Hi folks,
I was wondering about Svarcom translated strings and accents... with old PC-XTs with CGA/MDA video adapters there is no MODE CON CP functions, so the messages could be garbled...
Better is to avoid accented chars non existent on CGA/MDA character generator ROM?
Some old brazilian PC-XT manufacturers, such as S!D, Cobra, Itautec and others had special characters on video ROM tables for this purpose...
"á" can be shown on most adapters, but "ã" (with tilde) surely not.
The high-ascii bytes in messages are indeed likely to be garbled on MDA and CGA adapters. I do not think it is sensible to "break" the messages for all users, though - in some languages it would make the messages very hard to understand (Turkish, Polish, Russian, ...). XT machines are probably best left installed as EN-only - unless they come with a proper hardware code page, of course.
An alternative way could be to ship different translation sets with and without high ascii. For example "LANG=PL" would be normal, 8-bit enabled messages in Polish, while "LANG=pl" could be the same messages but with low ASCII only (ą -> a ; ć -> c ; ł -> l etc). This could be easily automated by utf8tocp, but would come with a doubled cost of storing the NLS messages. I do not think it's a desirable complication.
It's also worth noting that KEYB does not work properly on XT-class machines anyway --
https://github.com/SvarDOS/bugz/issues/114#issuecomment-2307899744
Mateusz
PS. Thanks for the BR updates in SvarCOM and INSTALL. BR INSTALL is missing the string 3.7, would you mind adding it please?
Done.
For sake of curiosity, I seen some NLS strings on MORE command source file... Can I add an "BR" dump byte section there too?
Something as
LANGLIST:
DB "EN", "MORE$" ; EN must be 1st to be used as default
; non-EN strings follow
DB "BR", "MAIS$"
DB "DE", "WEITER$"
DB "DK", "MERE$"
DB "ES", "M", 0B5h, "S$" ; "MAS" with an A-acute (CP 850)
DB "FI", "LIS", 8Eh, 8Eh, "$" ; "LISAA" - AA with diaeresis (CP 850)
DB "FR", "PLUS$"
DB "HU", "T", 99h, "BB$" ; TOBB with O-diaeresis (CP 852)
DB "IT", "PI", 0EBh, "$" ; "PIU" with U-acute (CP 850)
DB "LV", "T", 0A0h, "L", 0A0h, "K$" ; "TALAK" with A-makron (CP 775)
DB "NL", "MEER$"
DB "NO", "MER$"
DB "PL", "DALEJ$"
DB "RU", 84h,80h,8Bh,85h,85h,'$' ; "DALEE" (CP 866)
DB "SL", "VE", 0ACh, "$" ; "VEC" with C-caron (CP 852)
DB "SV", "MERA$"
DB "TR", "DAHA FAZLA$"
DB, 0 ; LANGLIST terminator
the word "MAIS" is codepage-agnostic.
I have added the "MAIS" translation to MORE, thanks!
This makes the MORE.COM binary 518 bytes large, so I will have to look where to save 6 bytes to make it fit inside a single sector again.
The good news is: It is not needed to add "PT", since the words are the same... ;-)
MORE 2024.0 published today.
- BR & PT languages
- size optimizations
- minor fix related to parsing the LANG env. variable
MORE 2024.0 is available in the SvarDOS repo of packages (through pkgnet), as well as on its own page at
http://svardos.org/more/.
While testing with 86Box 3.11, on a 486SX machine, installer crashes right after selecting any language but English.
Machine config: CGA card, 16MB RAM, no co-processor, HD 127MB