This file belongs to the Latin Modern package. The work is released under the GUST Font License. See the MANIFEST-Latin-Modern.txt and README-Latin-Modern.txt files for the details. For the most recent version of this license see http://www.gust.org.pl/fonts/licenses/GUST-FONT-LICENSE.txt or http://tug.org/fonts/licenses/GUST-FONT-LICENSE.txt Ver. 2.004, 30.10.2009: * a bug in *lssq.fd files corrected (thanks to Heiko Oberdiek); fonts were not touched, only the time stamp changed Ver. 2.004, 07.10.2009: * malformed `Euro' sign in oblique fonts corrected (thanks to G\"unter Spahlinger) * LaTeX *.fd files improved (better substitutions for unavailable variants) Ver. 2.003, 16.09.2009: main modification: shapes of LM glyps consistent with D.E. Knuth's recent changes in CMs plus lots of tiny changes; most important changes are described in details below: * fonts and glyphs modified according to D.E. Knuth's changes in CMs (note that the corrections related to glyph shapes, not metric data): -- leftward arrow (char '040), rightward arrow (char '041), upward arrow (char '042), downward arrow (char '043), left-and-right arrow (char '044), northeast arrow (char '045), southeast arrow (char '046), northwest arrow (char '055), southwest arrow (char '056), up-and-down arrow (char char '154) lmsy5 lmbsy5 lmsy6 lmsy7 lmbsy7 lmsy8 lmsy9 lmsy10 lmbsy10 -- upward arrow (char '013), downward arrow (char '014), lmcsc10 lmtcsc10 lmtti10 lmtt8 lmtt9 lmtt10 lmtt12 -- calligraphic F [less swashy], H [extended crossbow], I [lower right corner], T [foot] lmsy5 lmbsy5 lmsy6 lmsy7 lmbsy7 lmsy8 lmsy9 lmsy10 lmbsy10 -- lowercase beta (char '014), delta (char '016), omega (char '041) lmmi5 lmmib5 lmmi6 lmmi7 lmmib7 lmmi8 lmmi9 lmmi10 lmmi12 lmmib10 -- sans serif C and G [smaller overshoot] lmss8 lmss9 lmss10 lmss12 lmss17 lmssbx10 lmssdc10 lmssq8 lmssqbx8 very many thanks to Barbara Beeton, David M. Jones and Karl Berry * glyphs `eth', `Eth', `thorn', and `Thorn' reworked (as suggested by Arni Magnusson); it turned that there is no adequate template for all fonts, in particular the EC fonts contain apparent bugs; cf. e.g., `eth' (char 240) in ecdh1000 (counterpart of lmdunh10 -- enormously long cross) or eclb8 (counterpart of lmssqbx8 -- the cross hidden in the stem); COMMENT: Arni Magnusson wrote: ``I would love to use lmodern instead of cm-super''. Alas, this cannot be easily accomplished: the CM-SUPER fonts are based on the EC fonts, and both collections are not compatible with the CM (and thus LM) fonts! The reason is that the EC and thus CM-SUPER fonts have in most cases widths slightly smaller (with the coefficient ca. 99.975%) than the relevant CM fonts; exceptions are ecssdc10 and ecsql8, where the width coefficients are equal 100%, and ecrm1728, where the widths are greater by 1.5%--2.5% in comparison with cmr17, hence switching between EC/CM-SUPER and CM/LM is in general bound to produce different results * a bug in encodings fixed (sometimes instead of `backslash' other characters were typeset) * typos in ligature schemes in `comm_mph.mp' corrected (`quotedblbase' and `quotedblleft' instead `quotedblright') * `\pound' in math used to be typeset as dollar -- fixed; in ordinary text, however, the result may depend on the current encoding (for T1 is OK) -- thanks to Caity Ross for pointing out the problem * OTF structure corrected and improved (in particular, the ligature ffi works now properly) * `visiblespace' renamed to `uni2423' (suggestion of Werner Lemberg and Karl Berry) -- hopefully everywhere... * `Dblgravecomb' renamed to `dblGravecomb' * `nomero' in MT1 sources renamed to `numero', name `afii61352' no more used * undertie added (suggestion of Werner Lemberg) U+203F -- undertie (smile-shaped glyph placed below baseline) U+2040 -- tie (frown-shaped glyph placed at the top line) U+2054 -- undertieinverted (frown-shaped glyph placed below baseline) nota bene, there is no code point in the Unicode table for the inverted tie... * uppercase Greek completed (duplicates of Latin capitals; suggestion of Hans Hagen) * glyphs `acaron', `Acaron', `icaron', `Icaron', `idieresisacute', `Idieresisacute', `jcaron', `J_caron', `ocaron', `Ocaron', `ucaron', `Ucaron', `udieresisacute', `Udieresisacute', `udieresiscaron', `Udieresiscaron', `udieresisgrave', and `Udieresisgrave' added (proposal of Oliver Corff) -- all the glyphs but `J_caron' have code points in the Unicode table and names assigned in the Adobe Glyph List (incidentaly, in J_caron apostrophe is used alike in lcaron and tcaron -- cf. http://en.wikipedia.org/wiki/J-caron) * glyps orogate and Orogate (`rogate' in Polish means `with horns') added -- they are used to denote nasal vowels in ancient Polish texts (ca. XIV century) * uni2010 (hyphen duplicate) i uni2011 (non-breakable hyphen duplicate) added (suggestion of Hans Hagen) * LaTeX *.sty and *.fd files reworked, options `nomath' (prevents loading LM math fonts), `lighttt' and `variablett' added; sample usage: \usepackage{lmodern} \usepackage[nomath]{lmodern} \usepackage[variablett]{lmodern} \usepackage[lighttt,variablett]{lmodern} very many thanks to Marcin Woli\'nski REMARKS: * we are not going to introduce any extensions that would result in yet more TFMs files (e.g., for more convenient handling of oldstyle/normal digits) -- in such cases we suggest using the OTF fonts versions * improving hints is a kind of Sisyphean job -- there are no tools for automatic hinting checking; the only method is to check optically/manually each and every character with various display agents, on various screens under various operating systems and various resolutions. We are not totally opposing Sisyphean jobs in general, but having had that much time as we had, we decided to spend it on other improvements... Ver. 1.106, 22.01.2008: several meaningful fixes and enhancements * at the suggestion of Hans Hagen and Jonathan Kew (very many thanks for the interesting discussions and willing assistance), the `size' feature (opical sizing) implemented in the OTF files; this triggered a chain of changes of internal names in order to provide adequate family grouping (complying nolens volens with the "4-member family" paradigm) and resulted eventually in a completely new naming scheme * beware of the change of the OTF file names: OLD NAMES CHANGED NEW NAMES lmroman10-bold.otf lmroman10-bold.otf lmroman10-bolditalic.otf lmroman10-bolditalic.otf lmroman10-boldoblique.otf + lmromanslant10-bold.otf lmroman10-capsoblique.otf + lmromancaps10-oblique.otf lmroman10-capsregular.otf + lmromancaps10-regular.otf lmroman10-demi.otf + lmromandemi10-regular.otf lmroman10-demioblique.otf + lmromandemi10-oblique.otf lmroman10-dunhill.otf + lmromandunh10-oblique.otf lmroman10-dunhilloblique.otf + lmromandunh10-regular.otf lmroman10-italic.otf lmroman10-italic.otf lmroman10-oblique.otf + lmromanslant10-regular.otf lmroman10-regular.otf lmroman10-regular.otf lmroman10-unslanted.otf + lmromanunsl10-regular.otf lmroman12-bold.otf lmroman12-bold.otf lmroman12-italic.otf lmroman12-italic.otf lmroman12-oblique.otf + lmromanslant12-regular.otf lmroman12-regular.otf lmroman12-regular.otf lmroman17-oblique.otf + lmromanslant17-regular.otf lmroman17-regular.otf lmroman17-regular.otf lmroman5-bold.otf lmroman5-bold.otf lmroman5-regular.otf lmroman5-regular.otf lmroman6-bold.otf lmroman6-bold.otf lmroman6-regular.otf lmroman6-regular.otf lmroman7-bold.otf lmroman7-bold.otf lmroman7-italic.otf lmroman7-italic.otf lmroman7-regular.otf lmroman7-regular.otf lmroman8-bold.otf lmroman8-bold.otf lmroman8-italic.otf lmroman8-italic.otf lmroman8-oblique.otf + lmromanslant8-regular.otf lmroman8-regular.otf lmroman8-regular.otf lmroman9-bold.otf lmroman9-bold.otf lmroman9-italic.otf lmroman9-italic.otf lmroman9-oblique.otf + lmromanslant9-regular.otf lmroman9-regular.otf lmroman9-regular.otf lmsans10-bold.otf lmsans10-bold.otf lmsans10-boldoblique.otf lmsans10-boldoblique.otf lmsans10-demicondensed.otf + lmsansdemicond10-regular.otf lmsans10-demicondensedoblique.otf + lmsansdemicond10-oblique.otf lmsans10-oblique.otf lmsans10-oblique.otf lmsans10-regular.otf lmsans10-regular.otf lmsans12-oblique.otf lmsans12-oblique.otf lmsans12-regular.otf lmsans12-regular.otf lmsans17-oblique.otf lmsans17-oblique.otf lmsans17-regular.otf lmsans17-regular.otf lmsans8-oblique.otf lmsans8-oblique.otf lmsans8-regular.otf lmsans8-regular.otf lmsans9-oblique.otf lmsans9-oblique.otf lmsans9-regular.otf lmsans9-regular.otf lmsansquotation8-bold.otf + lmsansquot8-bold.otf lmsansquotation8-boldoblique.otf + lmsansquot8-boldoblique.otf lmsansquotation8-oblique.otf + lmsansquot8-oblique.otf lmsansquotation8-regular.otf + lmsansquot8-regular.otf lmtypewriter10-capsoblique.otf + lmmonocaps10-oblique.otf lmtypewriter10-capsregular.otf + lmmonocaps10-regular.otf lmtypewriter10-dark.otf + lmmonolt10-bold.otf lmtypewriter10-darkoblique.otf + lmmonolt10-boldoblique.otf lmtypewriter10-italic.otf + lmmono10-italic.otf lmtypewriter10-light.otf + lmmonolt10-regular.otf lmtypewriter10-lightcondensed.otf + lmmonoltcond10-regular.otf lmtypewriter10-lightcondensedoblique.otf + lmmonoltcond10-oblique.otf lmtypewriter10-lightoblique.otf + lmmonolt10-oblique.otf lmtypewriter10-oblique.otf + lmmonoslant10-regular.otf lmtypewriter10-regular.otf + lmmono10-regular.otf lmtypewriter12-regular.otf + lmmono12-regular.otf lmtypewriter8-regular.otf + lmmono8-regular.otf lmtypewriter9-regular.otf + lmmono9-regular.otf lmtypewritervarwd10-dark.otf + lmmonoproplt10-bold.otf lmtypewritervarwd10-darkoblique.otf + lmmonoproplt10-boldoblique.otf lmtypewritervarwd10-light.otf + lmmonoproplt10-oblique.otf lmtypewritervarwd10-lightoblique.otf + lmmonoproplt10-regular.otf lmtypewritervarwd10-oblique.otf + lmmonoprop10-oblique.otf lmtypewritervarwd10-regular.otf + lmmonoprop10-regular.otf * and beware of the change of the PostScript Type 1 internal font names -- we've fallen into this trap (in particular, all font maps have changed, also lm-rep-*.maps -- thanks, Mojca) * ligatures I_J and i_j are now available (in OTFs; feature `liga') only for Dutch; similarly, the ligature f_k -- for Polish; the `locl' feature reimplemented -- using an artificial glyph i.TRK is no longer needed (thanks to Arthur Reutenauer for comments and suggestions) * the metric bugs (spotted by Karel P\'\i\v{s}ka -- thanks!) corrected; actually, they turned out to be the consequence of a bug in CM fonts -- this is a serious matter, hence a few words of explanation; Karel observed that widths of `j' (lmri*), `x' (lmu10) and `sterling' (lmu10) were different from their CM counterparts; the latter two were just mistakes, while the discrepancy between the widths of the letters `j' in all italic fonts (hence lmu10, and also lmbxi10) is the result of apparently wrong setting of the width of `dottlessj' in MF sources: italsp.mf: cmchar "Dotless italic letter i"; beginchar(oct"020",5u#,x_height#,0); [...] cmchar "Dotless italic letter j"; beginchar(oct"021",5.5u#,x_height#,desc_depth#); [...] itall.mf: cmchar "Italic letter i"; beginchar("i",5u#,min(asc_height#,10/7x_height#+.5flare#),0); [...] cmchar "Italic letter j"; beginchar("j",5u#,min(asc_height#,10/7x_height#+.5flare#),desc_depth#); [...] The value `5.5u#' was perhaps brought to `italsp.mf' from romansp.mf... Whatever, we decided that the glyphs `dotlessj' and `j' should have the same width as the letter `dotlessi' (and `i') in italic fonts in both CMs and LMs -- retaining the discrepancy between the width of `j' and `dotlessj' we considered unfounded; note that the width of j-derivatives changed * indefatigable Karel P\'\i\v{s}ka pointed out also that small caps have non-uniform heights which resulted in uneven positioning of accents -- corrected * dots over the ij ligature aligned horizontally with dots over i and j (for a misterious reason they were placed slightly higher) * the heights of digits are now uniform (spotted by Mojca Miklavec and Taco Hoekwater -- thanks); they differ, however, from CM ones because of the limitation of 16 different heights per TFM * at the request of Idris Samawi Hamid, several glyphs added for romanized Arabic transliteration: `[Dd]linebelow', `[Hh]brevebelow', `[Hh]dieresis', `[Hh]tilde', `[Ll]tilde', `[Ss]dotbelow', `[Tt]linebelow', `[Th]dieresis', `[Th]tilde', `[Zz]dotbelow', ringhalfleft, ringhalfright, and, additionally, narrower `macron.alt' and `[Ii]macron.alt'; these changes implied also the addition of `linebelow' and `macronbelow' accents (also as ``combined'', i.e., zero-width variants) which, in turn, resulted in changes of Private Unicode Area codes (we abandoned the idea of being fully consistent with Adobe's PUA or PUA of such popular fonts like Minion); by the way, all accents having attributes `below' and `inverted' are named uniformly -- first `below', then `inverted' * `[Ee].reversed' renamed to `[Ee]reversed' (exist in Adobe Glyph List 2.0), added `eturned' and `schwa', all glyphs assigned proper unicodes (suggestion of Mojca Miklavec) * as an ``at leisure'' exercise, one more glyph, `infinity', was added (programmed on the base of MF sources) * at the request of V\'{\i}t Z\'yka, variants of the `cs' encoding for small caps and monospace fonts were introduced; at the suggestion of Staszek Wawrykiewicz, a similar variant was introduced for the `qx' ecoding (Polish; actually, `qx' had already a buggy implementation for monospaced fonts) and -- at the suggestion of Marcin Woli\'nski -- for the `rm' encoding (``Regular Math'', i.e., for fonts to be used as text fonts in math, e.g., \fam0) encodings. * wrong path directions corrected in `onequarter' and `threequarters' in lmth10 and lmtk10 -- thanks to Karel Hor\'ak * kerns specific only for the T5 encoding (i.e., Vietnamese) removed from AFMs and thus from OTFs; in particular, the deprecated `ka' kern pair removed from serif fonts (from TFMs it was removed long ago); also, strange pairs `eacute V', `ecircumflex V', `ecircumflexacute V', ..., etc., occuring in T5 encoding, now do not migrate to AFMs * math fonts (PostScript Type 1) have been equipped with PFMs for installing under Win systems; there are two AFM files for each PFB file: (1) *.AFT -- consistent with TFM; (2) *.AFM -- consistent with PFB * LaTeX fd files slightly modified by Marcin Woli\'nski -- thanks (no functional changes expected) Ver. 1.010x, 28.02.2007 (internal version number not changed): * only OTFs re-generated (files lm1.010x-bas.zip, lm1.010x-otf.zip); reasons: 1. hhea and OS/2 tables set explicitly following the recommendations contained in: http://typophile.com/wiki/Vertical%20Metrics%20How-To http://typophile.com/node/28277 2. widths of characters in the CFF table used to be rational (inherited from Type 1), but should be integers; the documentation http://www.microsoft.com/OpenType/OTSpec/hmtx.htm says: In CFF OpenType fonts, every glyph's advanceWidth as recorded in the hmtx table must be identical to its x width in the CFF table. and entries in the hmtx table must be unsigned short numbers. 22.01.2007: no version change * Lithuanian *.fd files corrected (reported by Sigitas Tolusis): L7X -> L7x Ver. 1.010, 16.01.2007: * important bugs fixed: Richard Kinch in 1995, http://www.truetex.com/tug95ps.zip, Karel P\'\i\v{s}ka, http://www-hep.fzu.cz/~piska/lm2006.html thanks to both; moreover, `perthousand' and `permyriad' corrected in all italic fonts (wrong path directions) * LaTeX support fixed and enhanced with LM math (thanks to Marcin Woli\'nski) * math text metric files (rm-*.tfm) provided * the letter `mu' corrected (other Greek letters must wait for a while) * `[lL]dotaccent' renamed to `[lL]dot', wrong shapes (dot above) corrected * `cedilla' adjusted in lmri*, lmu10, and lmdunh10 * `[CcSs]cedilla' reprogrammed in lmri* and lmu10 * `ogonek' and `cedilla' have forced height=0 now (in TFMs) * new encodings (cs, l7x) added * some glyphs renamed: `*superior' -> `*.superior', `*.oldstyle' -> `*.taboldstyle', `*.oldstyle.prop' -> `*.oldstyle' (names `*oldstyle', i.e., without a dot, removed) * several glyphs added: combining accents, `sfthyphen' (a duplicate of `hyphen' for non-TeX applications), `[eE].reversed', `[tT]cedilla', Romanized Sanskrit letters (`[dD]dotbelow', `[hH]dotbelow', `[lL]dotbelow', `[lL]dotbelowmacron', `[mM]dotbelow', `[nN]dotaccent', `[nN]dotbelow', `[rR]dotbelow', `[rR]dotbelowmacron', `[tT]dotbelow' -- the suggestion of R (Chandra) Chandrasekhar) * the structure of OTF files corrected -- grouping within families changed (following Adam Twardoch's suggestions -- thanks), features made more reasonable, glyph naming adjusted to standard recommendations, OTF version number consistent with the PS one * the technical documentation provided * a few minor changes in sources Ver. 1.000, 13.04.2006: the first official release ... Ver. 0.500, 09.11.2002: the first Latin Modern pre-release Ver. 0.100, 17.09.2002: the first results (then called AE, after L. Engebretsen) Ver. 0.000, 29.04.2002--03.05.2002: project started (EuroBachoTeX)