Reply 60 of 76, by VileR
- Rank
- l33t
Seems like none of the Memotek encodings are 100% identical to the Acer one. The closest one is the block at offset 800h, but nine high-ASCII characters are still different:
0xEB: Acer= Φ , Memotek= Έ
0xEC: Acer= ∞ , Memotek= Ή
0xED: Acer= ∅ , Memotek= Ί
0xEE: Acer= ∈ , Memotek= Ό
0xEF: Acer= ∩ , Memotek= Ύ
0xF0: Acer= ≡ , Memotek= Ώ
0xF9: Acer= ∙ , Memotek= £
0xFA: Acer= · , Memotek= ΄
0xFC: Acer= ⁿ , Memotek= ¨
Of course the glyph shapes for the Greek alphabet block also differ, but the characters themselves seems to be the same there (correct me if I'm wrong). So unless those magazine programs actually depend on these *non*-greek characters in the Acer ROM, then the Memotek one should be close enough 😉
Creating the data for a custom CGA ROM shouldn't be a problem... it's the same structure that the Acer ROM already has:
- 0h: 4 KB of 8x14 font data (unused on CGA, but Acer has a Greek encoding there too, probably for using the same ROM on mono cards)
- 1000h: 2 KB of 'thin' 8x8 font data (which Acer replaced with its Greek encoding)
- 1800h: 2 KB of 'thick' 8x8 font data (Acer keeps the US CGA charset at this location)
...with the last two 2 KB banks being jumper-selectable. So you could simply make a copy of the Acer ROM data - it should work just fine with an IBM CGA, and jumpering it to use the 'thin' font will give you the Acer-style encoding you want.
Of course, you could also replace those 2 KB with the 0800h Memotek data, or merge the two charsets any odd way... for instance, taking the Memotek charset (if you prefer that font style) and replacing just those 9 'problem' chars with the Acer ones.
Perhaps there are ready-made tools to do that, but I'd just use a hex-editor and CrystalTile 2... anyway, if you do want to "mix-and-match", let me know and I'd be happy to help!