Now there's an FAQ, by request:
http://www.gnu.org/software/freefont/FAQ.html
These are paraphrases of questions that have really been asked, most of them repeatedly (so it's at least an RAQ.)
There's a new release! It's
20120503
Find binaries and sources at
http://ftp.gnu.org/gnu/freefont/
There are lots of improvements, thanks to the efforts of many contributors.
Main efforts
============
Latin, Greek, Cyrillic
----------------------
Overall glyph-by-glyph standardization of spacing.
Re-kerned based on the new spacing.
Arabic
------
(Serif regular and bold) all letters re-done, better to resemble modern type in the Nask style, and better to use positional forms.
Main range and Supplement-B completed, as well as the first section of Supplement-A (all but ligatures).
Took steps to make the range functional for
Arabic, Persian (Farsi), Urdu, Pashto, Seraiki, Uighur, and Malay
as well as several African scripts.
(Mono) added basic range and tables.
Cyrillic
--------
Serbian/Macedonian localized forms for de, cursive gje, sha, be, ta,
Bulgarian alternative forms for several letters (serif, as Style Set)
(Sans) letters for Abkhaz and Azerbaijani, and Nivkh.
Devanagari
----------
(Serif and SerifBold) imported outlines from the Velthuis latex font, to support for Hindi, Marathi, and Sanskrit, including localized glyphs for Calcutta, Bombay, and Nepali.
(Sans) overhauled range, adding some glyphs and rearranging some tables.
(Sans Bold) Made range based on medium weight version.
Gujarati
--------
(Serif) Implemented range using glyphs from Samyak Gujarati font.
Deleted whole range in Sans (needs to start over again anyway).
Oriya
-----
(Serif) Implemented range in Serif using glyphs from Samyak Oriya font.
Malayalaam
----------
(Sans) Replaced range with with Meera 04 font of Swathanthra Malayalam Computing
Gurmukhi
--------
(Serif) Replaced range with glyphs from the Punjabi font by Hardip Singh Pannu
Sinhala
-------
(Serif) made a bold face
Old Italic
----------
(Mono) new letters constructed by S. White
Other additions
===============
IPA
---
Letters for North American languages
Saanich/Musqueam, Kwak̕wala
Some replacements for d-, l-, t-apostrophe
Runic
-----
(Mono) completely re-did range to fit weight, and to look better.
Syriac
------
Made new Unicode letters like those from Tim Erikson's Carlo Ator
Thaana
------
(SerifBold) derived from medium version
Bengali
-------
Ganda mark
Malayalam
---------
New Unicode characters, corrected some marks in Sans.
Greek
-----
Lunate letters
Latin Extended-A
----------------
Shona letters, other African letters
Latin Extended-D
----------------
H-hook (U+A7AA) for Chadian languages
Combining Diacritical marks
---------------------------
(Mono) Completed range
Currency
--------
(Serif) Indian Rupee courtesy of Daniel Johnson
Specials
--------
Filled out range in all faces
Control Pictures
----------------
Filled out range in regular style of each face
various additions to
--------------------
Letterlike symbols
Miscellaneous symbols
Mathematical Operators
Miscellaneous Mathematical Symbols-A
Miscellaneous Mathematical Symbols-B
Phonetic Extensions
Punctuation, new Punctuation Supplemental range
Geometric Shapes
Typographic
===========
small caps
----------
(Serif,SerifBold,Sans,SansBold)
Replacement glyphs and tables for core Latin letters.
also glyphs for script-size letters.
new substitution look-ups
-------------------------
Technical/legal
===============
WOFF (Web Open Font Format) version, as well as directions for constructing
WOFF derivatives.
Hashed out how the license works with Web Fonts technologies.
New file webfont_guidelines.txt contains instructions.
Project version control switched from CVS to SVN.
Re-arranged the tools/ directory;
now package .sfd files in a "*src*" tarball with a top-level Makefile.
Filled in IBM Classifications
Removed angle-brackets from license text that triggered bugs in some software.
Likewise scripts to automatically clean up the SVN version string in fonts.
Cleaned up some incorrect Indic language lookups, that weren't being triggered
correctly due to FontForge bugs.
Added script to generate Arabic test pages, courtesy of Emmanuel Vallois.
Clean-ups
=========
Many validation issues (some due to font changes, some due to FontForge changes)
Adjustments
===========
Latin, Greek, Cyrillic
----------------------
Much fiddling with accents, for various reasons (including bug reports.)
(Sans) acute and grave accent on capitals made to fit in line,
(Serif, Sans) Hungarian umlaut and double-grave relaxed angle a bit.
Regularized bullet-like characters somewhat.
Greek, Serif Cyrillic
----------------------
Made Phi, Ef more distinct
Reduced sizes of hooks on many letters.
Hebrew
------
Adjusted marks, cantillation points; widened Hebrew wide letters more.
Made yod in Yiddish distinct from Hebrew.
Osmanya
-------
Corrected Unicode values.
Malayalam
---------
Rearranged some incorrect lookups.
Tamil
-------
Several tables added to resolve spacing issues.
Super/Sub scripts
-----------------
Regularized positioning--All the "superior" numerals and a few superscript
Latin letters got positioned so they go over the em height.
Latin ordinals are positioned parallel to em height.
Many glyphs made to reference new small caps and script-size glyphs.
Thanks to
=========
Daniel Johnson
Tim Erikson
Emmanuel Vallois (!)
Zdeněk Wagner
Monika Shah
Swathanthra Malayalam Computing (Praveen A., Santhosh Thottingal)
Sarovar.org, who host the Samyak fonts (especially Rahul Bhalerao)
Jason Woolman (of Musqueam First Nation)
Chris Harvey (LanguageGeek)
Timothy Montler
Samyak font admin Rahul Bhalerao
Pavel Skrylev
Karl Berry
Alessandro Ceschini
Tim Erickson
Harshula
Sander van Geloven
Masoud Pourmoosa
Savannah.org maintenance team
And the numerous issue reporters, question-askers, detractors, and users!
The version control system for the project source files has been changed from CVS to SVN, for the improvements in manipulating directories, etc.
Anyone who checks the project out read-only, should now do so using
svn co svn://svn.savannah.gnu.org/freefont/trunk/freefont
Cheers!
A year and a half has passed since the last release. The delay was largely my fault--you have my apologies. A lot has happened though. Several new ranges have been added, some of which are hard to find anywhere as free software. Sinhala and Vietnamese should now be much more functional. Lots of symbols have been added. Many glyphs have been cleaned up and some more kerning has been specified, for a better overall look.
Get it at http://ftp.gnu.org/gnu/freefont/
Look for files dated
20100919
What follows is a summary of changes made to Gnu FreeFont since the last release. For a more complete listing, see the Changelog.
Unicode ranges
==============
new
---
Cherokee (Daniel Johnson)
Osmanya (Daniel Johnson)
Unified Canadian Aboriginal Syllabics (Daniel Johnson)
Tifinagh (Daniel Johnson)
Armenian in Serif (Daniel Johnson)
Vai (Daniel Johnson)
Kayah Li (Daniel Johnson)
Coptic (Steve White)
Glagolitic (Steve White)
deleted
-------
Tamil from Sans (due to absence of necessary ligatures)
replaced
--------
Sinhala is now taken directly from the TeX font 'sinh' by Yannis Haralambous.
It has been reviewed and corrected by a native speaker.
altered
-------
Armenian (spacing; Aiw to reference Phi on Daniel Johnson's advice)
Vietnamese (better accent placement, esp. in Sans)
Latin (general letter form improvements and clean-ups, numeral 1 and braces in mono re-designed for distinctness)
Greek (various letter form improvements: made nu distinct from Latin v)
Cyrillic (mark placement, letter forms)
Mahjong Tiles (got rid of a bamboo shoot)
Superscripts and subscripts (standardize placement)
Georgian (added missing letter in Mono, Sans)
Western Musical Symbols (much cleaning up, correcting)
Miscellaneous Symbols (re-design of musical flat symbol)
Hebrew (corrected some vowel marks)
expanded
--------
Currency Symbols
Latin Extended-D (Mayanist letters from Daniel Johnson)
Supplemental Arrows-A (completed in Serif)
Arrows (completed in Serif, corrected some switched glyphs)
Combining Diacritics
Spacing Modifier Letters
Letterlike Symbols
Mathematical Symbols
Miscellaneous Technical Symbols
Scan Lines
Geometric Shapes (completed in Sans)
Number Forms (new vulgar fractions; completed in Serif roman)
Combining Diacritical Marks for Symbols
Internatonal Phonetic Alphabet
Phonetic Extensions Supplement (completed in roman)
IPA Extensions (completed in roman bold and oblique)
Cyrillic (Abkhasian letters, yi_yi Ukrainian ligature)
Cyrillic Extended-B
Georgian (completed in Serif)
Gurmukhi (completed Unicode range)
Gujarati (completed Unicode range)
Devanagari (completed Unicode range)
Tamil (completed Unicode range)
Technical/Legal
===============
Added full GNU copyright notice to each SFD file.
Reduced cases of nested character references, to
ameliorate a bug in Mac OS 10.6
Various improvements to kerning
Removed TTF Names for sample text
TTF Names tables to give locale-specific names to the
various faces. Checked each with a native speaker.
Added special characters for TrueType,
.notdef .null nonmarkingreturn
Corrected the "replacement character"
Added GASP tables for control of hinting/smoothing (anti-aliasing) of TrueType.
PS Blues Values improvements; removed BlueFuzz altogether
Deleted TrueType tables 'cvt', 'fpgm', 'prep' because of FontForge complaints.
Arabic ligatures: turned on "ignore combining marks" on advice of Khaled Hosny.
tools
=====
range_report.py has been improved a lot and expanded for Unicode 5.2
ligature_lookups.py makes an HTML list and display of the ligatures in a font
Thanks
======
Foremost to Daniel Johnson, who drew and contributed several ranges, as well as providing technical and moral support.
To Harshula Jayasuriya for edits, advice, and other collaboration on the Sinhala range.
To Yannis Haralambous for permission to include his TeX sinha font
The Wellcome Library
Robert Kiley, Head of e-Strategy for the Wellcome Library
Thanks to Dominik Wujastyk, for providing us with feedback and contacts to repsonsible people at the Trust.
To Mina Magdy for much advice and criticism concerning the Coptic range (and for her tolerance of my choice of letter forms, which was different from hers).
To Pavel Skrylev for many additions to old Cyrillic ranges.
To Mohammed Anwari for correct ligatures in Buginese.
To Khaled Hosny for his patch for Persian.
To Karl Berry for hard work and overall sanity.
-- and to the Debian packaging team
-- and to the many bug reporters!
This is a maintenance release on the previous release of GNU FreeFont.
Get it at http://ftp.gnu.org/gnu/freefont/
Look for files dated
20090104
There are three main changes.
1) The Glagolitic range has been removed, due to an unfortunate mistake (mea culpa!) regarding licensing of the font on which the range is based.
Efforts to contact the original author were made but no reply was actually received. We are trying again, but until proper permission is given, these glyphs will remain excluded.
2) The old bug regarding excessive line spacing in Mono was inadvertently re-introduced in the last release. It has been fixed again, and the test scripts now contain tests for the problem.
3) Daniel Johnson has provided a bold version of his Cherokee range. Oblique faces have also been filled in.
Also, the license has been upgraded to GPL v3
Thanks!
Another release of Gnu FreeFont, now with *kerning*, and several new ranges!
Download it at http://ftp.gnu.org/gnu/freefont/
Find versions dated 20081226
See below for more details, and the release ChangeLog file for even more.
Big technical improvement
=========================
Ranges added
============
[These by George Douros]
Ranges adjusted
===============
Ranges filled out
=================
Latin Extended Additional
Miscellaneous Symbols
Combining Diacritical Marks
Currency Symbols
Spacing Modifier Letters
Superscripts and Subscripts
International Phonetic Alphabet
Phonetic Extensions
General Punctuation
Letterlike Symbols
Enclosed Alphanumerics
added APL characters (but why?)
Mathematical Alphanumeric Symbols
=================================
The suitability of this whole range is of arguable,
On the other hand it is in the standard, and other characters in Unicode (e.g. in Letterlike Symbols) refer to it, and I'm personally pushing for completion of math ranges. And George Douros has collected the glyphs, so in they go.
I replaced most of George's glyphs with copies of those in
other ranges of FreeFont, and some with Young Ryu's TXFonts.
His calligraphic, script and blackboard bold,
and his character naming scheme, remain.
More languages supported
========================
I find Daniel's Cherokee to be the best I've seen, and I've
downloaded most of the freely-available ones.
The extensions to Cyrillic should provide support for many
living languages, including:
Russian, Ukrainian, Macedonian, Serbian, Belarusian, Rusyn, Bulgarian, Altay, Abkhasian, Azerbaijani, Bashkir, Chuvash, Kazakh, Tatar, Uzbek, Yakut, Tajik, Kildin Sami, Mari, Khanty, Nenets, Mordvin, Komi, Enets, Moldavian, Chukchi, Itelmen, Aleut, Yupik
There is a movement on to revive Coptic, so we'll call it live.
I think the Thai range wasn't really functional before the mark lookups
were added. It has now been declared quite good by a native reader.
Kerning
=======
Finally realized why kerning didn't work in Linux, and wasn't
recognized in Windows. Fixed.
But I never liked the kerning tables that came with Nimbus.
I think they were automatically generated, but often didn't
make sense technically (a kern of 1EM out of 1000 is invisible)
or visually (many letters under-kerned, many badly over-kerned).
So I made my own kerning tables for Latin and basic Cyrillic.
It is likely that I made mistakes, or that I've generally over-done it.
(As always, comments are welcome!)
The kerning in Latin is elaborate--maybe too elaborate.
It was done with consideration at three different levels:
* repairs: letters so widely spaced that words appear broken
* evenness: even-looking white space between letters
* flow: letter spacing in printed text somehow flows in bulk.
This is rather more subtle than what some automatic area-averaging might
give (and harder to explain). Basically, I printed out a bunch of text
in various languages, and eyeballed it--a lot.
Latin tables were intended to provide kerning for all European languages,
and Turkish. Besides English, I printed out pages of text in
French, Spanish, and German
and specifically kerned for pairs in these languages.
I didn't do Vietnamese yet.
The policy for kerning in Cyrillic went the other way:
I resolved only to fix problems here. The base font was already pretty
tight--any further condensation would degrade readability.
Other technical repairs
=======================
Misplaced or mis-numbered glyphs in
Tamil, Gujarati, Devanagari, Hebrew Presentation Forms, Latin Extended B.
Many glyphs in Private Use area that incorrectly had Unicode numbers.
Lookup issues in Gurmukhi.
Various spline issues in Tamil.
Thanks!
=======
Special thanks for the contribution by Daniel Johnson of the Cherokee range!
Thanks also for the work of Darko Žubrinić on Glagolitic!
http://www.croatianhistory.net/etf/gl-font.html
Thanks to George Douros for his beautiful fonts!
See "Unicode Fonts for Ancient Scripts"
http://users.teilar.gr/~g1951d/
Thanks to the work and cooperation of Hiran Venugopalan
and Swathanthra Malayalam Computing on the Malayalam range
http://savannah.nongnu.org/projects/smc/
Thanks to the bug reporters! Keep 'em coming!
Recent FreeFont releases finally appear in Linux distros!
This week I installed Ubuntu Linux version 8.10 (Interepid Ibex), which was released on Oct. 30. It includes the FreeFont release of March 2008, pulled down from Debian unstable!
That amounts to approximately eight months between FreeFont releases and the bulk of its user base seeing the changes.
I'm working on another release right now, and hope to get it out before year's end. So maybe by August or so, the public will see the benefits.
Cheers, and thanks to all who contributed!
A major release has been made of GNU FreeFont, including many technical improvements, the addition and removal of some script ranges, and many bug fixes.
Download it at http://ftp.gnu.org/gnu/freefont/
Find versions dated 20080912
See below for more details, and the release ChangeLog file for even more.
Thanks to all the people who wrote in with recommendations!
==========================================================================
Main changes
So many sweeping changes were made, that they are difficult to characterize.
Bug fixes
---------
Hundreds of small fixes, mostly as reported by Jacob Poon.
Writing systems
---------------
Enforced a policy of including a writing system character range only if it serves to display text in its language. This resulted in several ranges being deleted, for now.
Added several writing systems (see Ranges)
Technical
---------
Line spacing: Serif and Mono are now within stated vertical bounds.
Various lookup table repairs in Indic ranges.
Made to pass FontForge validation (.sfd, .ttf, and .otf versions).
Moved auxiliary glyphs out of high Unicode.
==========================================================================
Testing
Used new FontForge scripts to test that Mono is really mono-spaced, and to validate all faces automatically, for .sfd, .ttf, and .otf files.
All tests pass without warnings*, using FontForge 20080429.
(* The OpenType versions produce a warning about the KERN tables when opened with FontForge. But other Windows OpenType files do this as well.)
Checked that the FreeFont installs and works minimally in Linux, Windows Vista, and Mac OS X.
Looked at test pages of all the main script ranges to see if everything was still in order.
==========================================================================
Ranges
Removed
-------
Hiragana/Katakana
since they aren't useful without Kanji.
Telugu
Oriya
Kannada
as they were missing many glyphs, and necessary ligatures, etc.
Added
-----
Hanunóo
Buginese
Tai Le
Old Persian
Ugaritic
Scaled
------
Sinhala, Thaana, Georgian to be in better agreement with rest of font.
Latin Extended accents
----------------------
Overall revision of accents in Latin extended ranges: strove to display accents without clipping. Finished Serif, Mono: Sans still has problems.
Greek, Cyrillic
---------------
Greek accents, and horizontal spacing. and letter shapes improvements.
In Serif, replaced all of Cyrillic and most of Greek with Tempora.
IPA
---
Completed in roman faces
Thai
Check them out!
http://www.gnu.org/software/freefont/
This is in anticipation of a new release of the font itself.
Particularly, note the "coverage" page, which contains a table of character ranges vs. their coverage by the different faces of FreeFont. The table was generated by a FontForge script, now in CVS.
Also, the article "Why Unicode fonts" goes into some detail into why we are making the font in the first place, and explains some new policies concerning quality and inclusion of scripts.
To do:
I have here a lot of test pages I would like to put up as well, but they really aren't fit for public consumption yet. I thought a gentle comparison of FreeFont with other Unicode families (especially DejaVu) would be in order.
I'm pleased to announce another release of GNU FreeFont.
This release addresses mostly technical issues with the fonts. Several outstanding bug reports are fixed, and some characters are added, especially Math symbols.
Download it at http://ftp.gnu.org/gnu/freefont
You want the tarballs dated 2008-03-23.
Thanks to the several people who offered advice!
A lot of changes were made, and probably a few things are screwed up. Certainly not all the pending bug reports have been dealt with.
I expect to make another release soon, dealing in more detail with bugs.
Please have a look, and drop me a line!
Changelog
===========================
A day-by-day summary of changes:
http://cvs.savannah.gnu.org/viewvc/freefont/ChangeLog?revision=1.248&root=freefont&view=markup
Main technical improvements
===========================
Brought into line with Debian ttf-freefont
Updated file format to Spline Font Database (SFD) 2
(so now we can use recent versions of FontForge)
The recent versions of FontForge are much superior to those of a
couple of years ago. Besides being much more stable, they now
do quite thorough checking for font problems. Thanks to George!
Line height improved
(it is still slightly increased: see discussion below)
Glyph outlines in Windows are now good
(see discussion below)
Now works in Windows and Mac OS
I have installed and used it in Windows Vista, and it has been
tested on Mac OS X
Path problems fixed
open paths
intersecting paths
outermost paths not clockwise
no control point at path vert. or horiz. extrema
flipped references
mixed references and contours
many "points too close" fixed
Lookup table (ATT) problems
especially in Sans: Indic ranges were a complete mess
Validation
all faces now pass FontForge validate
Back layer
removed all back layers: fixed "ghost glyph" in printing
Font Info: Manufacturer (like OS/2 Vendor ID) is now GNU
Copyright date updated
Changes to character ranges
===========================
Combining Diacriticals
Many were incorrectly placed
DPOS 'mark' table for Latin
Latin
Ligatures ff ffi ffl fi fl put in standard lookup table
Superscript digits and fractions regularized
Extended Additionals: fixed incorrect stacking of many accents
Added double hyphen uni2E17
Greek
Made 'tonos' to be the same as combining accent acute
Moved capital letters with tonos so tonos doesn't cover preceding letter
Arabic
I hope is basically functional now.
[Late correction: it somehow got out the door with some contextual character replacements that had no glyphs. The fix is already in CVS and should be in the next release.]
Added lookup tables to do contextual character replacement,
Added the required ligature look-up for lam-alif.
Offset so baseline agrees with Latin.
Japanese Hiragana, Katakana
Offset so baseline agrees with Latin.
Fixed U+30FB, KATAKANA MIDDLE DOT to be full-width
Cyrillic
Increased height of letter "i"
Indic (Devangari, Gurmurkhi, Bengali, Gujarati)
many changes of references to outlines
in Sans, many alterations to lookup tables, which are very buggy
many fixes of glyph outline problems
HTML Character Entities
completed set for all faces
Math operators
Completed LaTeX 2e range in Roman faces
Worked on extendable integrals
To do
=====
kerning (and related issues) needs to be changed to "kerning by classes"
documentation
* policies
* development tricks page
* test pages
maybe a find a better way to provide font snapshots
perhaps glyphs from http://packages.debian.org/sid/ttf-mph-2b-damase
perhaps replace Hiragana/Katakana with more modern set,
* such as Sazanami Mincho, VL-Gothic
Many ranges (Ethiopic, Japanese, Indic) need a lot of work
* many glyphs are plain ugly
* much work was done on lookups in Sans, but lookups in other faces
are probably non-functional
many outstanding bugs in the bugs list
=======================================================================
Line height
Ever since GNU Freefont started acquiring language script ranges, line height has been a problem. The underlying reason is, some of those new glyphs were very high or low.
The result was that in many applications, the line spacing (of FreeSerif especially) was extremely large.
In the Debian distribution of FreeFont, a patch was applied to remedy this. I applied this patch to the current FreeFont in CVS. However, I have since had to replace it. It had some bad side-effects: the most obvious was that in Windows, the ascenders and descenders of Latin letters were chopped off.
The current state is a compromise. The line spacing is now slightly increased (by about 15%) to allow for most accents and high and low letters.
The purpose of putting scripts from different languages into a single font can only be, to make it easy to make text in mixed languages. A criterion for them being together must therefore be that they look good together. They should therefore at least be commensurate.
Now, I found that some scripts (Japanese, Arabic, and Tamil) were greatly offset from the baseline, for no reason I could see. I have put them where I think they belong, and that helped a lot.
I hope in the future the situation can be further improved.
These scripts have some characters that are conventionally very high or low: Arabic, Thai, Tamil. There are also standard typographic practices that can make them work well with Latin.
The extended Latin ranges are a rather worse problem. There are cases of stacked accents which may be difficult to fit into a comfortable height.
Technical details:
ascent and descent: 800 EM, 200 EM.
OS/2 Win ascent and descent: 900 EM, 300 EM.
Unfortunately, this Win ascent/descent is interpreted variously. It was meant to serve simply a graphical clipping region, but many applications also take it to indicate line spacing (and some as line spacing but not clipping region).
=======================================================================
Glyph outlines in Windows
The appearance of the font in Windows has evidently always been very poor. The outlines appeared very choppy, very much like a badly-scaled bitmap.
After a great deal of detective work, comparing FreeFont to other fonts, and switching my (single home) computer between operating systems, I figured it out.
Now the outlines look quite good in my Windows Vista.
Technical details:
Windows will fail to do what is required unless some "TrueType tables" are included in the font: in SFD format, these have names like "TtTable" and "ShortTable". These were absent in FreeFont.
(Just what Windows does isn't clear to me: the tables are "hinting" tables, but the effect looks to me like anti-aliasing.)
Now, FontForge generates such tables, with the Auto Instruct command--but only for fonts with quadratic outlines. FreeFont has always had cubic outlines...so the command was disabled.
To maintain the cubic outlines (I don't know why, but let's say it is desirable), here is the trick:
Font Info -> Layers -> All layers quadratic
Auto-hint everything
Auto-instruct
Save As (different file name)
With text editor, copy the TrueType tables from the cubic .sfd file
to the quadratic .sfd file.
Note: this problem goes away if the font is bundled in OpenType format. This is the main format in Windows Vista, but many important applications in the free software world don't support it yet.
The svn/ directory of Freefont has been brought up to date with the sources in the Debian ttf-freefont package, which has been updated independently of Freefont for several years now.
I have set a CVS symbolic tag in the .svn files, named
DEBIAN_MERGE_2008
to mark the revisions that comprise this update.
With this merge, it should be possible for Freefont development to continue unhindered.
After a considerable amount of time, a new release of free UCS outline fonts is available on Savannah download area.
All the fonts now conform to MES-1 (Minimum European Subset) of Unicode/ISO 10646. In addition, thanks to Solaiman Karim, Omi Azad, K.H. Hussain and R. Chitrajan, OpenType support for Bengali (in FreeSerif, FreeSerifItalic, FreeSans and FreeSansOblique) and traditional Malayalam (FreeSerif) might be mentioned as major new features. Most reported bugs were also corrected.
We plan to continue development towards MES-2, which will bring complete support for Latin, Cyrillic and Greek alphabets. With the current pace, we will reach this goal in early 2006. Along the way, we are also improving on other areas, such as non-European Latin use, support for Indic scripts etc.