LM fonts: Bug reports and proposals

WWW: http://www-hep.fzu.cz/~piska/lm.html

Karel Piska

Institute of Physics, Academy of Sciences, Prague (CZ)

E-mail: piska@fzu.cz

(My surname with diacritic signs: P\'{\i}\v{s}ka)

LM: Bug reports and proposals

LM version 0.99.2 --- metric and encoding data
2005 Sep 19: Bug report: Oslash width (more details) is wrong in lmb* (bold) and lmss* (sans serif)
2005 Sep 19: Proposal: add lm-rep-csit.enc >fonts/enc/dvips/lm (sterling replaces dollar)
2005 Sep 19: Proposal: add lm-rep-csitt.enc >fonts/enc/dvips/lm (sterling replaces dollar)
2005 Sep 19: Proposal: change lm-rep-cstext.map >fonts/map/dvips/lm (uses the *it.enc and *itt.enc)
2005 Sep 19: Question: "j" in lmri(10|12|7|8|9) and lmbxi10 is wider than in the corresponding fonts cmti(10|12|7|8|9) and cmbxti10; the dots in "i" and "j" are greater.
Of course, (as always) all the products and proposals produced by any author(s) must by strictly checked by verificators.

CS-LM: Comparison of glyph shapes

2005 Nov 25: Uring/uring comparison (PDF) between LM and CS (PS version) and a proposal. LM is cyan, CS is red, the intersection is pink, a proposal is dark.
2005 Nov 22: Comparison of LM and CS glyphs (ZIP) (~4.7MB), especially the letters with accents. LM is cyan, CS is red, the intersection is pink.
cs*_lm*-0993.pdf (with hyphen) show the single glyphs per page in a large scale.
cs*_lm*_0993.pdf (with underscore) show one font (all the selected glyphs) per page with a smaller magnification. The boxes are from cs*.tfm.

CS-LM: Comparison of glyph widths

14 Sep 2005:
To allow replacement of PostScript Type 1 CSfonts with the corresponding Type 1 LMfonts in the final PS/PDF documents we expect the metric data (glyph sizes: widths) and the graphic representation will be the same. The next tables present a result of a comparison of widths between corresponding glyphs in CS and LM fonts.
Glyph widths in the PFB files have been compared!
Comments with an explanation of columns in the tables. The table with significantly different glyph widths ordered by fonts and ordered by glyph names .
List of all differences (most of them are negligible -- the difference 0.005 in a 10pt font means 0.00005 bp).

19 Sep 2005:
A complete list of compared fonts and glyphs in the units of the 1000x1000 glyph coordinate space (to check what have or have not been compared exactly).
The list of differences rescaled to pt in the final output device (if each font is rendered at its own size); but without RIP dependent roundings.
Missing fonts in the comparison:
(Comparison/verification has not been finished or the results are not complete.)
a) csu10
b) csinch (only common glyphs compared)
c) csdunh10,csff10,csfi10,csfib8 (not covered by LM: "funny" fonts)

A mistake with sterling
In the csti*(italic) Type 1 fonts (csti10,csti12,csti7,csti8,csti9,csbxti10) "dollar" denotes sterling!
(In the MAP file an *it* encoding should be used for the csti* fonts and in the corresponding ENC files the code 36 must be named "sterling".)
------------------
.022 csbxti10 dollar 15634/18=868.556 lmbxi10 sterling 13028/15=868.533
.020 csti10 dollar 13844/18=769.111 lmri10 sterling 8460/11=769.091
.007 csti12 dollar 13594/18=755.222 lmri12 sterling 10573/14=755.214
= csti7 dollar 15844/18=880.222 lmri7 sterling 7922/9=880.222
= csti8 dollar 816 lmri8 sterling 816
.027 csti9 dollar 14149/18=786.056 lmri9 sterling 9433/12=786.083
------------------
The real differences are negligible: e.g. 0.022 for a 10pt font 0.022*10/1000 bp = 0.00022 bp is the width difference; while the glyph width is 868.556*10/1000 bp = 8.68556 bp.

LM: Bug report n. 001 (glyph outlines)

21 Apr 2005: Bug report 001 for LM fonts version 0.98.3 (of 05.04.2005)
zipped files (ZIP) (~1.5MB) without 2 huge proofsheet files.

Appendix: Detailed description or visual explanation

under preparation: "Typical mistakes in conversion fonts to outline formats"

Local messages for CSTUG (also in Czech)

Bug report - CSfonts in the METAFONT version

2005 Dec 1: Uppercase acute depends on mode (resolution). The examples show only 2 fonts: csr10 (PDF) and csbx10 (PDF) - in the low resolutions the acute is thicker and in the high resolutions it is thinner than it should be.
Acrobat Reader ver. 7 (or 6 and higher) is recommended to see bitmaps properly.
Srovnani CS a LM - 3 (21 Oct 2005)
dopis vyboru CSTUG (txt)
Srovnani CS a LM - 2 (19 Sep 2005)
dopis do konference CsTeX (txt)
Ring.cs,ring.cs - proposal ver. 1.2 (14 Apr 2005)
(PFB) ringcs.pfb: Ring.cs/ring.cs glyphs collected in one font
(ZIP) Corresponding file extensions with Ring.cs/ring.cs in a METATYPE1 source format
Uring.cs,uring.cs - proposal ver. 1.1 (25 Mar 2005)
(PDF) lmacr.pdf: Comparison Uring/uring in CS/PS, CS/MF-MP-PS, LM and proposal 1.1
(PDF) lmacrdbl.pdf: dtto with scale factor 2
(TXT) commring.txt: Komentar k navrhu 1.1 (Comments in Czech)
(TXT) cslmtab.txt: Table of font names in CS and LM
(PFB) uuring.pfb: Uring/uring glyphs collected manually in one font


Created: 25 March 2005
21 Apr 2005: LM: Bug reports
14 Sep 2005: CS-LM: Comparison of glyph widths
19 Sep 2005: LM bug reports and proposals
25 Nov 2005: CS-LM: Comparison of Type 1 glyphs with accents
Last modified: 2 Dec 2005: CS bug report