summaryrefslogtreecommitdiff
path: root/Master/texmf-dist/tex/generic/ibycus/psibycus.RME
diff options
context:
space:
mode:
Diffstat (limited to 'Master/texmf-dist/tex/generic/ibycus/psibycus.RME')
-rw-r--r--Master/texmf-dist/tex/generic/ibycus/psibycus.RME99
1 files changed, 99 insertions, 0 deletions
diff --git a/Master/texmf-dist/tex/generic/ibycus/psibycus.RME b/Master/texmf-dist/tex/generic/ibycus/psibycus.RME
new file mode 100644
index 00000000000..0abc81d1dae
--- /dev/null
+++ b/Master/texmf-dist/tex/generic/ibycus/psibycus.RME
@@ -0,0 +1,99 @@
+This file is psibycus.RME
+
+The Type1 version of the Ibycus Greek font (regular weight only)
+available on orhan.classics.washington.edu, in the directory
+/ftp/pub/tex has been sell enough received that it is being added to
+the ibycus4 package. The font is still largely unhinted, but I am
+beginning to wonder whether that matters. I did plan to provide a
+rather thorough set of hints, since the primary purpose for making up
+this font was to get away from the crude bitmap scaling that is all
+you can get out of PDF readers. Hinting, however, is even worse than
+filing matrices. It is a soul-destroying bore, and I cannot do very
+much of it at any given session. The absence of hints will not be
+noticed at resolutions of 600dpi and above. On paper, the weight is
+just about right. In PDF files created with Adobe Distiller, the
+screen appearance is remarkably good. In xdvi it is even better. In
+ghostscript the rasterization uncontrolled by hints produces too dark
+an effect, but it does make it that much more readable. The font has
+a private UniqueID in the open range for now, but a registered
+UniqueID will be applied for.
+
+The original METAFONT realization of ibycus4 remains the controlling
+form of the font. The TeX user will see absolutely no difference
+in the set-widths of the Type1 font because the set widths in the
+TFM file are directly related (with infinitesimal roundings) to
+those in the the TFM file for the METAFONT version. NOTE: My previous
+hope that I could use exactly the same TFM for the type1 version
+and the METAFONT version failed because some renderers such as
+pdftex choke on even the slightest difference in set width.
+
+The Type1 control points have been derived from METAFONT log output
+generated by "tracingspecs". This is not impossible, as has sometimes
+been claimed, but it does take work.
+
+The additions to the METAFONT based package consist of:
+
+1. IbycusHTG-Regular.pfa ( = fibr.pfa = IBYHTGR_.PFB ).
+ The character designs are Silvio Levy's. Some small differences
+ in accent positioning and vertical positioning on the classic
+ METAFONT typeface grid have been allowed. My own lowercase lunate
+ sigma has been improved. These changes will ultimately be read
+ back into the ibycus4 METAFONT source.
+
+2. fibr.tfm (blended from the TFM file produced by afm2tfm and
+ the ligature and kerning from fibr84.tfm. The use of the Type1 font
+ is specified by calling on fibr, rather than fibr84.
+ Obliqued versions are invoked with fibo.tfm (derived in the same way).
+ Bold versions of the font are still exclusively METAFONT.
+ They will continue to be invoked as fibb84[89]?
+ until I make up Type1 versions of them (if I ever do).
+
+3. fibr.vf (fibo.vf), which serves to provide a reference into a dvips map
+ file. The raw TFM for this VF file is fibr84.tfm (fibo84.tfm), which
+ has the interesting effect of making METAFONT generated PK files a
+ fail-safe alternative in sizes above 10pt if the dvips map lookup fails.
+ The checksums for fibr.tfm fibr84.tfm and fibr.vf are identical.
+ likewise for fibo.tfm fibo84.tfm and fibo.vf.
+ This is arbitrary, since fibr.vf has integer escapement values
+ rounded from the METAFONT values in the tfm files.
+
+4. config.iby and iby.map. The iby.map file shows how to associate
+ fibr84 and fibo84 with IbycusHTG-Regular.
+ fibr84 IbycusHTG-Regular <IbycusHTG-Regular.pfa
+ fibo84 IbycusHTG-Regular ".167 SlantFont" <IbycusHTG-Regular.pfa
+ or
+ fibr84 IbycusHTG-Regular <fibr.pfa
+ fibo84 IbycusHTG-Regular ".167 SlantFont" <fibr.pfa
+ If you still use the large explicit psfonts.map from the web2c
+ days, these lines can be copied into psfonts.map for general use.
+ A better idea is to put them into a file iby.map and add the
+ line p +iby.map to the sort of psfonts.map that is provided
+ with teTeX.
+
+5. IbycusHTG-Regular.afm and IbycusHTG.enc. These are provided for
+ information only. The AFM file contains only the character info
+ and the pair-kerning data. There is no point in going further
+ since the Adobe convention for AFM is incapable of specifying the
+ ligature sequences needed for ibycus4. The encoding file could
+ supply both ligatures and pair-kerns, but to what purpose?
+
+6. Various TeX input files rewritten to use the fibr invocation in
+ place of the fibr84 invocation. (This is the only change that
+ users will need to make, other than installing the above files in
+ the TEXMF directory tree).
+
+IbycusHTG-Regular is licensed without charge for use in the creation of
+documents in all media, in accordance with the included copyright
+notice. Users are welcome and indeed encouraged to adapt the font
+to other typesetting systems. Note that iota subscript is applied to
+the affected vowel as a following zero-width character kerned
+drastically to the left. (Other well-known Polytonic Greek fonts
+do the same thing. This trick saves many, many glyph spaces for more
+constructive use.)
+
+ mackay@cs.washington.edu Pierre A. MacKay
+Smail: Department of Classics Emeritus Druid for
+ Denny Hall, Box 353110 Unix-flavored TeX
+ University of Washington
+ Seattle, WA 98195
+ (206) 543-2268 (Message recorder)