Return-Path: Received: from relay2.vsu.ru ([62.76.169.17] verified) by vsu.ru (CommuniGate Pro SMTP 3.2b8) with ESMTP id 1178315 for CyrTeX-en@vsu.ru; Fri, 24 Dec 1999 11:31:34 +0300 Received: by relay2.vsu.ru (Postfix, from userid 5) id 67B3319B6; Fri, 24 Dec 1999 11:31:33 +0300 (MSK) Received: (from vvv@localhost) by vvv.vsu.ru (8.9.3/8.9.3) id LAA01904; Fri, 24 Dec 1999 11:20:02 +0300 To: CyrTeX-en@vsu.ru Subject: Re: Malyshev's t2 fonts References: From: Vladimir Volovich Date: 24 Dec 1999 11:20:02 +0300 In-Reply-To: Laurent Siebenmann's message of "Sun, 19 Dec 1999 03:35:13 +0100" Message-ID: Lines: 79 User-Agent: Gnus/5.070099 (Pterodactyl Gnus v0.99) Emacs/20.4 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Dear Laurent, "LS" == Laurent Siebenmann writes: LS> As for the "mistakes" in my friend Malyshev's "t2" encoding, I LS> understand they are not mistakes at all, but a conscious effort LS> to serve almost everyone with the few Type1 fonts we have LS> today. (And T2 is no standard, just T2A, T2B, T2C ;=) i wrote: VV> i will check the latest t2r10 etc virtual fonts in BaKoMa VV> distribution, and will report if there are still such VV> incompatibilities. Okay; i've made a check of "t2" fonts. I've installed BaKoMa TeX distribution v2.20, and generated a font table for "t2r10" using "tex testfont". I did the same for "c1r10" (a font from c1fonts bundle) and for "larm1000", and compared the results. Here is a non-full list of errors which are present in "t2r10": Slot correct glyph (T2A encoding) used glyph ("t2r10" font) ==================================================================== 0x0d CYRILLIC LETTER PALOCHKA comma 0x0e LEFT-POINTING ANGLE BRACKET guilsinglleft 0x0f RIGHT-POINTING ANGLE BRACKET guilsinglright 0x12 CYRILLIC CIRCUMFLEX quotedblbase 0x13 DOUBLE GRAVE ACCENT guillemotleft 0x14 CYRILLIC BREVE guillemotright 0x22 QUOTATION MARK (quotedbl) quotedblright 0x80/0xA0 GHE WITH UPTURN empty (should be at least GHE) 0x89/0xA9 KA WITH DESCENDER BYELORUSSIAN-UKRAINIAN I (like latin I) 0x8F/0xAF DZE empty (should be latin S) 0x98/0xB8 CHE WITH DESCENDER YO 0xAB KA WITH VERTICAL STROKE guillemotleft 0xBB NJE guillemotright ==================================================================== A few other empty slots could be filled with generated glyphs (by combining them with accents or descenders). Some of them are important for e.g. byelorussian language. yes, there are 3 pairs of guillemets, and two pairs of YO in the "t2" fonts! :-) LS> With his t2: a) Russian and Ukranian users can conform to T2A b) LS> Russian users conform to CP1251 even ukrainian users could not conform to T2A, because there is neither the GHE WITH UPTURN, nor a substitute for it (in the slot for GHE WITH UPTURN): GHE, in Basil's t2 fonts. not speaking about byelorussian users, who also could not conform to T2A, because of missing SHORT U (which, btw, could be easily generated in virtual font by putting an accent over CYRU even if there is no SHORT U in the type1 fonts). LS> Your standard input macros work for Russian and Ukranian, LS> although they may give nonsense beyond. Laurent, could you please explain what do you mean? LS> If one uses T2A encoded "la" fonts in TeX, then one can convert LS> to good vectorial PS and PDF output involving these type 1 fonts LS> --- by trivial substitutions using ".map" files. Because of the above mentioned mistakes in "t2" fonts, they could not serve as a valid replacement for "la" fonts, because the results will be incorrect. One could not recommend to use "t2" fonts if they will not be corrected. Moreover, the number of these mistakes (in accents and in letters) makes these fonts almost useless. Currently, virtual c1fonts (ftp.vsu.ru/pub/tex/c1fonts*) could be used as a (more valid) substitute for "la" fonts. We suggest Basil to include free c1fonts into BaKoMa TeX distribution instead of "t2" fonts (we will release soon a new version of c1fonts with more glyphs supported, and will upload them to CTAN; then there will be full support for at least Russian, Ukrainian, and Byelorussian languages; current version of c1fonts supports only Russian, but there are no known errors in the accents range, as in "t2" fonts). Best regards, -- Vladimir.