57 PM, Jose Senna via Freedos-user > <[email protected]> wrote: > > I think the differences in vocabulary > are not the problem of portuguese versions, > but the different code mappings. For > instance, the (html) ã (/html), > much used in portuguese, maps as 84h in > both 437 and 860 codepages, while in > 850 it is C6h. This results in noticeably > weird characters if one is using the > 'wrong' keyboard /video settings.
You make an interesting point. While you could possibly map a translation to a different codepage, that may not be the proper codepage for that language. So, If the user has the correct codepage for a language active, they will end up with garbled or junk text. Side note… That is actually one of the things the conversion utility I was working on took into consideration when performing conversions. For example: when converting from an unknown codepage to UTF-8, it would test convert as many different codepage each to UTF-8. It would then use the multiple languages word data to create a “score”. The higher the score, the more likely it was that the correct codepage and language had been discovered. Not perfect. But, it worked reasonably well and accuracy improved in proportion with the amount of text being translated. When the attempting to perform a conversion using the incorrect codepage, a lot of junk words are generated. Things like “pi??a” (can’t really type out such things) are not words in any language and resulted in a much lower score for the conversion. :-) _______________________________________________ Freedos-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/freedos-user
