Changelog since this site went live on 2024/08/11

October 2024

• Created lipsum-CN-{kai,sans,sung} files to replace the old
  lipsum-simplified-chinese files. Unlike Japanese, Chinese has two styles of
  font which I regard as serif.
• I realised that the identifiers for the Noto Serif CJK fonts in CJK-CN-kai and
  CJK-CN-sung were the numbers for the Noto Sans CJK fonts. Corrected.

September 2024

• Updated to fonts-horai-umefont_670. I now show the Ume Hy Gothic font for
  item 1.131 because it includes blank non-breaking spaces and an identifier for
  missing codepoints. I also attempt to explain the available variations, and the
  problems with the Ume fonts for non-Japanese languages.
• Reworded tools.html and cjk-comparisons.html.
• Updated VL-Gothic to the current version.
• Looking at older CN fonts, I now find that JP and TW are not adequately covered.
  I have revised the languages files for the following: fireflysung, odokai, odosung,
  odosungmono, UKaiCN, UMingCN.
• I started to look at other Japanese fonts, then noticed some fonts admit to not
  having as many Kanzi glyphs as desired. I eventually created templates for
  testing the coverage of CN, JP and TW for a font. I also discovered that,
  particularly for Japanese, variant shapes of some Han glyphs might be
  preferred. Looking at google translate while creating these files, I finally
  understood why CJK languages usually omit 'gibberish' dummy text: almost any
  combination of syllables can have a somewhat-reasonable translation. These
  templates are in files/tools/templates and the comments mention where I got
  them from.
• Karl Berry suggested my usage of 'sc' for both Small Caps and Simplified
  Chinese could cause confusion. He suggested zh-CN. As far as I can see, only
  the files comparing shapes of codepoints are affected. I have renamed these to
  CJK-CN-kai, CJK-CN-sans, CJK-CN-sung.
• Updated the Hanazono fonts - although Debian has a newer version than what I
  used to use, the latest version, v8.030, is at github v8.030. That changed the
  fontconfig names to Hanazono Mincho A,B and added a C version (item 3.022 in
  my table) to cover later CJK extensions. Comments on how to use the B and C
  fonts are in 'HanaMinA-languages.pdf'.
• Created comparative files showing certain codepoints for Simplified Chinese:
  CJK-sans-sc, CJK-serif-sc-sung (includes Fang), CJK-serif-sc-kai.
  These all start by showing glyphs from Noto and WenQuanYi Zen Hei and are
  linked from cjk-comparisons.html. Some of those codepoints are not used in
  Simplified Chinese, I intend to do other CJK languages using similar files.
• Revised gbsn00lp, gkai00mp, UKaiCN, UMingCN, template languages-full.tex.
• Revised odokai to include angle brackets, fixed typo in odosung{,mono}.
• Revised the lipsum-greek-sans and lipsum-greek-serif files.

August 2024

• Fixed the Junicode languages files, Greek had accidentally been omitted.
• I have now revised the 'PDF-substitutes' files.
• Attempted to simplify CJK descriptions: replace Pan-CJK by the actual language,
  consistently name as (Monospace) Language Sans/Serif. For Chinese recognise
  that Serif is split into Sung, Kai, Fang.
• Revised odosung and odosungmono to include angle brackets.
• Fixed an error in the "identify CJK language part of languages-full.tex:
  I had shown U+5377, should have been U+537F as in Japanese part.
• For small caps in Azeri and Turkish I had wrongly added the combining dot above
  to a normal lowercase i instead of to a dotless i. For most fonts this makes very
  little difference to my conclusions about whether or not they are usable for Turkic
  languages. But I noticed oddities in hyphenation for Linden Hill and OFL Sorts Mill
  Goudy which suggest that at least those two fonts may give problems with
  hyphens after dotted i in Turkish small caps using XeLaTeX and polyglossia. For
  the fonts which support small caps in Azeri (for which I have no hyphenation
  patterns) the same may be true if they are used for Turkish.
• Revised the languages files for all Noto CJK fonts, both WenQuanYi Zen Hei, and
  Droid Sans Fallback to better show their coverage of the Korean Hanja I know about
  and to include angled brackets and double angled brackets for quotations.
• Added Hangul-Hanja files for Droid Sans Fallback. Some of the many missing
  codepoints in that highlighted duplications in my earlier uploads and the template,
  for 'ga' and 'joon' . I have fixed the template but will not revise the other files, the
  additions are benign and I'm sure there are other duplications.
• Added CJK-sans-sc (Simplified Chinese, Sans) files showing Han glyphs which vary
  across languages. Unfortunately, the differences seen in Serif fonts are not always
  visible in Sans fonts.
• Added a page for linking to comparisons of CJK fonts. Deleted the Korean-mixed
  template, replaced it with a Hangul-Hanja template.
• Updated the Fandol fonts to v0.3. This adds FandolFang which is a FangSong font.
  Unfortunately, I discovered that Han uses angle brackets, not angle quotes. Will need
  to revise a lot of the CJK fonts for that.
• I discovered that my version of NimbusSanL was not recognized as a low-priority alias
  for Helvetica by fontconfig. I had an old fork, now replaced by the last version
  which now has both names for fontconfig - 'Nimbus Sans L,NimbusSanL'.
• In the indexes, removed '‡' markings from monospace fonts :
  I had originally intended to delay revising those, but once I became aware of some of
  the issues I decided to fix them all before making the site live.
• Minor text changes.